Meat & Livestock Australia Limited v Cargill, Inc

MLA's contentions 641 MLA contends that claims 1, 6, 8, 9, and 12 to 14 are anticipated by the Gengler Patent, filed 11 March 2003, but claiming priority from the French priority document (11 March 2002). 642 MLA also says that if the number of SNPs used in the claims is an arbitrary limitation, then claims 3 to 5 are also anticipated by the Gengler Patent. Accordingly, MLA seeks leave to amend its Notice of Appeal to also assert anticipation of those claims by the Gengler Patent. I will allow such an amendment. But as I have already rejected MLA's arbitrariness submission, MLA's case concerning claims 3 to 5 rises or falls with what I say concerning claim 1. 643 The Gengler Patent is relied upon by MLA as a "whole of contents" anticipation because, although it was not published before the earliest priority date (31 December 2002) of the 253 Application, its claims (so MLA contends) are entitled to an earlier priority date from the French priority document. Therefore, the claims of the Gengler Patent have a priority date before the earliest priority date asserted for the 253 Application. 644 Now MLA also says that the 253 Application is not entitled to claim priority from the US priority document filed on 31 December 2002, as that document does not disclose the specified SNPs claimed. MLA says that the information regarding the specified SNPs was first included in the parent application filed on 31 December 2003. And each claim of the 253 Application expressly relies on the specified SNPs as part of the invention. Therefore MLA says that the claims of the 253 Application are not entitled to claim priority before 31 December 2003: see s 43 of the Act, and regulations 3.12(1)(b) and (c), (2)(c), (2A) and (2C) of the Patents Regulations 1991 (Cth) (as they then were). It is said that if this is the case, the claims are not novel over the Gengler Patent. 645 Further, MLA also says that if the priority date is deferred to 31 December 2003, MLA can rely on the Gengler Patent as a novelty document without the need to rely on the French priority document and apply the "whole of contents" approach, as the Gengler Patent was published on 22 September 2003. The Australian copy was apparently published on 22 September 2003, but the international version was, on its face, published on 18 September 2003. 646 Before proceeding further I should say now that none of this matters given my conclusion, as I will later explain, that the Gengler Patent does not make a clear and unmistakeable disclosure of each and every integer of the relevant impugned claims of the 253 Application. But in case I need to express my views on these subjects I would say two things. First, in my view the Gengler Patent is entitled to an earlier priority date from the French priority document. Second, given that conclusion, I do not need to decide whether the 253 Application has the earlier priority date of 31 December 2002. If the Gengler Patent had made the requisite clear and unmistakeable disclosure, then it would have anticipated the 253 Application even with the earlier priority date. Now the 253 Application may have had the earlier priority date, but if it does not it certainly has the deferred priority date of 31 December 2003; but for present purposes I do not need to decide between the two. For completeness, I would reject the other MLA contention of a 17 October 2013 priority date if it is persisted with.

Claims 1 and 3 to 5 647 MLA says that the Gengler Patent is directed to the development of methods for genotyping animals for DNA-based markers in two candidate genes for milk production, and their use in identifying (selecting) animals with desirable phenotypes for various milk-production traits. At 2, lines 27 to 29, the Gengler Patent states: "[t]he aim of the invention is to provide a method for identifying a mammal having a genotype indicative of advantageous milk production traits." The Gengler Patent specifically refers to cattle (Examples 5 and 6). 648 MLA relies upon Professor Visscher's explanation that: [t]he DNA-based markers referred to in Gengler comprise a previously-known polymorphism (with alleles A and B) in the κ-casein gene, a previously-known polymorphism (with alleles A and B) in exon 6 of the pituitary transcription factor 1 (PIT-1) gene, and another polymorphism (with alleles C and T) in exon 2 of the PIT-1 gene that appears to be novel to this document. 649 Professor Visscher also said that all three of the polymorphisms disclosed in the Gengler Patent are SNPs. 650 MLA points to the fact that the authors of the Gengler Patent stress that the novelty of their invention lies in selecting animals based on combined consideration of their genotypes in both the K-Cas and PIT-1 genes. For example, at 3, lines 24 to 26 it is stated that: "the method is more effective than known prior art methods based on the use of the PIT-1 marker alone or on the kappa-casein marker alone". 651 MLA also relies upon Professor Visscher's explanation that the Gengler Patent: discloses a method to "identify a trait in a bovine subject" from 3 SNPs in 2 genes [i.e. A and T alleles in PIT-1, and B allele in kappa-casein] that are associated with the trait as described by claim 1 of the [253 Application]. 652 Further, as to Branhaven's assertion that the Gengler Patent does not teach the use of all three SNPs together, MLA points to claim 1 of the Gengler Patent which claims: A method for identifying a mammal having a genotype that is indicative of advantageous milk production traits, comprising the following steps: (a) obtaining a biological sample comprising the DNA of said mammal; (b) analyzing the polymorphism of the PIT-1 and κ-casein genes of said mammal, in which the simultaneous presence of allele A and/or T of the PIT-1 gene and allele B of the κ-casein gene is indicative of high potential for milk production and protein production in said mammal. [emphasis added] 653 MLA says that the abstract, claim 2 and the body of the Gengler Patent (at 4, lines 1 to 5), are all in similar form. That is, they all disclose genotyping animals for all three SNPs (i.e. 2 SNPs in the PIT-1 gene and 1 SNP in the κ-casein gene) satisfying the "at least three SNPs in more than one gene" requirement of claim 1 of the 253 Application. MLA says that it does not matter that the examples in the Gengler Patent do not disclose use of all three SNPs together since they are merely experiments leading to the invention of all three SNPs and, in any event, the disclosure is not limited to the experiments that are performed. 654 Now MLA refers to the fact that Professor Plastow originally gave evidence that there was no point in using a SNP which was in high LD with another SNP but that he conceded that the Gengler Patent is drafted to cover the use of 3 SNPs together. Further, Professor Visscher explained why that would be useful and it is said that Professor Plastow did not refute that. Accordingly, MLA says that in the circumstances, not only does the Gengler Patent direct the use of three SNPs at once to infer a trait, but that it would be understood by a person skilled in the art as logical to do so. 655 Further, MLA contends that on the basis of the information provided by Professor Hayes, the PIT-1 and K-Cas genes on the bovine genome fall within 500,000 bp of specified SNPs of the 253 Application. In particular, based on the mapping results presented by Professor Hayes, the SNP disclosed in SEQ ID No: 20015 of the 253 Application ("SNP20015") is located at base 35,186,040 of chromosome 1, and therefore the entire PIT-1 gene lies within 500 kb of SNP20015. Given that the two SNPs in PIT-1 are located in the gene (one in exon 6 and the other in exon 2), the two SNPs must also lie within 500 kb of SNP20015. Furthermore, the SNP disclosed in SEQ ID No: 20139 of the 253 Application ("SNP20139") is located at base 87,471,038 of chromosome 6, and therefore the κ-casein gene lies within 500 kb of SNP20139. Given that the SNP in κ-casein is located in the gene (at nucleotide position 5345), the SNP must also lie within 500 kb of SNP20139. Accordingly, MLA says that the Gengler Patent claims the features of claim 1 of the 253 Application, with an earlier priority date. Further, MLA contends that if the number of SNPs used is an arbitrary parameter, then claims 3 to 5 are also anticipated. As I have said, I have rejected that submission of arbitrariness.

Claim 6 656 With respect to claim 6 of the 253 Application, MLA contends that by carrying out the method claimed in claim 1 of the Gengler Patent, animals are identified as having advantageous milk production qualities. That implicitly provides a method for sorting of those animals. In any event, it is said that the method in the Gengler Patent is suitable for that purpose.

Claims 8 and 12 657 With respect to claims 8 and 12 of the 253 application, the Gengler Patent states at 7, lines 23 to 24 that "… other techniques can be used to determine polymorphism in the genes under consideration …". MLA contends that these other techniques would be understood by the skilled addressee to include standard nucleic acid hybridisation as set out in claims 8 and 12 of the 253 Application, as Professor Visscher explained.

Claim 9 658 Claim 1 of the Gengler Patent concerns the "high potential for milk production and protein production". MLA contends that this falls within the list of traits in claim 9 of the 253 Application (e.g. milk production and milk quality).

Claim 14 659 As to claim 14, MLA contends that the methods disclosed in the Gengler Patent for identifying the SNP in the K-Cas gene (Example 1) and the SNP in exon 6 of the PIT-1 gene (Example 3) each involve isolation of genomic DNA and PCR amplification of a fragment of DNA containing the SNP.

Analysis of the Gengler Patent 660 I would also reject MLA's lack of novelty case concerning the Gengler Patent. 661 In my view, even if the Gengler Patent forms part of the relevant prior art base, it does not anticipate. 662 The Gengler Patent discloses a candidate gene approach to identify animals possessing certain milk production traits; I accept though that the claims of the 253 Application are not limited by the approach taken to get to the 3 SNPs. There are two candidate genes disclosed, PIT-1 and K-Cas. The Gengler Patent describes 2 different SNPs in the PIT-1 gene and 1 SNP in the K-Cas gene. The authors test the effect of the K-Cas SNP alone, or in combination with either the exon 2 SNP of PIT-1 or the exon 6 SNP of PIT-1. 663 The method of the invention is described in terms of a single method using both the PIT-1 marker and the K-Cas marker: [T]he inventors have developed a method for identifying a mammal that has a genotype indicative of advantageous milk production traits, in which both the PIT-1 gene and the kappa-casein gene are used as genetic markers. They have shown that this method is more effective than known prior art methods based on the use of the PIT-1 marker alone or on the kappa-casein marker alone. 664 As to exemplification, the inventors show in example 6 that one PIT-1 SNP combined with the K-Cas SNP gives a better result than either SNP alone. Further, in example 7, they show that the other PIT-1 SNP combined with the K-Cas SNP gives a better result than either SNP alone. But only in the relevant claims and the consistory clauses do the inventors use the term "and/or" to encompass use of one or other or both of the SNPs in the PIT-1 gene together with the SNP in the K-Cas gene. 665 First, I agree with Branhaven that the inventors are not disclosing 2 separate methods to identify certain milk production qualities, but a single method that is capable of being carried out using 2 SNPs (1 of the 2 PIT-1 SNPs and the K-Cas SNP) following all the examples in the Gengler Patent. 666 Now I would note that in relation to the evidence before me, Professor Goddard accepted that in the examples in the Gengler Patent, the inventors only carry out the method with 2 SNPs, one or other of the PIT-1 SNPs and the K-Cas SNP. He accepted that the inventors do not exemplify the method using both PIT-1 SNPs. 667 Indeed, Professor Taylor explained that the reason that only one of the two SNPs in the PIT-1 gene is used is because the PIT-1 SNPs are in tight linkage disequilibrium with each other. Professor Plastow also explained that because it was well apparent from the disclosure in the Gengler Patent that the 2 SNPs in the PIT-1 gene were in linkage disequilibrium, one would only use one or the other. Dr Sonstegard also confirmed that because the two PIT-1 SNPs are so tightly linked, there would be no reason to test for both. 668 Now I accept that Professor Visscher gave evidence that "it would not always be the case that the one [SNP] at exon 6 predicts perfectly the one on exon 2", so you would get some extra information from using both the PIT-1 SNPs. But the inventors for the Gengler Patent apparently did not think this necessary. Further, I agree with Branhaven that even if Professor Visscher's evidence is accepted, it only rises as high as supporting the proposition that there was an alternative way of carrying out the method to the way the method was carried out in the examples in the Gengler Patent, which was to use 3 SNPs instead of 2. In other words, using 3 SNPs is only one option in carrying out the method disclosed in the Gengler Patent. Alternatively expressed, the Gengler Patent contains a direction which is at least as likely to be carried out using 2 SNPs as it would be using 3 SNPs. Indeed, the evidence of Professor Plastow, Professor Taylor and Dr Sonstegard supports the conclusion that the method of the Gengler Patent would be more likely to be carried out using 2 SNPs, rather than 3 SNPs. 669 Second, the Gengler Patent does not disclose nor is compelling evidence given of an implicit disclosure that any of the 3 SNPs disclosed in the Gengler Patent are in linkage disequilibrium with any of the 2,510 SNPs of the 253 Application. 670 In summary, in my view there is not a clear and unmistakable direction to use 3 SNPs, an integer in claim 1 of the 253 Application. For at least this reason, the Gengler Patent is not an anticipation of the relevant claims. It is not inevitable that a skilled person following the Gengler Patent would necessarily follow a method that would infringe claim 1 of the 253 Application. A similar conclusion follows concerning claims 3 to 5, 6, 8, 9 and 12 to 14 which have the 3 SNPs (or indeed greater number) requirement. 671 Finally, given my conclusion I do not need to resolve priority date questions as I have said.

(f) Other matters 672 At one stage, MLA relied upon the Gelhaus publication (Gelhaus A and Förster B, "Cattle MHC Genes DOA and DOB: Sequence Polymorphisms and Assignments to the Class IIb Region" (2001) 28 European Journal of Immunogenetics 429), but this now appears to have been abandoned on this aspect at least. Nevertheless as my attention was drawn to it, it is convenient to dispose of it. 673 Gelhaus reported on a study of the genetic polymorphism of second exons of the cattle DOA and DOB genes. As it stated, two and four allelic variants were detected. The SNPs detected enabled the relevant loci to be used as markers in trait analysis. The context was a consideration of genetic variants that might influence disease resistance in cattle; unsurprisingly therefore, Gelhaus was reported in the European Journal of Immunogenetics. 674 The conclusion of Gelhaus was distinctly stated at 432 in the following terms: The polymorphisms detected in the BoLA-DOA and -DOB genes allow these loci to be used as markers in genetic studies on disease resistance and related traits. Assuming functionality as in sheep, mouse and human, further characterization of the gene products is likely to contribute to a better understanding of the genetic control of immune responses in cattle. 675 Now no association study had been carried out although clearly Gelhaus was teaching that one could test a SNP(s) with an association study. 676 But there is no express disclosure of the method of claim 1 of the 253 Application or limb (a) or limb (b) thereof.

(g) Conclusion 677 MLA's ground of opposition concerning a lack of novelty fails. Its arguments did not even approach satisfying the high threshold for establishing such a ground on this appeal, as I discussed at the outset of my reasons.

(a) Some legal principles 678 The question is whether the claimed invention lacks an inventive step over the prior art base. An invention is taken to involve an inventive step when compared to the prior art base unless it would have been obvious to a person skilled in the relevant art in light of common general knowledge as described in Minnesota Mining and Manufacturing Co v Beiersdorf (Australia) Ltd (1980) 144 CLR 253 at 292 per Aickin J (Minnesota Mining) as it existed in the patent area (the then s 7(2) of the Act) before the priority date, whether that knowledge is considered separately or together with information of the kind described in the then s 7(3). 679 For the present context, the applicable provisions of ss 7(2) and (3) are the following: (2) For the purposes of this Act, an invention is to be taken to involve an inventive step when compared with the prior art base unless the invention would have been obvious to a person skilled in the relevant art in the light of the common general knowledge as it existed in the patent area before the priority date of the relevant claim, whether that knowledge is considered separately or together with the information mentioned in subsection (3). (3) The information for the purposes of subsection (2) is: (a) any single piece of prior art information; or (b) a combination of any 2 or more pieces of prior art information; being information that the skilled person mentioned in subsection (2) could, before the priority date of the relevant claim, be reasonably expected to have ascertained, understood, regarded as relevant and, in the case of information mentioned in paragraph (b), combined as mentioned in that paragraph. 680 The term "obvious" means "very plain" (Aktiebolaget Hässle v Alphapharm Pty Ltd (2002) 212 CLR 411 at [34] per Gleeson CJ, Gaudron, Gummow and Hayne JJ (Aktiebolaget Hässle) and Lockwood (No 2) at [51]). The inventive element needed to sustain a patent can be small. A "scintilla of inventiveness" will be sufficient and "no smallness or simplicity will prevent a patent being good" (Meyers Taylor Pty Ltd v Vicarr Industries Ltd (1977) 137 CLR 228 at 249 per Aickin J). Relevant further content has been given to determining obviousness in The Wellcome Foundation Ltd v VR Laboratories (Aust) Pty Ltd (1981) 148 CLR 262 at 286 per Aickin J in stating: whether the hypothetical addressee faced with the same problem would have taken as a matter of routine whatever steps might have led from the prior art to the invention, whether they be the steps of the inventor or not. 681 Further, in relation to experiments, his Honour said (at 280 and 281): In the present case it was admitted by the respondent that the test of obviousness was an objective one, but it was argued that evidence of a subjective character was admissible. That is no doubt true in some cases because expert witnesses are often properly asked whether they found a particular invention "surprising" to them. That however does not answer the question whether evidence of the steps which the patentee took is relevant and therefore admissible. Evidence of what was in the patentee's mind may be admissible as evidence of the state of the art, but could seldom be otherwise admissible. Evidence of what he did by way of experiment may be another matter. It might show that the experiments devised for the purpose were part of an inventive step. Alternatively it might show that the experiments were of a routine character which the uninventive worker in the field would try as a matter of course. The latter could be relevant though not decisive in every case. It may be that the perception of the true nature of the problem was the inventive step which, once taken, revealed that straightforward experiments will provide the solution. It will always be necessary to distinguish between experiments leading to an invention and subsequent experiments for checking and testing the product or process the subject of the invention. The latter would not be material to obviousness but might be material to the question of utility. 682 The question to be posed was whether putative experiments leading from the prior art to the invention as claimed were part of the inventive step or were of a routine character to be tried as a matter of course. That question has an affinity with the Cripps question posed in Olin Mathieson Chemical Corporation v Biorex Laboratories Ltd [1970] RPC 157 and paraphrased by French CJ in AstraZeneca AB v Apotex Pty Ltd (2015) 257 CLR 356 at [15] (AstraZeneca) in the following terms: Would the notional research group at the relevant date, in all the circumstances, which include a knowledge of all the relevant prior art and of the facts of the nature and success of [the existing compound], directly be led as a matter of course to try [the claimed inventive step] in the expectation that it might well produce a useful alternative to or better drug than [the existing compound]? 683 Now the question does not import as a criterion that the inventive step claimed would be perceived by the hypothetical addressee as "worth a try" or "obvious to try" (AstraZeneca per French CJ at [15]). Further, Aktiebolaget Hässle (at [72]) rejected the adoption of a criterion expressed in terms of "obvious to try" or "worth a try". 684 Let me say something further about common general knowledge. 685 Section 7(2) first requires consideration of what would have been obvious to a person skilled in the relevant art in the light of the common general knowledge as it existed before the priority date, putting to one side for the moment s 7(3) information. Common general knowledge is knowledge "generally known and accepted without question by the bulk of those who are engaged in the particular art" (British Acoustic Films Ld v Nettlefold Productions (1936) 53 RPC 221 at 250 per Luxmoore J). Information cannot be treated as part of the common general knowledge unless there is evidence of its general acceptance and assimilation by persons skilled in the art. Information does not constitute common general knowledge merely because it might be found, for example, in a journal, even if widely read by such persons. Further, as Luxmoore J said at 250: In my judgment it is not sufficient to prove common general knowledge that a particular disclosure is made in an article, or series of articles, in a scientific journal, no matter how wide the circulation of that journal may be, in the absence of any evidence that the disclosure is accepted generally by those who are engaged in the art to which the disclosure relates. A piece of particular knowledge as disclosed in a scientific paper does not become common general knowledge merely because it is widely read, and still less because it is widely circulated. Such a piece of knowledge only becomes general knowledge when it is generally known and accepted without question by the bulk of those who are engaged in the particular art; in other words, when it becomes part of their common stock of knowledge relating to the art. Whatever else common general knowledge may be, it has never in my judgment included public knowledge of particular documents reports or scientific papers and the like. The knowledge of a number of individuals that a particular suggestion or particular suggestions has or have been made for the use of biasing in a particular apparatus, or a number of particular apparatus, cannot be held to be common general knowledge. It is certainly difficult to appreciate how the use of something which has in fact never been used in a particular art can ever be held to be common general knowledge in the art. 686 Further, and as stated in Minnesota Mining at 293 per Aickin J, the notion of common general knowledge: involves the use of that which is known or used by those in the relevant trade. It forms the background knowledge and experience which is available to all in the trade in considering the making of new products, or the making of improvements in old, and it must be treated as being used by an individual as a general body of knowledge. So it must be knowledge that is known and available to all in the trade or at least the bulk of those who are engaged in the relevant art. Accordingly, information ascertainable by a routine literature search is not of itself taken to be common general knowledge. And patent specifications do not form part of common general knowledge without evidence that they have been absorbed into common general knowledge. 687 And as further elucidated by Jagot J in Gilead Sciences Pty Ltd v Idenix Pharmaceuticals LLC (2016) 117 IPR 252; [2016] FCA 169 at [216] (affirmed on appeal in Idenix Pharmaceuticals LLC v Gilead Sciences Pty Ltd [2017] FCAFC 196 per Nicholas, Beach and Burley JJ), it is erroneous to treat a document as being part of common general knowledge simply because skilled persons could readily locate and assimilate its contents. Her Honour went on to explain (at [217]): It may be accepted that instant recall of an article is not required. This does not mean, however, that documents found by searching for a subject-matter, rather than by some form of recall or reminder of what is already known to exist, are common general knowledge. This is so irrespective of the fact that experts in the field read widely. Further, it is not the case that mere publication and republication proves that a document and its contents have entered the common general knowledge. Nor is it the fact that a document and its contents necessarily form part of the common general knowledge merely because one expert knows or has managed to locate it and assimilate its contents. Such a document may or may not form part of the common general knowledge. The relevant inferences are to be drawn on the basis of the whole of the evidence. 688 There is also a further point to be made concerning common general knowledge. There is no general principle permitting admissions in the specification of a patent to be used to establish in and of themselves that information is common general knowledge. Whether information has become so widely assimilated that it forms part of common general knowledge must be determined on the evidence, although admissions can be considered as part of that evidence. But I would note in this context that [0211] of the 253 Application provided that: Reference to any prior art in the specification is not, and should not be taken as, an acknowledgment, or any form of suggestion, that this prior art forms part of the common general knowledge in Australia or any other jurisdiction or that this prior art could reasonably be expected to be ascertained, understood and regarded as relevant by a person skilled in the art. 689 Now in addition to using common general knowledge on a stand-alone basis, common general knowledge can be aggregated with s 7(3) information. I have already set out s 7(3) in its then applicable form. That part of the prior art base which is common general knowledge and the information referred to in s 7(3) are considered for the purpose of looking forward from the prior art base to see what the skilled person is likely to have done when faced with a particular problem. In a case where the problem is known and is part of the common general knowledge, the problem may be similar to that which the patentee claims to have solved with the claimed invention. But where the problem addressed by the patentee does not form part of common general knowledge, the relevant starting point is the prior art base, but not including the problem as identified in the patent specification. As Besanko, Foster, Nicholas and Yates JJ noted in AstraZeneca AB v Apotex Pty Ltd (2014) 226 FCR 324 at [203]: If the problem addressed by a patent specification is itself common general knowledge, or if knowledge of the problem is s 7(3) information, then such knowledge or information will be attributed to the hypothetical person skilled in the art for the purpose of assessing obviousness. But if the problem cannot be attributed to the hypothetical person skilled in the art in either of these ways then it is not permissible to attribute a knowledge of the problem on the basis of the inventor's "starting point" such as might be gleaned from a reading of the complete specification as a whole. 690 The purpose of the inquiry is to determine whether the invention is obvious to solve the perceived problem, looking forward from the prior art base. But of course this may not have been the patentee's starting point. 691 Now impermissible hindsight should be avoided in determining whether a claimed invention lacks an inventive step. The plurality in Aktiebolaget Hässle said (at [21]): The defendant to an infringement action who cross-claims for revocation on the ground of obviousness bears the onus of establishing that case. This obliges the defendant to lead evidence looking back to the priority date, sometimes, as here, many years before trial. In those circumstances, the warnings in the authorities against the misuse of hindsight are not to be repeated as but prefatory averments and statements of trite law. The danger of such misuse will be particularly acute where what is claimed is a new and inventive combination for the interaction of integers, some or all of which are known. It is worth repeating what was said by Lord Diplock in Technograph Printed Circuits Ltd v Mills & Rockley (Electronics) Ltd [[1972] RPC 346 at 362] … 692 Further, for a combination invention, the question is whether the combination, not each integer, is obvious. It is simply impermissible to take any one integer or take each integer seriatim and ask whether each integer involved an inventive step. The invention is the combination. The combination is what is claimed as the monopoly. And the relevant question is whether the combination involves an inventive step (see Lockwood (No 2) at [69]). As the plurality in Aktiebolaget Hässle (at [41]) said: … The claim is for a combination, the interaction between the integers of which is the essential requirement for the presence of an inventive step. It is the selection of the integers out of "perhaps many possibilities" which must be shown by Alphapharm to be obvious, bearing in mind that the selection of the integers in which the invention lies can be expected to be a process necessarily involving rejection of other possible integers. This expression of the issue follows what was said by Aickin J in Minnesota Mining. (citation omitted) 693 Further, the misuse of hindsight is most common in relation to combination claims (Minnesota Mining at 293 per Aickin J). 694 Further as to combination claims, Lord Davey stated in In the matter of Klaber's Patent (1906) 23 RPC 461 at 469 (in terms approved by Dixon J in Palmer v Dunlop Perdriau Rubber Co Ltd (1937) 59 CLR 30 at 73): A proper combination for a patent is the union of two or more integers, every one of which elements may be perfectly old, for the production of one object which is either new, or at any rate is for effecting an old object in a more convenient, cheaper or more useful way. But the point in a combination patent must always be that the elements of which the combination is composed are combined together so as to produce one result.

(b) Common general knowledge 695 MLA contends, which I of course accept, that common general knowledge constitutes part of the mental equipment of those concerned in the art under consideration being the background knowledge and experience which is available to all in the trade in considering the making of new products, or the making of improvements in old. 696 MLA further contends that where the specification itself describes matters as being known, I should proceed on the basis that such an admission is correct (see Bristol-Myers Squibb Co v F H Faulding & Co Ltd (2000) 97 FCR 524 at [30]) and indeed accept that to be common general knowledge. That submission overreaches. I would make two brief observations consistently with what I have already intimated. First, for the patentee to say that something is known as at the priority date does not necessarily entail that it was common general knowledge in the context of applying the statutory concept in s 7(2). Second, if an admission does arise, it is only part of the evidence for me to consider, attributing to the admission what I consider to be its appropriate weight. Statements in a specification cannot be definitive of the question. 697 MLA contends that the following eight matters were common general knowledge as at the earliest priority date of 31 December 2002 (and deferred priority date of 31 December 2003) of the claims of the 253 Application. It is convenient to deal with each matter and my response in the order advanced by MLA. As I have previously stated, where I use the expression "priority date", it is to be taken as a reference to either the earliest priority date or the deferred priority date unless I state otherwise.

Need for improved methods for predicting genetic potential in animals 698 MLA says that there was a need for improved methods for predicting genetic potential in animals for numerous quantitative traits which were known to be desirable. Further, it points to the fact that the specification states at [0022] that "three different experimental approaches have been used with limited success to identify genes, chromosomal regions or DNA markers that account for a large proportion of the genetic variation observed in economically important traits in livestock species" and that the results "have not been widely utilized to date because they do not account for enough of the total genetic variation to allow accurate prediction of an animal's performance for a specific trait." MLA says that this accords with the evidence of the experts that the approaches at the time were of limited value for predicting an animal's performance with respect to the majority of quantitative traits. 699 MLA says that by the late 1990s/early 2000s, there was a lot of interest in using molecular genomics in cattle breeding and that was where research and the industry was heading. It contends that ultimately the evidence of all experts was that before December 2002, DNA markers "were the way to go". 700 MLA also says that in the review published in February 2001 of Andersson L, "Genetic Dissection of Phenotypic Diversity in Farm Animals" (2001) Nat Rev Genet 130 (cited in Wiggans GR, Cole JB, Hubbard SM and Sonstegard TS, "Genomic Selection in Dairy Cattle: The USDA Experience" (2017) 5 Annu Rev Anim Biosci 309), the author said that "[i]t is likely that large-scale marker analysis will be used routinely, as soon as the cost for genotyping has been reduced by a factor of around ten" (at 136 and 137, second column, second last line). The review also said that "[l]inkage disequilibrium mapping will be a very powerful approach for mapping and finding trait loci in domestic animals once dense SNP maps become available and the cost for genotyping has been reduced such that genome scans using thousands of SNPs can be done" (at 137, first column, second full para). The review also observed that "[i]t is also only a matter of time before initiatives will be taken to sequence the genomes of farm animals. This will most probably be carried out using a whole genome shotgun approach …" (at 137, second column, first full para). MLA says that this review was generally consistent with the views expressed by Dr Sonstegard in his 2001 review of dairy cattle genomics (Sonstegard TS, Van Tassell CP and Ashwell MS, "Dairy Cattle Genomics: Tools to Accelerate Genetic Improvement?" (2001) 79 J Anim Sci E307). 701 Further, MLA points to the fact that prior to the priority date, Professor Taylor had set up Genomic FX to do something which was similar to what is described in the 253 Application. It asserts that his company's aim was to develop a simple diagnostic test for genotyping animals for the purpose of improving their genetic potential. His plan was to scan the genomes of thousands of animals and to identify SNPs evenly spaced throughout the genome and find SNPs that correlated with useful traits. It is said that his company did not proceed with the plan because it was forced to "shut its doors". MLA says that the only difference between the work of Genomic FX and the 253 Application that Professor Taylor could identify was that he was looking at within family QTLs, not the whole bovine genome. But MLA says that the only difference between the within family QTL approach and the genomic selection approach is that GS looks at markers across the whole genome, whereas the within family QTL approach only looks at segments of the genome. Otherwise, so MLA contends, Professor Taylor's proposed approach before the priority date was the same as that of the 253 Application. 702 I would make the following observations in response to these contentions. 703 It is not in dispute that the state of the art as at the priority date was that there were a variety of methods being developed and used to determine the genetic potential of bovines for economic traits, including progeny testing, the candidate gene approach and QTL mapping. 704 It is also not in dispute that as at the priority date, traditional phenotypic selection techniques such as progeny testing were commonly used in practice and that those techniques were sophisticated and being continuously improved. Further, as at the priority date there was limited use of molecular genetic techniques. And at that time the laboratory use of molecular genetic techniques was directed more to the candidate gene approach and the within family QTL mapping approach. 705 As I have previously summarised, the candidate gene approach is based on knowledge of the gene's biological function and its relationship to the trait of interest. It is typically based on knowledge of analogous genes in other species. As Dr Sonstegard said, it was a popular approach amongst researchers at the time to identify genes responsible for quantitative traits. Further, as I have already previously summarised, QTL mapping typically involved genotyping and trait-association experiments conducted within families to identify particular chromosomal regions associated with a trait of interest. Further, Dr Sonstegard also gave evidence that when the candidate gene approach was combined with QTL mapping, the approach was referred to as a "positional candidate gene approach". The search for candidate genes was limited to a specific region of the genome that was known to be associated with a trait (i.e. the QTL). 706 It is also not in dispute that as at the priority date neither QTL mapping nor the candidate gene approach involved the use of genome wide dense markers. Such approaches were rather focused on identifying the location of a causative gene(s) within a specific region of a chromosome. 707 Now what I have just said was reflected in the research work at the time, indeed even after the priority date, being conducted by the experts who were called before me. Professor Goddard's publication closest to the priority date involved the fine mapping of a QTL locus for twinning rate in cattle which had all bulls genotyped for 15 markers located on a specific region of chromosome 5, but did not disclose the construction or use of a dense genome-wide panel of markers or map (see Meuwissen T, Karlsen A, Lien A, Olsaker I and Goddard ME "Fine Mapping of a Quantitative Trait Locus for Twinning Rate Using Combined Linkage and Linkage Disequilibrium Mapping" (2002) 161 Genetics 373 (accepted for publication on 11 February 2002 and published in May 2002)). Further, Professor Visscher's work in bovines until at least 2005 predominantly involved family-based linkage studies using microsatellite markers. Further, Dr Sonstegard was involved in the fine-mapping and sequencing of a QTL on bovine chromosome 6 and identified osteopontin as a candidate gene affecting milk production in or around 2005, which research was published in the Proceedings of the National Academy of Sciences in 2005 (Schnabel RD, Kim JJ, Ashwell MS, Sonstegard TS, Van Tassell CP, Connor EE and Taylor JF, "Fine-Mapping Milk Production Quantitative Trait Loci on BTA6: Analysis of the Bovine Osteopontin Gene" (2005) 102 PNAS 6896). Further, in evidence before me reference was made to a well-known study in 2001 in which the positional candidate gene approach was successfully used to identify a causative polymorphism in a gene called DGAT1 which accounted for about 30% of the variation in particular milk traits (Grisart B, Coppieters W, Farnir F, Karim L, Ford C, Berzi P, Cambisano N, Mni M, Reid S, Simon P, Spelman R, Georges M and Snell R, "Positional Candidate Cloning of a QTL in Dairy Cattle: Identification of a Missense Mutation in the Bovine DGAT1 Gene with Major Effect on Milk Yield and Consumption" (2001) 12 Genome Research 222). Further, Professor Plastow gave evidence about his use of the candidate gene approach on the melanocortin receptor 4 gene (MC4R), published in 2000 (Kim KS, Larsen N, Short T, Plastow G and Rothschild MF, "A Missense Variant of the Porcine Melanocortin-4 Receptor (MC4R) Gene is Associated with Fatness, Growth and Feed Intake Traits" (2000) 11 Mamm Genome 131). The MC4R candidate gene explained variation in a polygenic trait, being growth and feed intake. And according to Professor Plastow, although this gene explained a very small amount of the variation in the trait, the ability to identify variants through the use of markers was commercially very valuable. Further, I note for completeness that Professor Plastow later discussed in a 2005 article the MC4R gene, the use of functional genomics studies to identify candidate genes and also high-density marker approaches (described as Phase 3), none of which was inconsistent with his evidence before me as to what was relevantly known or expected as at the priority date (van der Steen, H, Prall G and Plastow GS "Application of Genomics to the Pork Industry" (2005) J Anim Sci 83:E1). 708 Now MLA has made reference to Professor Taylor's involvement in Genomic FX prior to the priority date. But the described objective of Genomic FX was to use within family QTL mapping with fine marker mapping to identify predictive variants or positional cloning, rather than to scan the genomes of thousands of animals and to identify SNPs evenly spaced throughout the genome. And in any event such work did not proceed. Now I accept that positional cloning was one of the methods being used in laboratories before the priority date, but this method was not a whole of genome approach. 709 Further, the evidence of Professor Taylor and Dr Sonstegard was that although there was interest in genetic markers, progeny testing was still a very common technique in the industry as at the priority date. 710 Further, and as I have endeavoured to explain previously, in contrast to the candidate gene and QTL mapping approaches, a genome wide approach does not require identifying any particular causal gene or region of the chromosome. And the possibility of using genome wide approaches in livestock breeding was theoretical as at the priority date, with it being unconfirmed whether such approaches could successfully be applied in cattle. Further, the tools required such as a genome-wide marker map, were not available. And the absence of a sufficient number of markers was a technical hurdle. Further, it was speculation whether a SNP chip would become available. 711 As at the priority date, to apply a genome wide approach it would have been necessary to identify a large number of genetic markers evenly spaced throughout the genome. And to design and carry out the necessary experiments would have presented many challenges, particularly when as at the priority date, work on sequencing the whole bovine genome had yet to commence. As Dr Sonstegard gave evidence of, which I accept, the challenges included the need to sequence and assemble a large proportion of the bovine genome, the need to identify a large number of informative markers, the need to determine the relative position of those markers, the need to generate a genome-wide panel of relatively evenly spaced markers, and the need to identify those markers that are associated with a trait. I would just interpolate at this point that MLA spent a considerable time attacking Dr Sonstegard's evidence by reference to observations made in articles that he had co-authored which discussed in part genomic selection and the problems involved and resources required. But I do not consider that MLA established at all that Dr Sonstegard's evidence was not reliable on what was known or expected as at the priority date on such matters. 712 Further, on the evidence before me, it is not seriously in doubt that as at the priority date, the relevant work was technically difficult, expensive and time consuming. The then existing techniques were inadequate. They did not provide the required specificity for SNP identification, and certainly not on a large scale. Further, SNP chips were not developed until the mid to late 2000s. I would note that they were not used by the inventors of the 253 Application. Further, as Professor Plastow explained, the necessary high throughput assay technology to screen multiple bovine markers was not developed until well after the priority date. 713 Further, the preponderance of the evidence also established that as at the priority date, microsatellites were the most widely used genetic marker in bovines. They had various advantages over other markers, including SNPs, as various of the experts accepted. Further, fragment analysers only became available in the early 2000s to enable the rate at which microsatellites could be genotyped to be substantially increased. Contrastingly, SNPs were seldom used as a genetic marker, and as at the priority date very few had been discovered in cattle. Moreover, it was only after the priority date that a significant transition towards the use of SNP markers occurred when the BovineSNP50 SNP chip was commercially released. 714 Further, according to Professor Plastow, even by 2005 the consensus was that there were still significant hurdles to be overcome before a genome wide approach could be utilised in the breeding selection and management of cattle. 715 Further, I would also note that Professor Visscher for MLA accepted that the main practical limitation to implementing a genomic selection approach as at December 2002 was the identification of a sufficient number of markers. Now although Professor Visscher said that there was no technical difficulty at the time in assaying or genotyping each individual marker, that is different to what was considered to be the main practical limitations that I have outlined. In any event, this is a reference to genotyping individual markers, rather than a genome-wide panel of markers. Now Professor Visscher accepted that the information in Vignal (Vignal A, Milan D, San Cristobal M and Eggen A, "A Review on SNP and Other Types of Molecular Markers and Their Use in Animal Genetics" (2002) 34 Genet Sel Evol 275) was consistent with his understanding of what was known prior to December 2002 about the use of genetic markers in animal genetics. But in my view Vignal does not assist MLA. Vignal does not describe the use of genomic selection as the direction of research in 2002. And I note that it does not cite the Meuwissen paper; MLA asserted that Meuwissen was well known at the time. Vignal does not suggest producing a genome wide map of evenly spaced markers. And it does not suggest how such a map could be constructed and applied. Further, Professor Visscher accepted the reference to "high densities of markers" being needed was in the context of a fine QTL mapping approach. It is well apparent that the expressed desire for more markers within the region of a QTL is different from a genome wide dense marker map necessary to carry out a GWAS or genomic selection. The purpose of identifying more markers for a fine QTL mapping approach is to more accurately locate the causal gene. But contrastingly, the GWAS and genomic selection approaches do not involve determining the identity or location of any causal mutation. 716 Further, what is apparent from Vignal, consistently with the expert evidence led before me, is that at the priority date, to the extent that genetic markers were being used, they were microsatellites. Now it might have been accepted that in the future SNPs might produce equivalent information, but that uncertainty does not suggest any expectation that SNPs could be used successfully (as claimed in the 253 Application). It is convenient to set out some extracts from Vignal. 717 Vignal, which was published in 2002 (after 8 March 2002 (the date of acceptance)), stated in section 2.2: What is the reason for the increasing popularity of SNPs, whereas in terms of genetic information provided, as simple bi-allelic co-dominant markers, they can be considered as a step backwards when compared to the highly informative multi-allelic microsatellites? Are we not only putting a new name on what has just been considered until now as a common polymorphism and originally studied as RFLPs? In fact, the more recent SNP concept has basically arisen from the recent need for very high densities of genetic markers for the studies of multifactorial diseases, and the recent progress in polymorphism detection and genotyping techniques. 718 Further, Vignal from section 6.3 and in conclusion stated: Several approaches can be taken for fine QTL mapping, such as increasing the number of meiosis events by increasing the size and/or the number of families for genotyping, selecting recombination events in recurrent backcrosses, using advanced intercross lines (AIL) or performing linkage disequilibrium and haplotype-based studies in outbreed populations. However, whatever the approach taken, high densities of markers will be needed. In some instances, when the populations studied are closely related, even the microsatellite markers may not be heterozygous for the F1 animals. Also, for some species, such as chickens, the density of microsatellites will be low. Testing of candidate genes and candidate polymorphisms in exons, promoters or other important regions such as splice sites, promoters or other regulatory regions, will have to be done using the SNP approach, since this will be the most common polymorphism and the more likely responsible for phenotypic variation. When testing for the association between complex phenotypic traits and candidate loci, single-loci SNP analyses present a loss of information due to the bi-allelic nature of the markers, as compared to the multi-allelic microsatellites. However, by performing haplotype frequency estimations over several SNPs from a locus, this can be overcome and even possibly improved, due to the fact that SNPs will more often be close to the site responsible for the variation than microsatellites. 7. CONCLUSION Although in a strict molecular sense, SNPs are just what has been previously known as base substitutions, the fact of naming molecular markers by this acronym meaning single nucleotide polymorphism, is an indication of the new importance that this type of polymorphism has in molecular genetics. Indeed, if in some instances, the lack of information due to the bi-allelic nature of SNPs is a limitation, there are cases in which they can provide valuable data on associations between specific genes or other DNA structures and phenotypes, or on population and genome dynamics. The very high density of SNPs in genomes, usually allows to develop several of them in a single locus of a few hundred base pairs. By reconstructing haplotypes, multi-allelic systems can eventually be defined for analyses, to overcome the limitations due to the low heterozygosity of SNPs. With increasing progress being made in the molecular techniques used to produce SNP data, in the automation of allele scoring and in the development of algorithms for genetic analyses, the effort needed to produce an equivalent amount of information as with microsatellites may some day be equivalent. 719 There is little doubt that the broader use of SNPs was expected to evolve. 720 In summary, although I readily accept that as at the priority date there had always been a desire to improve breeding methods in domesticated animals, it is readily apparent that at that time traditional non-molecular genetic approaches were the most widely used. Moreover, if genetic markers were used to identify causative genes, these markers were principally microsatellites.

Venter 721 Let me now deal with the next topic concerning common general knowledge. 722 MLA says that by the priority date, the Venter paper (Venter JC et al, "The Sequence of the Human Genome" (2001) 291 Science 1304) was well known. This paper was published in February 2001 in Science. Venter reveals that with the help of Celera, over a period of 9 months a 2.91 billion base pair consensus sequence of the human genome was generated by the shotgun sequencing method, which involved collating 14.8 billion base pairs of DNA sequence made up of 27 million high quality sequence reads from five individuals. Further, MLA says that although Professor Taylor appeared to suggest that the method taken by the inventors of the 253 Application was different to Venter because they took DNA from a number of individual cattle, whereas the human genome project did not, he was wrong. MLA says that both projects used multiple sources of DNA. 723 Further, MLA points to the fact that Venter states that 2.1 million single-nucleotide polymorphisms were identified, but less than 1% would result in variation in proteins. 724 Now in my view it has not been established that the information in Venter had become part of common general knowledge as at the priority date, although I accept that shotgun sequencing was well known. Venter was published before the priority date and the news that the human genome had been sequenced no doubt attracted some attention. But the relevant question is the nature of the information, if any, which had become generally accepted in the field of bovine genetics and had become relevantly common general knowledge in the context that I am considering. But even if the information in Venter had become part of common general knowledge, I do not consider that it significantly assists MLA. 725 Venter focuses on the human genome. It provides a general discussion of the identification of SNPs, but it does not describe an association between a SNP and a trait. 726 Further, the sequencing of the human genome published in Venter was done for the purpose of generating a reference assembly, and was not done to create a genome wide panel of informative SNPs. Moreover, the discussion of SNPs in Venter related to issues of human evolutionary history, rather than the potential use of SNPs in animal breeding programs. 727 Further, as compared with Venter, the genome sequencing and assembly methods used by the inventors of the 253 Application were different. Those inventors sequenced DNA from 4 diverse breeds of cattle. These were selected to be representative of relevant genetic diversity. This approach, which is different to the approach of Venter, had complicated contig generation and SNP identification. The inventors were able to identify 786,777 SNPs by comparing each of the sequenced DNA fragments from the four bovine subjects. The inventors then used the human genome as a scaffold to map the location of 242,181 SNPs. This step (also not described in Venter) presented technical challenges as explained by Professor Taylor; now MLA says that this is "pure speculation", but I consider that there is an air of unreality to that assertion. The disclosure in the 253 Application was the first time that a genome wide panel of SNPs had been positioned for a non-human species using the human genome as a scaffold. Further, the inventors then identified which of the 242,181 SNPs would be highly informative across breeds. The inventors proceeded to identify which of the 242,181 SNPs were polymorphic across a majority of 210 animals from diverse breeds. SNPs that were not sufficiently polymorphic across the 210 animals were discarded and replaced with another SNP within the same region. This process was repeated until a map of 6,189 relatively evenly spaced SNPs was obtained; see Example 1 of the specification. The experts agreed that the scale of genotyping required to perform this process would have been very difficult; again, this process was not described in Venter. Then the inventors in order to identify SNPs that were associated with commercially valuable traits collected phenotypic data relating to marbling, tenderness, fat thickness, daily gain and retail yield from 4,791 bovine subjects. Animals at the extremes of each phenotype were separated and their DNA samples pooled. Each pool of DNA was genotyped for each of the 6,189 mapped SNPs to identify those SNPs that were statistically significantly associated with a trait. This genome-wide association study (GWAS) identified 2,510 SNPs that were associated with the five traits tested. Now Venter does not at all describe a genome-wide association study. 728 So, as I said above, even if Venter was part of common general knowledge, it hardly assists MLA. 729 Further, if it needs to be said at this point, the work carried out as part of the 253 Application could hardly be characterised as routine or involved "normal clinical trials, which utilise certain standards and procedures" (AstraZeneca, at [95] per Kiefel J) or trials that "would conventionally be carried out" (AstraZeneca, at [123] per Nettle J). A careful review of the 253 Application suggests the very opposite.

SNPs as DNA markers 730 Let me deal with another topic concerning common general knowledge. 731 MLA also contends that it was also known that DNA markers, including SNPs, were useful for identifying traits in animals. It is said that at the least, the utility of SNPs (and indeed a dense array of SNPs) was described and known through Venter. In any event, MLA says that regardless of Venter, it was known that SNPs would be useful markers to conduct genome wide association studies and infer/identify traits in bovines. 732 Further, MLA says that although Branhaven seeks to make much of the fact that microsatellites were in greater use than SNPs as at December 2002, that is not the relevant question. It is said that both kinds of markers were available and being used. Further, MLA says that while both were suitable for genotyping animals, the utility of SNPs was not in doubt. MLA says that it was known that there was a far greater prevalence of SNPs in the genome than microsatellites and that such prevalence accounted for much more than the greater polymorphism inherent in microsatellites. MLA also says that it was known that SNPs were easier to genotype than microsatellites and that statistical studies using SNPs were easier to perform than for microsatellites. 733 MLA also says that the methods of identifying SNPs in the bovine genome were also well known. It submitted that such techniques had been used in relation to the human genome (i.e. shotgun sequencing) and were equally applicable to the bovine genome. For example, before the priority date and according to Professor Taylor's evidence, Genomic FX was using a computer to identify differences in sequences and therefore the presence of SNPs. 734 In response to these contentions I would observe the following. 735 As at the priority date, and indeed for a time thereafter, if genetic markers were being used in cattle breeding they were likely to be microsatellites rather than SNPs. The markers that Dr Sonstegard used during 1997 to 2003 when mapping QTL were microsatellites. His experience was that microsatellites were the most widely used up until the mid-2000s, even though other types of markers may have been available. Professor Taylor also used microsatellites in his own research and Professor Visscher continued to rely on data generated from microsatellites in his quantitative research work in cattle up to 2005. 736 The widespread use of microsatellites as at the priority date meant that they were the most common marker present on the bovine linkage map. They were therefore a valuable resource for QTL mapping experiments as they produced results that were more reproducible across laboratories. Further, microsatellites being more polymorphic were also considered by some of the experts to be more informative than SNPs. 737 Further, as at the priority date very few SNPs had been discovered in cattle. Moreover, the technology that was available for genotyping them was either expensive or low throughput. Indeed, many more bovine microsatellites had been identified as compared to SNPs. As Dr Sonstegard said, as at the priority date researchers were discovering new microsatellites and developing denser linkage maps. 738 Further, as I have already said when discussing the issue of novelty, the Meuwissen simulation was based on markers which more closely resembled microsatellites or behaved more like microsatellites. Vignal also illustrated a practical preference for microsatellites. 739 Now although MLA has sought to diminish the significance of the fact that microsatellites were in greater use than SNPs, I disagree. In assessing what steps the notional skilled person would have taken as at the priority date, the practical advantages (actual and perceived) of microsatellites over SNPs are important to consider. Indeed, Professor Hayes gave evidence that as at the priority date it was possible to implement genomic selection using microsatellites. Professor Goddard also said that although there were only about 4,000 bovine microsatellites known as at December 2002, this would have been sufficient to implement genomic selection. Now I accept that it was disputed by other experts whether there were in fact at the time a sufficient number of microsatellites across the entire genome to carry out a genomic selection approach. But it is apparent from the evidence that even assuming that the notional skilled person had decided to use a genome wide approach, they would have been likely to have used microsatellites. 740 But notwithstanding these observations, I would accept that as at the priority date, and generally speaking, it was part of common general knowledge that SNPs could be used to identify traits in animals. There were examples in the evidence where this had been done, although such examples were not themselves shown to be part of common general knowledge. For example, the Rothschild patent no 5,550,024 (1996) discloses genetic markers for pig litter size, methods for identifying such markers and methods for screening pigs based upon various polymorphisms, preferably a restriction fragment length polymorphism. One can see from the discussion of Figure 11, reference to the Pvu II polymorphism with underlined bases corresponding to polymorphisms between the two sequences generating the Pvu II recognition site. In other words, the patent identifies and uses SNPs as relevant markers.

Comparative genome mapping 741 Let me address another topic. MLA says that comparative genome mapping of species, for example the bovine genome against the human genome, was a well-known technique. It says that it was well known that such mapping allowed for spacing markers evenly along a putative genome designed with the human genome as a scaffold. But MLA says that in any event, it was also known that it was not necessary to have a genomic map in order to do genomic selection in some species, because the screening of many SNPs would inherently involve a spread of SNPs throughout the genome. 742 In my opinion, it may be accepted that comparative mapping between species was known as at the priority date. But it was typically used in relation to discrete loci to identify the specific location of a gene or the order of multiple genes.

Techniques for determining association between SNPs and traits 743 MLA also says that techniques for determining if SNPs were associated with traits were well known. It is said that as at the priority date this involved genotyping animals using known nucleic acid hybridisation techniques and then using standard statistical techniques to assess whether an association between a marker and a trait existed. As noted by Professor Taylor, the least squares approach had been used in a GWAS as early as 1996, and BLUP had been known for a considerably longer time. I accept that these techniques were well known as at the priority date.

Genotyping animals 744 Let me deal with another topic. MLA submitted that as the 253 Application explained at [0152], there were various "[m]edium to high-throughput systems" available for genotyping animals for identified SNPs. In this context, MLA said that Professor Taylor explained that as at the priority date the then high-throughput devices could genotype up to 50 SNPs at a time. In my view it is more accurate to say that Professor Taylor said that devices as at the priority date would generally genotype fewer than 50 SNPs at a time. Professor Taylor also gave evidence that in 2002, DNA samples could be sent to laboratories (e.g. Geneseek) and genotyped for fairly small numbers of SNPs. 745 MLA also says that by December 2002, people in livestock research knew that very high-throughput genotyping tools to use in DNA marker selection for cattle "were coming". MLA says that SNP chips "as a concept" were not only known as at the priority date but were being used for markers found in the human genome. 746 Further, MLA says that Meuwissen anticipated using this technology. MLA says that the two statistical approaches in Meuwissen that used all markers for genomic selection (being BLUP and Bayes A) were based on the premise that rapid genotyping of SNPs would become possible with new technology. MLA has also pointed out that Professor Visscher referred to this in the 1998 paper by Haley and Visscher (Haley CS and Visscher PM, "Strategies to Utilize Marker-Quantitative Trait Loci Associations" (1998) 81 J Dairy Sci 85). Professor Visscher has also cited a 1999 paper by Lipshutz et al (Lipshutz RJ, Fodor SPA, Gingera TR and Lockhart DJ, "High Density Synthetic Oligonucleotide Arrays" (1999) 21 Nature Genetics Supplement 20), which said it would be possible to have up to 50,000 nucleotides on a chip. I would interpolate at this point though that the first commercial bovine SNP chips (being the 10,000 SNP Chip) that enabled very high throughput genotyping of the bovine genome were not commercially available until 2005. In any event, MLA says that the inventors of the 253 Application did not invent any new method of genotyping animals, but simply used the available devices at the time. 747 Further, MLA says that once Dr Sonstegard became aware of Meuwissen, he immediately set out to make a tool to prove the method of the paper; I would note though that Dr Sonstegard did not become aware of Meuwissen until 2005. Further, Professor Taylor stated that he was so excited by the method in Meuwissen that he spent 5 or 6 years trying to build a SNP chip to test the method once he had heard about it. 748 Now as to these submissions, I accept that as at the priority date, techniques for genotyping animals were known. However, SNP genotyping could only be accomplished in small numbers as high-density SNP genotyping assays had not been developed for any species. Further, in my view the anticipation that very high throughput genotyping tools "were coming" does not necessarily qualify them as forming part of common general knowledge as at the priority date.

Sorting animals and cloning animals 749 MLA contends that it was commonly known how to sort and select animals once the presence of DNA markers for desirable traits were identified, and how to clone animals once they were selected as being desirable. I accept that these methods were part of the common general knowledge.

Meuwissen and the concept of genomic selection 750 Let me deal with the final aspect of common general knowledge concerning Meuwissen that I touched on earlier and discussed in greater detail in the novelty section. 751 MLA says that Meuwissen, the concept of genomic selection and, in particular, the four statistical approaches to genomic selection described in Meuwissen, were well known as at the priority date. Nevertheless, MLA had to accept that the molecular geneticists, Professor Plastow and Dr Sonstegard, had no knowledge before the priority date of Meuwissen. But the quantitative geneticists all knew about the paper before the priority date and considered it to be significant. It is said by MLA that this case is a good example of how the hypothetical skilled addressee might have a greater breadth of common general knowledge than any one particular type of expert. 752 Now although I have discussed Meuwissen previously when discussing the question of novelty, it is appropriate to repeat some aspects. 753 Meuwissen described how to use genetic markers, preferentially microsatellites, throughout the genome of a livestock species (cattle being one example) to predict the genetic potential of an animal for a quantitative trait. The four statistical methods described used differing numbers of markers and differing approaches to detecting an association, which led to different degrees of accuracy of prediction for the trait. Two methods used every genetic marker that was available and were the most accurate (BLUP and Bayes A). The least squares and Bayes B methods chose the most statistically significant number of markers to proceed with, with the least squares method (using the least number of markers) being the least accurate method. 754 MLA says that it was well known at the priority date that the initial genotyping and phenotyping of animals for a dense map of genetic markers as recommended by Meuwissen would be costly and time-consuming. Now the genotyping systems at the time as described by Dr Sonstegard were limited to about 50 SNPs at a time, but MLA says that such work was "simply a matter of cost and time". It is said that that is what the inventors did in relation to the 253 Application in that they genotyped a dense map of markers in populations of animals which had been phenotyped. But MLA says that they did not invent any new way of doing that work. 755 Further, MLA says that in terms of making a commercial genotyping tool that could be used to select animals as at the priority date, two of the statistical approaches in Meuwissen involved the use of a selection of the most significantly associated markers. That is, while the initial work in selecting the best markers out of the dense map of markers would have been costly and time consuming, it was possible to choose a much smaller number of markers. So, for the Bayes B method, you would discard 95% of the markers. It is said that that could provide some accuracy for prediction of the trait using existing genotyping devices as described by the specification. Further, MLA says that while SNP chips, which could handle thousands of SNPs, were not at the priority date commercially available for bovine species, they were in use in humans and were known to enable the use of all or most of the markers used for the BLUP and Bayes A methods disclosed in Meuwissen. 756 Now in my view, MLA has not established that Meuwissen had become part of the common general knowledge (although this may not matter as it can be brought in under s 7(3)). The fact that the publication Genetics may have been widely read, and the information presented and discussed at some conferences, is not sufficient to establish that Meuwissen had become common general knowledge. Professor Visscher did not know how widely accepted Meuwissen had become in December 2002. And it is not to be forgotten (as I have said) that Vignal, a reasonable summary of what was known at the relevant time, did not even cite Meuwissen. 757 Indeed, there was some evidence before me suggesting that Meuwissen (or the information therein) only became widely known after 2004 when the first genome reference assemblies and SNPs were deposited in the public domain and the first results were produced based on the analysis model disclosed in Meuwissen. 758 Further, the practical application of the approach suggested in Meuwissen referred to, inter-alia, a dense genome wide panel of markers which did not become available until the mid-2000s; but I do accept that it disclosed that a subset of markers could be used. Further, as Professor Taylor noted, it was also not clear how marker density would influence the accuracy of the approach suggested in Meuwissen or whether there was a marker density below which the approach would not work.

(c) Obviousness in light of common general knowledge 759 Having addressed in some detail what was common general knowledge, it is appropriate to now consider the question of obviousness in that light. 760 MLA has contended that before the earliest priority date, the hypothetical skilled person (or team) armed with the common general knowledge at the time, would have desired to produce a molecular approach for the selection of cattle for traits of economic significance. The team would also know that such traits were generally quantitative traits with a large number of genes each accounting for a small amount of the trait. 761 MLA contends that in addressing that desire, the team would know of the Meuwissen genomic selection approach. The team would know that the approach required at the outset a large set of genetic markers spread throughout the genome to take into account average LD in the target population. 762 MLA says that the team would be aware that LD not only changes from population to population, but even within the genome of a single animal. Accordingly, the team would pick an average LD when deciding how dense the marker map needed to be. MLA relies upon Professor Taylor's evidence that at the time, for bovine it was known that there was widespread disequilibrium within the cattle genome and that it was anticipated that LD would be in the order of 500,000 bp. Further, MLA says that in any event, it did not matter if the estimate of LD was wrong. It says that reducing the number of markers would not have resulted in a large loss in accuracy, and relies accordingly on the evidence of Professor Goddard and Professor Visscher. It was simply a matter of degree. 763 MLA says that the skilled team would also be aware of the Venter paper and so was aware that it was possible to map the entire bovine genome within at most 9 months and, by the priority date, within 26 days, by shotgun sequencing (assuming that you had the number of machines that Celera had). The team would also know from Venter that the approach would disclose a dense set of SNPs throughout the genome which would be useful for association studies. Alternatively, the same skilled team was aware that it could use a less complete shotgun sequence and map it to the human genome. 764 MLA says that the skilled team would also be aware of how to phenotype animals. It had been practised as part of the progeny selection approach since well before 2002. The skilled team would also know how to genotype animals in a test population for the dense set of SNPs spread throughout the genome for associations between the SNP markers and a trait, or indeed, numerous traits. The only barrier to these things was time and cost. 765 MLA also says that the skilled team would have been aware that it could use two of the approaches disclosed in Meuwissen immediately on existing genotyping devices (i.e. Bayes B and least squares), though such approaches would be less accurate than the approaches using all markers (i.e. BLUP and Bayes A). Alternatively, the skilled team would appreciate that SNP chips were being used in humans and, once made commercially for bovine, could be used to employ the more accurate approaches disclosed in Meuwissen. MLA says that the skilled team was thus aware of how to correlate the SNP markers with traits using any one of the statistical methods reviewed in Meuwissen, with varying degrees of accuracy. Accordingly, in addressing the desire to arrive at a new or improved method for selecting bovine as at the priority date, the skilled team would directly be led as a matter of course to try each of the approaches in Meuwissen in the expectation that it might well produce such a method. Each of the Meuwissen statistical methods would involve the use of a number of SNP markers spread throughout the genome, being associated with a quantitative trait, and used to identify the genetic potential of a bovine subject for the trait. MLA says that on the balance of probabilities, at least three of those markers would inevitably be within 500,000 nucleotides of a specified SNP and at least two of the markers would be within a gene. 766 Further, MLA says that the skilled team would know that it could use the predictive power of any of the approaches in Meuwissen to sort animals, select the best animals for breeding or clone them if so desired. 767 Further, MLA has contended that there would have been no relevant hurdles or difficulties as at the priority date. But that if there were difficulties and hurdles, they only "boiled down" to time and cost. 768 The first difficulty advanced by Branhaven is said to be identifying a sufficient number of markers. But MLA says that at the time, the shotgun sequencing method used in the human genome project as described in Venter was available and would produce a significant number of SNPs in bovine. It says that that was the method in fact used by the inventors (253 Application at [0053], [0191] and [0192]) and the 253 Application does not identify any particular technical difficulty or hurdle overcome or experienced in applying that method, nor does it describe the method other than by reference to Venter. 769 The second difficulty advanced by Branhaven is said to be the task of evenly spreading the SNPs throughout the bovine genome. But MLA says that that had already been done with the human genome. Also, it could either have been done against an existing bovine linkage map (as was in fact done in Grosse and Gelhaus & Förster (Gelhaus A and Förster B, "Cattle MHC Genes DOA and DOB: Sequence Polymorphisms and Assignments to the Class IIb Region" (2001) 28 European Journal of Immunogenetics 429) or using the human genome as a scaffold, which is what the inventors of the 253 Application did. MLA accepts that both those methods would be inaccurate to some degree, but not useless. MLA says that otherwise, the SNPs could be mapped against the bovine genome once it was created, which was just a matter of time and cost (26 days for Celera), but not technical difficulty. But, in any event, MLA says that this difficulty would not in reality exist if sufficient SNP markers were used because they would necessarily be spread throughout the genome. That is, it was not in fact necessary to map the markers at all. MLA relies upon the evidence of Professor Visscher in support of such propositions. MLA says that the above analysis "does away with" the evidence from Professor Taylor regarding the "non-trivial" exercise of collating and mapping the shotgun sequences. 770 The third difficulty advanced by Branhaven is said to be genotyping thousands of SNPs. But MLA says that the ability to genotype up to at least 50 SNPs was not difficult and so the difficulty in genotyping many more was simply a matter of time and cost. Furthermore, SNP chips (array of SNPs) was the obvious solution, which approach was being taken for markers found in the human genome. In any event, so MLA contends, the 253 Application does not itself suggest that there is any such technical difficulty. The 253 Application describes various "known in the art" medium to high-throughput systems for analysing SNPs (253 Application at [0152]). Accordingly, MLA says that even if Branhaven was right to assert that it was speculation whether a SNP chip would become available, that does not answer the inventive step case because that was not the only way of achieving the invention. 771 Generally, MLA says that if a path is known or an obvious one to take, that is not altered by the fact that it might take a long time and be expensive. MLA accepts that it would have taken a long time and cost a lot of money to apply the claimed method to thousands of SNPs until the bovine SNP chip became available. But MLA says that does not mean that it involved an inventive step to seek to apply the method without a SNP chip, or to consider using one in the future. It says that a lack of resources is simply not relevant to the inventive step question. 772 In summary, MLA says that the evidence shows that the skilled team, faced with the problem of improving the methods for predicting the genetic potential of animals and armed with sufficient resources, would have been led directly as a matter of course to something falling within each claim of the 253 Application in the expectation of producing a better method than the existing approaches. It says that a finding of obviousness must follow. It contends that there is no inventive step in identifying naturally occurring SNPs and their natural associations with traits as a result of following an obvious path. Once the obvious path was followed, the result was inevitable. 773 Further, to Branhaven's assertion that as at the priority date no one was practicing genomic selection or GWAS because there were nowhere near enough markers available to apply either and that this was a profound technical hurdle, MLA says that the 253 Application neither describes a new way to identify such markers nor claims one. It simply sets out the use of known methods (Venter) to find them and claims not only the result of their work, but any SNP falling within the region of about +/- 500,000 bp of a specified SNP. 774 Further, MLA says that to the extent that Meuwissen can be said to have made the concept of genomic selection well known in livestock breeding, at most the inventors reduced that idea to practice using known techniques. It is asserted that Dr Sonstegard agreed that "all" the inventors did was reduce an existing concept to practice and that reduction to practice involved no new techniques. I do not consider that this assertion fairly reflects the totality of Dr Sonstegard's evidence (or for that matter Professor Plastow's evidence) or the appropriate concessions they were prepared to make under cross-examination. 775 MLA also says that no inventive step is given to the method by the addition of the requirement that at least three SNPs occur in more than one gene, because this is merely an arbitrary parameter. As I have said, I have rejected that contention. 776 In summary, MLA says that for these reasons, all of the claims lacked an inventive step in light of the common general knowledge as at the priority date. Further, the additional features of claims other than claim 1 were also all well known.

Analysis 777 In my view MLA has not established that common general knowledge would have directly led the skilled person to the claimed invention. Common general knowledge did not suggest using SNPs identified by a genome wide association study to identify a trait, and a fortiori the particular method claimed in the 253 Application. 778 Now the Venter paper (assuming in favour of MLA that it was part of common general knowledge) disclosed the shotgun sequencing of the human genome, but it was not concerned with the construction of an informative genome wide dense panel of markers of SNPs to identify commercially important traits in bovine. Further, and contrary to MLA's assertion (as I understood it), the task of evenly spreading the SNPs throughout the bovine genome had not already been done with the human genome. Further, Venter did not disclose any processes for selecting and refining SNPs to ensure that the SNPs were informative for traits. 779 Further, Meuwissen (assuming in favour of MLA that it was part of common general knowledge) simulated genomic selection assuming that a dense genome wide panel of markers already existed. As Professor Goddard said: Meuwissen "assume[d] that there are these markers known and can be used, so we assume that there are a panel of genome-wide markers available. We don't describe how you would actually discover them". 780 Further, it is not apparent why a skilled person would choose the least accurate approach (least squares) in Meuwissen, particularly if, as asserted by Professor Goddard, there was a need for high levels of accuracy. 781 Further, although MLA has contended that the skilled person would appreciate that SNP chips were being used in respect of humans, and once made commercially for bovine could be used to employ the more accurate approaches disclosed in Meuwissen, as at the priority date genome wide SNP chips were not being widely used in respect of humans. 782 Further, I would also reject a central proposition of MLA's inventive step case that the construction of an informative SNP map for the entire bovine genome at the priority date, involved only routine steps. 783 Let me begin by saying that MLA did not lead evidence from a molecular geneticist on this issue and it was accepted by all experts that the construction of a SNP map for the entire bovine genome was work done exclusively by a molecular geneticist. MLA's witnesses accepted that they had no relevant laboratory experience. Professor Goddard has never performed molecular genetic techniques and did not know how to. Neither Professor Hayes nor Professor Visscher had ever worked in a laboratory save for Professor Hayes saying that he was once "marginally involved". 784 I agree with Branhaven that little weight should be accorded to the assertions of MLA's witnesses that the steps that a molecular geneticist would have had to take to construct and characterise an informative SNP map for the entire bovine genome was entirely a matter of routine steps. I accept the evidence of both Professor Plastow and Dr Sonstegard, both highly experienced molecular geneticists, that the construction of a SNP map for the entire bovine genome was anything but routine. 785 As Professor Plastow and Dr Sonstegard said, the development of a sufficient number of markers would have presented a significant technical hurdle. Moreover, the development of a dense panel of markers, which was adequately spaced throughout the genome in order to adopt the genomic selection approach, presented even greater difficulty. In my view, and contrary to MLA's assertions, it was not just a matter of "time and expense" or a mere matter of "scale". 786 As Branhaven correctly pointed out, one key objective touchstone of the practical difficulties in constructing an informative SNP map is the time that it actually took for researchers in both the fields of human and bovine genetics to produce a sufficient number of SNPs to construct a dense genome wide panel of markers. 787 Professor Visscher referred to the 1996 publication of Risch and Merikangas in relation to research into human disease (Risch N and Merikangas K, "The Future of Genetic Studies of Complex Human Diseases" (1996) 273 Science 1516) which proposed a GWAS approach in humans. It was apparent from Professor Visscher's review article in 2012 titled "5 Years of GWAS Discovery" that the first results for a GWAS study in humans were not published until 2005 and 2006 and that the starting point was actually considered to be the 2007 Wellcome Trust paper (The Wellcome Trust Case Control Consortium, "Genome-Wide Association Study of 14,000 Cases of Seven Common Diseases and 3,000 Shared Controls" (2007) 447 Nature 661). This was all about 10 years after first being suggested by Risch and Merikangas. But Professor Visscher did say in his oral evidence that he was aware that in 2002 a GWAS study was carried out on humans where a SNP chip wasn't used but a different experimental technique was undertaken. 788 And even in 2005, it was recognised with reference to studies of humans that whole-genome association studies had not generally previously been possible because dense enough marker maps and high-throughput genotyping platforms were lacking (see Uimari P, Kontkanen O, Visscher PM, Pirskanen M, Fuentes R and Salonen, JT, "Genome-Wide Linkage Disequilibrium from 100,000 SNPs in the East Finland Founder Population" (2005) 8 Twin Research and Human Genetics 185). 789 As for bovines, the evidence of Dr Sonstegard and Professor Taylor establishes that the first 10,000 SNP chip did not became available until 2005. Further, the evidence of Professor Plastow and Dr Sonstegard also establishes that the 50,000 SNP chip did not become available until 2008. 790 Now MLA has asserted that people in the field knew that something was coming, for example, high density SNP chips or high throughput genotyping tools. But in my view that anticipation is not a proper basis upon which to answer the statutory test imposed by s 7(2) of the Act. As Branhaven rightly contended, that the construction of a SNP map for the entire bovine genome was theoretically possible does not make its construction a matter of routine. Nor does the fact that there was interest in the use of molecular genetics in cattle breeding as at the priority date. 791 Further, as I touched on earlier in my reasons, there was a definitional debate amongst the experts as to the distinction between GS and GWAS. To the extent that anything turns on it, I accept the distinction identified in the Rabier article between GS and GWAS (Rabier C-E, Barre P, Asp T, Charmet G and Mangin B, "On the Accuracy of Genomic Selection" (2016) 11(6) PLoS ONE e0156086). The difference between GS and GWAS is that GS involves all the marker effects being estimated simultaneously. To adopt the language of Professor Goddard, in GS all markers are "taken into account" and "treated impartially". Similarly, Professor Visscher suggested an approach that involved "estimating marker effects for all SNPs". Contrastingly, GWAS involves establishing specific associations between individual SNPs and a trait. But it is also important to observe that this distinction was not necessarily apparent as at December 2002. But it is not in dispute that both GWAS and GS require a dense, genome wide, panel of markers. 792 Let me now proceed to make some general observations. 793 If one is applying the Cripps question, it must be shown that the notional research group would directly be led as a matter of course to try the combination of integers claimed in the expectation that it might well produce a useful result. The combination of integers claimed in the 253 Application require, inter-alia, the use of at least 3 SNPs, one which must be selected from the panel of 2,510 SNPs or SNPs within about 500 kilobases either side of these SNPs. And all of the at least 3 SNPs must be associated with a trait. Now even if it were accepted that the notional skilled person would have routinely identified and characterised SNPs across the bovine genome, MLA has not established that the skilled person would arrive at the actual SNPs of the 253 Application, let alone the use of at least 3 SNPs in the claimed method. And even if it were assumed that a sufficient number of SNPs had been identified, different researchers would make their selection from available SNPs in different ways as Professor Plastow explained. And that is irrespective of whether the claims are construed to require that the limb (b) SNPs be in LD with the limb (a) SNPs. 794 Further, although MLA has asserted that each of the various steps taken by the inventors of the 253 Application were individually known and therefore the invention as claimed was obvious, I agree with Branhaven that such an approach invites error. The invention is directed to a method for inferring a bovine trait using a number of identified SNPs all of which must be associated with the trait. Even if it were accepted that each of the individual methods used by the inventors were known techniques, that would not provide a proper basis for finding that the claimed invention was obvious. 795 Further, although MLA also relies on the fact that the 253 Application does not identify particular technical difficulties in the steps undertaken by the inventors, there is no requirement for any such disclosure. 796 Further, let me deal with another point raised by MLA concerning claim 1 and inventive step. MLA says that where one is dealing with a combination of elements, it is their interaction which is the essential requirement of invention (Aktiebolaget Hässle at [6] and [41]). So much may be accepted. But MLA seeks to draw too much from the notion of interaction in suggesting that the integers of claim 1 do not interact. One is dealing with a method claim which has "component" parts. But the parts do not need to interact like a physical locking mechanism in order for the component parts to interact in a relevant sense to form a new invention. Each component is necessary to, and the combination is sufficient for, the invention as claimed. In the relevant sense they interact.

Meuwissen as a section 7(3) document 797 As I have said above, I have not accepted MLA's contention that the information in Meuwissen was common general knowledge. But MLA also says that it can at least be relied upon under section 7(3) of the Act. Now as I have previously indicated, as the 253 Application was filed after 1 April 2002 and examination requested before 15 April 2013, the amendments to section 7(3) of the Act brought about by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (Cth) do not apply to the 253 Application. 798 Under the applicable version of s 7(3), for a document to be relied on under that provision there must be some basis for concluding that the skilled person could be reasonably expected to have ascertained, understood and regarded as relevant the particular prior art in question. The opponent may need to adduce specific evidence to support a contention that a particular prior art document satisfies the requirements of s 7(3), although in some cases this may be readily apparent without direct evidence. 799 But even if information is able to be taken into account under s 7(3) of the Act, that is not the end of the matter. The opponent must still establish on the evidence, that the invention would have been obvious in the light of that information taken together with the common general knowledge. 800 In AstraZeneca, Kiefel J summarised the effect of s 7(3) as follows (at [68] to [70]): Before a document containing prior art information can be used along with the common general knowledge for the purposes of the s 7(2) inquiry, it is necessary that it meet the requirements of s 7(3). In Lockwood Security Products Pty Ltd v Doric Products Pty Ltd (No 2) it was explained that prior art information which is publicly available in a single document is "ascertained" if it is discovered or found out, and "understood" means that, having discovered the information, the skilled person would have comprehended it or appreciated its meaning or import. The Court also explained that the phrase "relevant to work in the relevant art" is directed to publicly available information, not part of the common general knowledge, which the skilled person could be expected to have regarded as relevant to solving a particular problem, or meeting a long-felt want or need, as the patentee claims to have done. Lockwood (No 2) also explains [at [127]] that, in answering the question of obviousness, the information referred to in s 7(3), like that part of the prior art base which is the common general knowledge, is considered for a particular purpose. That purpose is to look forward from the prior art base to see what the skilled person is likely to have done when faced with a problem similar to that which the patentee claims to have solved with the claimed invention. It is this aspect of the s 7(2) inquiry which assumes particular importance on these appeals. In addressing s 7(2), it is to be borne in mind that the skilled person is an artificial construct, intended as an aid to the courts in addressing the hypothetical question of whether a person, with the same knowledge in the field and aware of the problem to which the patent was directed, would be led directly to the claimed invention. The statute's creation of the skilled person construct for this purpose is not to be taken as an invitation to deal with the question posed by s 7(2) entirely in the abstract. Whilst the question remains one for the courts to determine, the courts do so by reference to the available evidence including that of persons who might be representative of the skilled person. (citations omitted) 801 French CJ made clear at [43] and [44] that data is not required in order for a prior publication to render an invention obvious, provided that experiments to create such data would be routinely carried out: The question posed by this Court in Wellcome Foundation Ltd and AB Hässle does not require that, in order to sustain an obviousness case, a party has to lead evidence which echoes the terms of that question. A similar conclusion was open, as the primary judge found, on Patent 471. AstraZeneca submitted in this Court that the "claimed invention" is a treatment using a once daily, 5-10 mg dosage of rosuvastatin. The only dosage expert, Dr Reece, confirmed that neither the Watanabe Article nor Patent 471 contained animal or human trial safety data. He had given evidence that such data were essential to determining what dosage should be tested in clinical trials. The person skilled in the art would never have chosen the dose to be tested simply by trying the doses that worked for other statins. That evidence was referred to by Jessup J, who said: But that evidence also made it quite clear that such trials would conventionally be carried out. They would fall within the concept of working towards the invention with an expectation of success referred to in AB Hässle. No error is disclosed in that reasoning. (citations omitted) 802 MLA says that that is the case here. It says that the evidence shows that a person skilled in the relevant art, faced with the task of improving existing methods to predict the genetic potential of cattle, even if not aware of Meuwissen before December 2002, would have found it in a search, understood it and regarded it as relevant, and would have been directly led by Meuwissen to practically apply each of the statistical approaches described therein as a matter of course, subject only to available resources. MLA says that the person skilled in the relevant art would have expected to improve existing methods by doing so. 803 MLA says that to the extent that Meuwissen does not defeat any of the claims of the 253 Application on the ground of lack of novelty, then it must defeat all of the claims on the ground of lack of inventive step (whether it is part of common general knowledge or relied upon under s 7(3)). 804 Now I accept that Meuwissen could reasonably be expected to have been ascertained by the skilled person at the priority date and accordingly can be relied upon under s 7(3). But the problem for MLA is that Meuwissen was a hypothetical simulation and accordingly did not involve any real data. Further the aim of Meuwissen was (at 1820, left hand column): to compare least squares, BLUP and Bayesian analyses for their accuracy of predicting total breeding value of individuals in a situation where a limited number of recorded individuals are genotyped for many markers with many alleles per marker. 805 Now MLA has relied on the proposition from AstraZeneca (at [43] and [44] per French CJ) that data is not necessarily required to be provided in a prior art document to render a claimed invention obvious. But that observation was subject to the qualification that the experiments required to create the data would conventionally be carried out. But contrastingly, the experiments required to produce the necessary data based on the Meuwissen simulation could not sensibly be said to be merely routine work. 806 Further, Meuwissen taught away from the claimed invention. The least squares method discussed in Meuwissen performed worse than the three other statistical methods discussed. It greatly overestimated some haplotype effects and underestimated others. Professor Hayes accepted that the least squares statistical method discussed in Meuwissen was the worst performing method of the four discussed, although he said that it did "work". 807 In my view it is likely that the skilled person would have been led by Meuwissen to use the BLUP method or Bayes A method which took into account the potential effect of all markers. He would not have been led to use the worst performing statistical method which only considered a sub-set of informative markers. So in my view he would have been led away from an approach which was closest to the claimed method of the 253 Application. 808 Further, it was accepted by both Professor Goddard and Professor Hayes that the simulation in Meuwissen was based on markers akin to microsatellites rather than SNPs. So even if it were accepted that Meuwissen would lead the skilled person to adopt a genome wide approach, he would have likely done so using microsatellites; this is fortified by the evidence that there were sufficient microsatellites available at the priority date to perform a genome wide analysis in cattle. 809 In summary, in my opinion there is nothing in Meuwissen that would lead to the use of at least 3 SNPs each of which is associated with a trait, as required by the method of claim 1 (and other claims) of the 253 Application.

Moody or Grosse as section 7(3) documents 810 As I have already explained, MLA's primary contention is that either of Moody or Grosse anticipate certain claims of the 253 Application. But MLA says that if, contrary to that contention, none of those documents anticipates at least claim 1 of the 253 Application, then the claims are bad for lack of inventive step in light of common general knowledge in combination with any one of those documents. 811 MLA points to Professor Visscher's explanation that in order to assemble a genome-wide panel of SNPs in bovines, he would have started by looking at what SNPs other researchers had already identified for inclusion in his panel of SNPs. Professor Visscher said that he would have ascertained, understood and regarded as relevant each of Moody and Grosse as disclosing SNPs that he could use in his SNP panel. Further, MLA points to the work by Professor Hayes that is said to confirm that a majority of those SNPs fall within the broad +/- 500,000 nucleotide ranges in the claim and all fall within "genes". 812 So, MLA contends that I can be confident that the person skilled in the art would have hit upon a method falling within the claims of the 253 Application, as a matter of routine, in light of common general knowledge and either of Moody or Grosse, considered separately. 813 Now in the light of Professor Visscher's evidence, I am prepared to accept that Grosse or Moody can be considered under s 7(3). But I am not sure where any of this takes MLA. 814 MLA's witnesses were not asked to consider whether the invention as claimed would have been obvious in light of Grosse or Moody when combined with the common general knowledge. The issue was also not addressed by the experts in the conclave questions and answers relating to Grosse and Moody, which were directed to the novelty of claim 1. Further, as I have said, Professor Goddard accepted that Moody disclosed no associations between any SNP and any trait. I reject MLA's assertions that either Grosse or Moody made out together with common general knowledge a lack of inventive step. Now MLA asserted that based upon the evidence of Professor Hayes, Moody or Grosse on his analysis gave you limb (b) SNPs. I must say that I was not sure whether all of this had been looked at carefully enough by the experts. But in any event I consider it a stretch to conclude from his findings and common general knowledge that this established a lack of inventive step.

The Gengler Patent as a section 7(3) document 815 In some respects it is unclear whether MLA is raising the Gengler Patent as a s 7(3) document. If it is it has not established the point. MLA has adduced no evidence that the skilled person would have ascertained the Gengler Patent. Professor Visscher gave evidence that as at 2002-2003 it was relatively rare for breeding companies to obtain patents in the field of animal breeding and genetics that directly affected his research. Professor Plastow agreed with this and did not consider patent literature to be a relevant source of information as to the state of knowledge in his field. Moreover, none of MLA's witnesses gave evidence that they would have ascertained the Gengler Patent. 816 In any event, none of MLA's witnesses were asked to consider whether the invention as claimed would have been obvious in light of the Gengler Patent when combined with the common general knowledge. The issue was also not addressed by the experts in the conclave questions and answers relating to the Gengler Patent, which was directed to novelty questions. 817 In my view, MLA has adduced no evidence to support its contention that the invention as claimed would have been obvious in light of the Gengler Patent when combined with the common general knowledge. And on my reading of the Gengler Patent, that would not be established in any event.

(e) Conclusion on inventive step 818 I reject MLA's obviousness case. The evidence does not establish, and certainly not clearly, that the invention as claimed does not involve any inventive step. 819 Further, although the above discussion focuses on the subject matter of claim 1, I accept that each claim of the 253 Application is to be considered independently for the purposes of this ground (as with other grounds). But MLA has not clearly satisfied me that any of the remaining claims lack an inventive step based upon the test(s) discussed in Aktiebolaget Hässle at [50] to [53].

Lack of utility 820 Section 18(1)(c) of the Act requires that an invention be "useful", with the onus of showing otherwise resting on MLA. Moreover, in the context of the present appeal, for MLA to succeed on this ground I must be clearly satisfied of a lack of utility. 821 Three questions may be posed. What has been promised (as gleaned from the specification) for the invention as delineated by the relevant claim? Is the promise useful? Has the promise been met? In this context the patent application (including the claims) must be construed from the perspective of a skilled person, and not in a way that such a person would appreciate would lead to unworkability when by construction it could be given a more limited meaning. 822 There is no need for the patent specification to provide support to demonstrate that a promise of the claims was satisfied (SNF (Australia) Pty Ltd v Ciba Speciality Chemicals Water Treatments Limited (2011) 92 IPR 46; [2011] FCA 452 at [296] per Kenny J). And as Jagot J accepted in Apotex Pty Ltd v AstraZeneca AB (No 4) (2013) 100 IPR 285; [2013] FCA 162 at [352], "[u]ltimately, an asserted lack of utility must be established by appropriate evidence, not by mere speculation that the invention will not work or meet the promise set out in the specification." 823 Now a lack of commercial viability or commercial practicality does not establish lack of utility. This point was made (in part) in Lane Fox v Kensington and Knightsbridge Electric Lighting Co Ltd [1892] 3 Ch 424 at 431 per Lindley LJ, cited by the High Court in Advanced Building Systems Pty Ltd v Ramset Fasteners (Aust) Pty Ltd (1998) 194 CLR 171 at [24], with that Court stating: It is no objection to the validity of a patent granted under the Act that it is commercially impracticable; its utility depends on whether, by following the teaching of the complete specification, the result claimed is produced. 824 Further, when construing the specification, it is important to consider the nature of any promises. Unless the specification would be understood to convey a clear assertion that the invention will achieve a particular outcome across the breadth of the claims, inutility will not be demonstrated by showing that in some cases that outcome may not be achieved. A claim may have utility even though the promised advantage is not achieved in all cases. In Rescare Ltd v Anaesthetic Supplies Pty Ltd (1992) 111 ALR 205; [1992] FCA 811, the claimed method of treatment utilised a nose-piece apparatus applied to patients suffering from snoring sickness so as to maintain air pressure sufficient to keep the nasal passages open. It was argued that the claim was not useful as not all patients would achieve a beneficial result. Gummow J said (at 232): One looks at the claim to see whether there is a failure to fulfil that promise. It is not necessary to show utility that the promise be fulfilled in every case. On the evidence, the claimed invention plainly is of considerable practical utility in the treatment of substantial numbers of persons who are "patients" within the meaning of claim 1. 825 See also Sanofi-Aventis Australia Pty Ltd v Apotex Pty Ltd (No 3) (2011) 196 FCR 1 at [245] per Jagot J. 826 Moreover, there are limits on any broad principle, if there be one, that everything within the scope of a claim must be useful (cf H Lundbeck A/S v Alphapharm Pty Ltd (2009) 177 FCR 151 at [217] per Bennett J). Lehane J in Aktiebolaget Hässle v Alphapharm Pty Ltd (1999) 44 IPR 593; [1999] FCA 628 at [227] observed in respect of such a broad principle that "[a] degree of caution, however, is required". If qualifications and expedients necessary to make the invention work are left to the reader to supply, that does not necessarily equate with inutility.

(a) MLA's contentions 827 MLA says that if a claim, properly construed, includes within its scope means that will not produce the promised result, even if a skilled addressee would recognise which means to avoid, then the claim will lack usefulness. 828 Now MLA says that the promise of the invention is the provision of a very accurate method of predicting genetic potential of animals. MLA places heavy emphasis on that part of the specification that says at [0101]: "This invention identifies animals that have superior traits, predicted very accurately, that can be used to identify parents of the next generation through selection." 829 MLA says that Branhaven has attempted to "run away" from the clear words of the specification at [0101] by asserting that, to the extent there is a promise, it is merely "to allow the identification and selection of cattle with superior genetic potential for desirable characteristics". MLA says that even if there was such a broad promise, it does not take away from the clear promise at [0101]. 830 Now on the assumption that the claims do not require a statistically significant association, a proposition that is now contestable as I will require this by way of amendment as I have already said, MLA says the following. 831 MLA says that as the claims only require SNPs to be "associated" with traits, with no degree of association specified, the claims will include at least three SNPs that will not, collectively, accurately predict genetic potential in relation to such traits. 832 MLA says that the claims include specified SNPs that are asserted to have an association with a trait in the discovery bovine population only. But as Professor Hayes has pointed out, none of those has been validated as associated in any other bovine population. Moreover, the claim includes non-specified SNPs for which no association has been shown, but is only assumed based on LD. Accordingly, so MLA says, it is not possible to say with respect to any specified SNP or non-specified SNP that it will accurately predict genetic potential in relation to any trait. Nor is it possible to say that a combination of at least three SNPs wherein at least one SNP is a specified SNP or a non- specified SNP will accurately predict genetic potential in relation to any trait. 833 MLA says that the experts agreed that there will be a number of false positive results, which means that there will be SNPs within the claim that will not predict association with a trait. I agree that the evidence establishes that there are likely to be false positives. But MLA had to accept that the experts disagreed as to the number of false positive results. But MLA pointed to the evidence of Professor Goddard that it is likely to be higher as a result of the way in which Example 2 was conducted. MLA says that it is also likely to be very significant as no degree of association between a SNP and a trait is required by the claims. 834 Further, MLA says that even the addition of an integer requiring that the non-specified SNPs be in LD with a specified SNP could not save the claims from a finding of lack of utility, as one SNP being in LD with another SNP does not necessarily indicate that both SNPs are associated with the same trait. Professor Goddard gave evidence that you can only be certain of this if both SNPs are in 100% LD. But in my view, if LD is injected into limb (b) and each of the 3 SNPs has to have a statistically significant correlation with the trait, this may not be as much of an issue. 835 Now on the alternative assumption that the claims require a statistically significant association, MLA says the following. 836 MLA has contended that even if the claims require that there must be a statistical significance in the association between at least three SNPs and the trait in question with a p value of equal to or better than 0.01, that does not resolve the issue. 837 First, it is said that a mere statistical significance in a test population is not useful until it is validated in the working population. MLA says that even if the word "associated" in claim 1 required statistical significance, it does not require validation. Accordingly, it says that given the number of false positives that would exist in the specified SNPs (Professor Hayes indicated there would be about 305) it is possible to choose at least three SNPs which will not work (and more SNPs as required in the later claims). 838 Second, MLA says that even if a claim required not only a statistically significant association but also a validated association, that still does not assist Branhaven because the degree of statistical significance it attributes is too low to very accurately predict multigene quantitative traits from just three SNPs. Professor Goddard gave an example of a study of human height. The study found that in 250,000 people, 700 markers were identified that were statistically significant. All of those markers explained 16% of the variation in height. Therefore three of those markers chosen at random would explain only about 1% of the genetic variation. Professor Goddard said that this was typical of virtually every trait that had been looked at. Further, Professor Hayes stated that traits in cattle are similar to the height trait in humans. Accordingly, MLA submits that 3 or even 10 of the specified 2510 SNPs are not enough to make an accurate prediction. 839 I must say that this particular issue, as well as various aspects of construction that I have discussed earlier that go to clarity and a description of the invention, has troubled me the most.

(b) What is the promise? 840 It is convenient to again set out [0022] and [0101] of the specification which are potentially relevant to identifying the promise: [0022] In summary, three different experimental approaches have been used with limited success to identify genes, chromosomal regions or DNA markers that account for a large proportion of the genetic variation observed in economically important traits in livestock species. The results achieved from research programs utilizing these methods have not been widely utilized to date because they do not account for enough of the total genetic variation to allow accurate prediction of an animal's performance for a specific trait. Furthermore, even when successful these approaches are only capable of identifying additive genetic components while ignoring non-additive genetic components such as dominance (i.e. dominating trait of an allele of one gene over an allele of another gene) and epistasis (i.e. interaction between genes at different loci) which are critical to the development of diagnostics that can be utilized by animal breeders to accurately predict genetic potential for economically important traits in livestock species. … [0101] This invention identifies animals that have superior traits, predicted very accurately, that can be used to identify parents of the next generation through selection. These methods can be imposed at the nucleus or elite breeding level where the improved traits would, through time, flow to the entire population of animals, or could be implemented at the multiplier or foundation parent level to sort parents into most genetically desirable. This invention provides a method for determining the optimum male and female parent to maximize the genetic components of dominance and epistasis thus maximizing heterosis and hybrid vigor in the market animals.

(c) Some of the expert evidence 841 It is also appropriate to record some of the expert evidence relevant to this aspect. 842 Professor Goddard did not believe that the inventors of the 253 Application accounted for possible statistical errors when they identified the 2,510 significantly associated SNPs. Consequently, the 2,510 SNPs would have included a number of false positive results. Referring to the process at [0195] to test each of 6,189 mapped SNPs for each trait, Professor Goddard explained that at least 60 (i.e. 1%) of the associated SNPs were likely to be false positives. As this process was carried out for each of five traits, it was likely that 300 of the 2,510 SNPs were false positives and these could not be identified or distinguished. Further, it was his opinion that, based on his experience, the statistical test performed by the inventors would have yielded a false positive result more often than the assumed rate of 0.01. Professor Goddard considered that there were other confounding factors that the experiments in the 253 Application did not account for, including differences between bovine breeds. These factors would have likely resulted in additional false positive results. In light of these factors, a p value of 0.01 was inappropriate and a more stringent statistic should have been applied. Further, Professor Goddard considered that the significance test applied to determine association undermined the utility of the 2,510 SNPs. It failed to assess whether one of the two alleles was consistently associated with a high value of a trait. This affected the results in Tables 1A and 1B, as it was not possible to know in which sort of cattle and under what circumstances one allele was better than another. 843 Further, Professor Goddard also noted that Tables 1A and 1B failed to report the effect sizes of the 2,510 SNPs, making it inappropriate to combine genotypes from multiple SNPs to predict the genetic value of an animal for a trait. 844 Further, Professor Goddard also considered that the 253 Application did not select enough SNPs to provide dense coverage of the bovine genome. In the absence of such dense coverage, one would not be able to obtain meaningful results that could be applied to bovine breeds not included in the Example 2 experiment. He considered that one would require greater than 300,000 SNPs in order to find SNPs that had consistent effects across breeds. 845 Further, Professor Goddard did not believe that a small number of SNPs would be useful for prediction of a trait, because each SNP may only predict a very small amount of genetic variance. For example, and as MLA submitted, Professor Goddard said that in relation to human height, 700 statistically significant markers explained only 16% of variation. Three markers would only explain 1%, a result that would be typical of "virtually every trait". Professor Goddard also calculated that the predictive power of the 2,510 SNPs was 4%, and that assuming 500 SNPs are associated with each trait the predictive power of three SNPs was 0.024%, which was so low as to not be practically useful. And although Professor Goddard was aware of the study that identified a causative polymorphism of the DGAT1 gene that accounted for a large degree of genetic variance for fat content in milk, he considered such a finding to be very rare, especially for complex traits that rely upon a large number of genes of small effect. 846 Further, Professor Goddard considered that the 253 Application provided no guidance as to how SNPs other than the 2,510 would be useful for predicting a trait's likelihood. He stated that the invention might have worked only if it was confined to the use of the 2,510 SNPs in respect of the five traits and in respect of a cattle population identical or very similar in breed composition to the population used in the 253 Application. 847 Professor Goddard considered that the specified SNPs would be little better than random SNPs for inferring bovine traits for the five disclosed traits. In respect of additional traits, he stated that the 2,510 SNPs were no better than random SNPs. In respect of the non-specified SNPs in limb (b), Professor Goddard considered that they would be no better than random SNPs for inferring any bovine trait. He stated that most SNPS within 500,000 bp of a specified SNP would be in low LD such that it would have little value as a surrogate for that SNP. 848 Further, Professor Goddard also considered it necessary that an association between a SNP and a trait be subjected to a validation study before concluding that it was useful for predicting genetic potential. 849 Professor Hayes stated that at a p value of 0.01, 305 results in the 253 Application would have been false positives. He said that this was simply a matter of statistics. He also recalled that in 2002, there were known problems with DNA pooling methods that could generate spurious results if the pools were not constructed well. 850 Further, Professor Hayes undertook work at the instruction of MLA's solicitors. He recorded that he identified the location of position 300 for 2,447 of the 2,510 sequences. He could not identify unique locations in the bovine genome for the remaining 63 sequences. Of the 2,447 aligned sequences, Professor Hayes noted that 100 of them contained flanking sequences of between one and 20 nucleotides, i.e. the polymorphism did not occur at position 300. He said that this typically indicated that the polymorphism may have been an indel rather than a SNP. Of these 100 sequences: (a) 57 had a polymorphism at either position 301 or 302. Professor Hayes said that an artefact may have been introduced by the way the relevant algorithm aligned the sequences, and that when the sequences were realigned, 54 of the 57 sequences aligned within one nucleotide of each other. (b) For the 46 sequences where flanking sequences were more than one nucleotide apart, Professor Hayes considered that there was a high probability that these sequences had an indel at the relevant position. This was supported by his observation that many of these sequences contained multiple polymorphic nucleotides within 6 nucleotides of position 300, indicating a region with multiple insertions. 851 Further, Professor Hayes also took issue with the absence of a validation study in the 253 Application. He considered that SNPs would only be useful in a selection method once they were shown to be associated in both a discovery population and a further validation population. In the absence of validation, one could not assess which of the 2,510 SNPs would be useful in selecting an animal for a trait. 852 Further, regarding the utility of the limb (b) non-specified SNPs, Professor Hayes noted that the +/- 500,000 nucleotides encompassed 60.3% of the bovine genome. Given this wide coverage, he thought it was likely that some of these SNPs would be useful in inferring a bovine trait. 853 Professor Visscher stated that due to the way that the experiments were conducted in the 253 Application, the number of false positive results could be higher than 300. Further, he considered that the utility of the invention was undermined by the absence of a validation study, which would have sorted truly associated SNPs from false positives. 854 Professor Visscher raised several concerns in relation to utility. Despite the objective of the 253 Application being a method to make more accurate and earlier predictions, there was no evidence that the method using the 2,510 SNPs was better than or could make more accurate predictions than existing methods. 855 Further, Professor Visscher said that there was no disclosure of association between the 2,510 SNPs and any traits other than the five disclosed. Consequently, he was not confident that the SNPs could make useful predictions for a trait other than these five. Further, he said that any uncertainty about the predictive accuracy of the 2,510 SNPs for the five disclosed traits would only be amplified in relation to other traits. 856 Further, the 253 Application did not quantify the usefulness of any predictive accuracy of the method, such that Professor Visscher considered that the usefulness of the invention for the five disclosed traits could be very low or non-existent. Professor Visscher believed that three SNPs could not be a useful predictor if associated SNPs only accounted for a small amount of genetic variance. 857 Further, regarding the utility of limb (b) non-specified SNPs, Professor Visscher noted that the +/- 500,000 nucleotides covered most of the bovine genome. Given this wide coverage, he thought that the scope would capture most of the associations with trait loci. 858 Professor Plastow explained that as at December 2002, using the method of claim 1 he would have employed as many of the markers from the 2,510 SNPs for as many traits as he could, because at the time it was a "remarkable" jump forward. He would have done this even without knowing the size of the genetic effect of each of the markers. 859 Further, Professor Plastow also said that using the 2,510 SNPs he would have expected a "good chance of success". 860 Professor Plastow considered that the specified 2,510 SNPs would be useful in identifying associations with bovine traits. In contrast to Professor Goddard, Professor Plastow thought that the SNPs disclosed in the 253 Application would be sufficiently robust to produce meaningful information from which trait associations could be inferred in other populations. In relation to limb (b) non-specified SNPs, he said that because the non-specified SNPs were to be in linkage disequilibrium with specified SNPs (his assumption concerning limb (b)), then they too would be useful in identifying associations with bovine traits. In relation to the necessity of conducting a validation study, Professor Plastow said he would expect that breeders would be able to calibrate the SNPs disclosed in the 253 Application against their own population of animals. 861 I would also note that Professor Plastow took issue with many of Professor Goddard's propositions. For example, he said: I do not share Professor Goddard's belief that the 2,510 SNPs disclosed in the '253 Application would only work to identify cattle of superior genetic potential in respect of a population of cattle identical or very similar in breed composition to the ones used in the experiments described in the '253 Application. In my view, the series of SNPs disclosed in the '253 Application would be sufficiently robust to produce meaningful and useful information from which an inference of trait association may be made in other populations. As I have also discussed above, I understand that what the method of the '253 Application seeks to achieve is to provide SNPs close to causative mutations. In this way, the SNPs disclosed in the '253 Application provide a useful tool to, say, commercial breeders, who want to improve the genetic characteristics of their animals. I consider it would be reasonable to expect that breeders would be able to calibrate the SNPs disclosed in the '253 Application against their own population of animals and to select those with superior genetic potential. In my view, the process would be iterative, but that does not mean that the method disclosed in the '253 Application would not work as Professor Goddard states. I disagree with Professor Goddard that the P-value selected was inappropriate. In my experience, a P-value of 0.01 is commonly used and considered perfectly acceptable as the benchmark by which the results obtained may be sorted. I do not consider that the occurrence of false positives, certainly at the low level asserted by Professor Goddard, would significantly undermine the usefulness of the 2,510 SNPs as suggested by Professor Goddard. I also disagree with Professor Goddard that it would take "a large amount of work" to identify which of the SNPs are false positives and which are not. As I have already indicated above, any "large amount of work" would have been carried out at the stage of identifying the disclosed SNPs and in determining that they can be used to infer a trait. Once provided with the disclosed SNPs, I do not consider that it would have been difficult or in any way burdensome to use those SNPs in accordance with the method claimed in the '253 Application. Professor Goddard suggests that in order for "a useful estimate" based on genetic markers to be determined, it is necessary that the number of SNPs and the size of the population be "very large" - specifically, that it is "usual" to use >10,000 cattle and at least 300,000 markers. I completely disagree. As at December 2002, the number of SNP markers that had been identified was very limited, and indeed the 50K lllumina SNP chip did not become commercially available until 2008. Moreover, in my experience, even by today's standards, a study involving 10,000 animals would be viewed as exceptionally large and certainly not usual. In my view, Professor Goddard has set a benchmark which even with the identified SNPs and SNP chip technology available today, the experimental design proposed by Professor Goddard would be extremely difficult to achieve. Professor Goddard's views in relation to the stringency that ought to be applied in considering the method claimed by the '253 Application are also, in my opinion, contrary to his view (set out above in the First Goddard Affidavit) that it is possible to have as little as 20 markers to explain 50% of variation in traits. If this were the case, I cannot understand why Professor Goddard considers it necessary to use at least 300,000 markers in order to produce a useful or meaningful result. Professor Goddard asserts that he is aware that a density of one SNP every 0.5 cM is not enough when studying breeds. In my view, this statement could not have been based on information known as at December 2002. I cannot see how Professor Goddard can state with such certainty, based on the information that was available in 2002, that the '253 Application does not disclose the selection of enough SNP markers to provide enough density of coverage of the bovine genome. As I have noted above, even the Meuwissen 2001 paper does not suggest that the density needs to be as great as Professor Goddard states is necessary. For the reasons I have already explained above, I do not agree that the process of validating the SNPs disclosed in the '253 Application must involve the matters discussed by Professor Goddard. In my view, the process can be iterative and can involve sorting through SNPs that do not apply to your population (what Professor Goddard considers to be a false positive) from those that do. In my view, that process would provide robust results that enable the selection of animals based on their genetic potential. Professor Goddard suggests that the LD of 500,000 nucleotides has resulted from "selective analysis". I do not agree. The '253 Application clearly indicates that tests were carried out in order to determine the extent of LD, and the LD of 500,000 nucleotides is fully supported by the results reported. I cannot see any reason why the LD analysis would not be justified. 862 Dr Sonstegard acknowledged that statistically there would have been false positive results in the method of the 253 Application. He also acknowledged that a validation study was necessary in order to reach a definitive view of whether a set of markers was capable of accurate prediction. 863 Dr Sonstegard stated that the specified SNPs would be useful for inferring one of the five bovine traits identified. He also said that they would have some predictive value in respect of other traits based on the extent of LD in cattle genomes and the distribution of the specified SNPs across more than half the genome. 864 Dr Sonstegard said that non-specified SNPs would be useful for inferring one of the five bovine traits if you knew what they were. They would also be useful for inferring other bovine traits but that this would require testing. 865 Commenting on the 2009 study by Hayes and Sonstegard, Dr Sonstegard said that the conclusion of a validation study that 56 SNPs of the 125 initially thought to be associated were actually associated was "pretty good". Further, he said that it was "pretty good too" when 20 of those SNPs were validated in a different bovine breed. 866 Professor Taylor considered that the 2,510 specified SNPs would be useful for inferring the five specified bovine traits. Further, he expected to find on average that 204 of those SNPs were associated with other bovine traits that possessed similar heritabilities to the five specific traits. 867 Professor Taylor considered that limb (b) SNPs would also be useful for inferring the five specific bovine traits. But he also said that not all SNPs within 500,000 nucleotides of the specified SNPs will be in LD with the QTL and that those SNPs not in LD with the QTL would not be predictive. This qualification also applied to inferring additional traits using the limb (b) SNPs. 868 Professor Taylor acknowledged the necessity of a validation study to identify issues around false positive results and a tendency in GWAS to overestimate the effect of SNPs.

(d) Analysis 869 Now there are some themes that can be distilled from the expert evidence, particularly MLA's witnesses. 870 First, it seems to be said that many more SNPs than the number stipulated in the claims of the 253 Application would be necessary in order to establish the requisite association with a trait, and therefore for the invention to be useful. 871 Second, it seems to be said that absent a validation process, the invention would not be useful. 872 Third, it seems to be said that on any view some high statistically significant association between the 3 SNPs and the trait is necessary, and that the invention would not be useful absent this. Moreover, it seems to be said that there would be difficulties in identifying a limb (b) SNP, although in my view that would have been well within the skill of the calling. 873 Fourth, it seems to be said that even if a high statistically significant association was required by the claims, that in any event there would be a substantial number of false positives. 874 Fifth, it seems to be said that on any view the invention would only have utility for the 5 traits referred to in the specification. 875 Now many of these propositions were challenged by Branhaven's experts. Given the conflict in the expert evidence and for other reasons, I am not clearly satisfied that the claims of the 253 Application lack utility in the dimensions asserted by MLA providing that amendments are made to deal with statistical significance and LD as I have indicated. Let me elaborate. 876 First, I agree with Branhaven that there is no promise in the specification that a definite degree of advantage will result from the use of the invention. MLA has referred to various passages in the specification, but these do not promise that any particular level of "accuracy" will be achieved by the use of the claimed method. Further, the specification must be construed as a whole and in context. Relevantly, [0101] cannot be taken out of context. 877 Second, if the amendments that I have suggested be made concerning "associated" and LD are made, then I agree with Branhaven that the claims of the 253 Application include (or will include), as limiting and characterising features, requirements that ensure the utility of the invention. 878 Third, matters such as the need to validate SNPs in the population of interest are qualifications and expedients of the kind readily left to the skilled reader to supply. 879 Fourth, I do not agree with MLA that many more SNPs than those specified in the claims would be necessary to make the invention work. Indeed, the specification and other evidence before me negates such a proposition. One SNP may be sufficient. 880 Fifth, as to traits other than the 5 stipulated traits, MLA has not adduced cogent evidence to support the proposition that the invention would not work for the other traits. 881 Sixth, if statistical significance is required (as I have said), the incidence of false positives does not prove a lack of utility. Any association established by statistical correlation will always have such an incidence. It is inherent in the very invention claimed. 882 In summary, in my view, MLA has not clearly established by evidence that the method would not achieve a useful result in any particular case. MLA has the onus which in my view was not discharged by its experts throwing up theoretical possibilities.

Lack of sufficiency 883 At the relevant time, s 40(2)(a) of the Act provided: (2) A complete specification must: (a) describe the invention fully, including the best method known to the applicant of performing the invention; and … 884 The "invention" for the purposes of s 40(2)(a) is the embodiment(s) which is described and around which the claims are drawn. Kimberly-Clark Australia Pty Ltd v Arico Trading International Pty Ltd (2001) 207 CLR 1 at [19] and [21] approved that formulation and its application to s 40(2)(a). 885 Only one embodiment within each claim need be enabled for sufficiency purposes. An invention is sufficiently disclosed if a skilled person could make a single embodiment of the invention which falls within the scope of the claims. The following test for sufficiency was formulated in Kimberly-Clark at [25]: The question is, will the disclosure enable the addressee of the specification to produce something within each claim without new inventions or additions or prolonged study of matters presenting initial difficulty? (footnote omitted) 886 In Lockwood Security Products Pty Ltd v Doric Products Pty Ltd (No 1) (2004) 217 CLR 274 (Lockwood (No 1)), the High Court said at [60]: For the purposes of s 40(2)(a), it is not necessary for the inventor to disclose all the alternative means; it is enough that there is disclosure in the sense of enabling the addressee of the specification to produce something within each claim without new inventions or additions or prolonged study of matters presenting additional difficulty. (footnote omitted) 887 In SNF (Australia) Pty Ltd v Ciba Speciality Chemicals Water Treatments Ltd at [234], Kenny J explained: A specification is not insufficient merely because some experiment of a routine character (as distinct from prolonged study of matters presenting initial difficulty) is necessary in the particular case … Nor is a specification insufficient because it fails to give detailed instructions as to matters which a "practical person … would naturally settle, and expect to have to settle … himself", provided he "would find no difficulty in so doing". (citations omitted) 888 Further, the plurality in AstraZeneca AB v Apotex Pty Ltd (2014) 226 FCR 324 observed (at [205]): although the complete specification must describe the invention fully, including the best method known to the inventor of performing the invention, it does not follow that the inventor must explain how he or she arrived at the invention. It is the invention itself that must be fully described, not the route that was travelled by the inventor to arrive at it. Again, the way in which the inventor came to the invention described and claimed may also have an evidentiary significance. However, an invention may be the result of chance or luck as much as long experiment and the question whether an invention, or an alleged invention, involves an inventive step is an objective one. (citation omitted)

(a) MLA's contentions 889 MLA contends that the claims are insufficient because there is no clear teaching of how to produce something with each claim. In particular, it is said that the person skilled in the relevant art does not know which three SNPs to choose to be able to predict genetic potential. It is said that this was clear from Professor Plastow's evidence about how he would put claim 1 into practice. 890 Further, MLA contends that where (as here) a claim in effect claims alternative methods, the specification must describe how to perform the invention for each such method, not just one method within the claim. It contends that a patentee should not be permitted to avoid a finding of lack of sufficiency by drafting a claim to cover two or more alternative methods and then only providing a sufficient description as to one. In support of this contention it cites Tramanco Pty Ltd v BPW Transpec Pty Ltd (2014) 105 IPR 18; [2014] FCAFC 23 at [206] to [208], per Nicholas J (referred to by me in GlaxoSmithKline Consumer Healthcare Investments (Ireland) (No 2) Ltd v Apotex Pty Ltd (2016) 119 IPR 1; [2016] FCA 608 at [693]). Nicholas J said at [207] (Allsop CJ and Greenwood J agreeing): In Kimberly-Clark and Doric the High Court was concerned with various product claims each of which consisted of a number of features all of which were essential (in the sense of non-optional) integers of the claim. However, there are some types of claim that may need to be approached slightly differently including, in particular, claims to methods for producing one or more specified results. For example, a claim for a method of producing one or more of outcomes A, B or C might be infringed if the alleged infringer uses the method to produce outcome A, but not outcome B or C. Whether there is infringement in such a case will depend upon (inter alia) the proper construction of the claim and, in particular, whether it requires the use of the method to produce only one or more of outcomes A, B or C, as opposed to all three of them. Assuming the former construction (in the present case it appears to have been assumed by both parties at the trial and on appeal that claim 1 is to be construed in this way), it would seem to me to be wrong in principle to hold that the description of the invention is sufficient if the specification enables the use of the method to achieve outcome A, but not outcomes B or C. It would be inconsistent with the purposes of the Act to confer a monopoly on a patentee for a method of producing any of outcomes A, B or C, if the patentee's disclosure only enabled the use of the method to produce some of those outcomes. 891 Further, MLA pointed out that in Apotex Pty Ltd v Warner-Lambert Company LLC (No 2) (2016) 122 IPR 17; [2016] FCA 1238 at [244], Nicholas J reinforced his observation that the Kimberly-Clark test for sufficiency may not be able to be applied to all forms of claim and said (at [245] to [247]): Tramanco was not concerned with methods of medical treatment, but with a method for determining one or more parameters relating to the performance of the suspension system of a vehicle. In the present case the method claims, as I have construed them, are to methods of treating pain, or different types of pain, in all kinds of mammals including, in particular, humans. As we are reminded from time to time, judgments are not to be construed as if they were statutes, and it is always necessary to read and understand them in context. Application of the test of insufficiency propounded by the High Court in Kimberly-Clark based upon a test originally formulated outside the context of methods of treatment claims cannot be sensibly applied to a case such as this. In my respectful opinion, it would be a nonsense to say that the invention was sufficiently described if it enabled the skilled addressee to perform the method of the invention on rats (or other mammals) but did not enable it to be performed on humans. In my view, s 40(2)(a) requires that the Patent in this case enable the skilled addressee to perform the claimed invention in relation to humans without new inventions or additions or prolonged study of matters presenting initial difficulty. It is only by requiring this degree of enablement that the patentee could be sensibly understood to have given consideration for the grant of a patent which the skilled addressee would understand to be essentially directed to the treatment of pain in humans. 892 Now MLA has relied heavily upon his Honour's observations. But as appropriate insurance, MLA has also reserved the right to argue in a suitable forum that Kimberly-Clark is wrong if I consider that contrary to what Nicholas J has said, Kimberly-Clark can not be so distinguished. 893 MLA says that in the present case, claim 1 specifies one method that requires one of the at least three SNPs to be a specified SNP, and another method that requires one of the at least three SNPs to be within 500,000 nucleotides of a specified SNP. Accordingly, MLA says that even if limb (a) is sufficiently described, which in any event it does not accept, that cannot provide an answer to limb (b) of claim 1. It says that its contention should apply, by parity of reasoning, to different traits. That is, the claims describe a different method for each trait to be identified or inferred. Therefore each method must be sufficiently described for each such trait. 894 Further, MLA says that the specification does not describe the following matters, which are required to perform the invention when construed as being to distinct methods. And the evidence shows that determining each of these matters would have required undue experimentation as at the priority date. It has essentially made four points. 895 First, MLA says that none of the specified SNPs has been shown in the 253 Application to occur in a gene. Accordingly, the person skilled in the art would need to work out if one of the specified SNPs, or a non-specified SNP, was in a gene. Further, given that the majority of genes in the bovine genome had not been identified by December 2002, this would have required significant experimentation. And even using comparative mapping with the human genome, the skilled addressee could not know whether a SNP was within a gene in the bovine genome where there was no equivalent gene in the human genome, as Professor Goddard explained. Further, Professor Taylor's evidence was that this exercise would not necessarily be trivial. Further, as Professor Hayes explained, many of the specified SNPs do not occur in genes. 896 Second, MLA says that none of the specified SNPs was validated in any animals and many will not have any predictive value at all. Further, not all of the specified SNPs were shown to be associated with each of the five traits. Only about 600 were identified as associated for each trait. Further, even then, a mere association is not sufficient to be predictive of genetic potential with any degree of accuracy according to the evidence of Professor Goddard and Professor Visscher. Further, it was submitted that Professor Plastow accepted that the SNPs need to be validated in a different population as having the relevant association. Further, each of Professor Taylor and Dr Sonstegard also agreed that markers need to be validated in another population to be useful in that population. 897 MLA points out that the 253 Application does not validate any SNP / trait association, leaving it to the skilled addressee to do so. But it is said that such work would be time consuming and costly. Professor Goddard said that the amount of work required is comparable to starting from scratch to do a genomic selection project. It involves genotyping and phenotyping work. Further, MLA points to the fact that Professor Hayes was directly involved in the validation work in the study reported in Hayes BJ, Bowman PJ, Chamberlain AJ, Savin K, van Tassell CP, Sonstegard TS and Goddard ME, "A Validated Genome Wide Association Study to Breed Cattle Adapted to an Environment Altered by Climate Change" (2009) 4(8) PLoS ONE e6676. He said that it was a significant amount of work involving 12 to 18 months to run an analysis, and 12 months to extract the DNA and to carry out genotyping. MLA points out that this was carried out in 2007, and that importantly Professor Hayes was able to use a SNP chip. But without access to a SNP chip, Professor Hayes described a genotyping process involving only 6 to 8 SNPs as "just a pain". I would just note at this point that this study tested 152 SNPs in the validation data sets. 898 MLA says that it is no answer to say that any use of the method will require validation. MLA says that if the inventors had, as they should have done, validated the associations for the specified SNPs in a working population of cattle, then the SNPs would have a much higher prospect of being associated with the trait of interest in any population. 899 Third, MLA says that none of the specified SNPs was tested for association with any trait other than the five traits tested. Accordingly, to the extent that the claims extend to other traits, they are not sufficiently described. It is said that that is because the skilled addressee would have to conduct his own association studies involving genotyping and phenotyping significant numbers of animals for the SNPs and trait of interest. That would effectively involve repeating the inventors' work which required undue experimentation in terms of time and cost. 900 Further, MLA says that if the skilled addressee wanted to use one of the specified SNPs or a non-specified SNP for any trait other than the five traits identified in the specification, he would need to re-do Example 2 of the 253 Application. This would involve measuring phenotypes and genotypes for all SNPs to be used with more than 4,000 cattle. This would be a major undertaking that could take one or two years and cost millions of dollars. And after all that, it could show that many SNPs are not valid. 901 Fourth, MLA says that no non-specified SNPs (that is, SNPs within +/- about 500,000 nucleotides of a specified SNP) were identified as being associated with any trait. In this regard, it says that Branhaven accepted that it was very difficult and time consuming to identify a sufficiently large number of SNPs and identify associations between those SNPs and traits. MLA says that the same must apply to any non-specified SNPs. Furthermore, even the task of determining whether a non-specified SNP is in fact within +/-about 500,000 nucleotides of a specified SNP would be extraordinarily time consuming. MLA says that it effectively required the entire bovine genome to be known, as Professor Goddard and Professor Visscher suggested. 902 MLA also referred to Gilead Sciences Pty Ltd v Idenix Pharmaceuticals LLC, where Jagot J found certain claims lacked sufficiency and accepted Gilead's submission (at [624]) that: … a screening process of any significant number of nucleosides - which of course require synthesising before screening - is a research project akin to those in which Biota and Avexa were engaged and exceeds the level of work permitted for a patent to comply with s 40(2)(a). 903 MLA says that the same applies here, as the skilled addressee would have to repeat the work done by the inventors. I would note that her Honour's decision was upheld on appeal (see Idenix Pharmaceuticals LLC v Gilead Sciences Pty Ltd [2017] FCAFC 196). 904 Further, MLA says that Branhaven, through its patent attorneys, has admitted that undue experimentation or invention was required to identify SNPs and determine their association with a specific trait. MLA relies upon the letter from Freehills Patent Attorneys to the Commissioner of Patents dated 11 October 2013 and in particular point 4. MLA says that Branhaven cannot have it both ways. If invention was required to identify relevant associated SNPs, then there was insufficient information in the specification for a person skilled in the relevant art to identify SNPs other than the specified SNPs (limb (a) of claim 1). 905 Generally, and applying the principle in Kimberly-Clark at [25], MLA says that the specification does not: enable the addressee of the specification to produce something within each claim without new inventions or additions or prolonged study of matters presenting initial difficulty. (citation omitted)

(b) Analysis 906 Now as a preliminary matter, I would observe that the observations in Tramanco Pty Ltd v BPW Transpec Pty Ltd at [207] and Apotex Pty Ltd v Warner-Lambert Company LLC (No 2) at [244] have little application to the present case. The claim under consideration in Tramanco at [207] was in form and in substance a claim to a method for producing alternative results or outcomes. It is apparent that his Honour regarded the claim in Apotex v Warner-Lambert in similar terms. But claim 1 of the 253 Application is not a claim to such a method. It is not a claim to a method for producing alternative results or outcomes. Rather it is a claim to a general method that may be employed using any combination of features within the scope of the claim. 907 Further, I agree with Branhaven that MLA's contentions fail to apply the Kimberly-Clark test. The skilled person could readily put the invention into practice, by performing an embodiment of the invention within the scope of each claim. Moreover, this could be done without undue experimentation. And this is all that is required. The evidence does not establish that the skilled person could not use an embodiment of the method within claim 1 of the 253 Application without undue experimentation. For example, he could take any three or more of the 2,510 specified SNPs and use them in the method of the invention to infer the potential for a trait with which they were shown to be associated. In my opinion, the evidence does not clearly establish otherwise. 908 Now MLA has pointed to four matters in respect of which it is said that the description in the specification is deficient. But none of these matters supports any finding of a lack of sufficiency. 909 First, MLA says that the skilled person would need to work out if the SNPs proposed to be used in the claimed method were in genes. But this is something that the skilled person could readily do, as Professor Plastow and Dr Sonstegard explained. At the priority date this could be done by comparing sequence homology with publicly available information concerning the human genome. And from the time of filing of the 253 Application, the bovine genome could be used as the reference point. The location of a SNP within a gene is for molecular geneticists well within the skill of the calling. Moreover, the difficulties and uncertainties said by MLA to be associated with this process are overstated. As a skilled person could employ a method within claim 1 using any three or more SNPs that meet the requirements of the claim, it is not the case that large numbers of SNPs would need to be assessed and located within genes in order to apply the invention. Moreover, it would have been well within the skill of the calling to identify a limb (b) SNP. 910 Second, although MLA says that the SNPs would need to be validated, this is something that the skilled person could readily do. As some of the evidence established, validation is something that would be done as a matter of course in any case where the population of interest is genetically different to the research population. Moreover, I agree with Branhaven that it is not to the point that the 253 Application does not show that the 2,510 specified SNPs were validated in a population other than the research population. There would have been little point to doing this for the purposes of the 253 Application. As a practical matter, the SNPs would have needed to be validated again in any event by the skilled person seeking to practise the invention in his population of interest. It would only be where the population of interest happened to be the same as or genetically similar to the validation population (if there had been validation) in the 253 Application that this could be avoided. Further, in my view the work involved in validating SNPs in a population of interest is routine, and not undue experimentation. I accept Branhaven's submission that it is part of putting the method into effect and would be done as a matter of course. 911 Third, although MLA submits that it would be necessary to re-do at least example 2 for traits other than the five traits the subject of the association study in the 253 Application, such a proposition does not properly apply Kimberly-Clark. The method of claim 1 is not specific as to the trait of interest. It extends to any trait. I agree with Branhaven that enablement in relation to any one or more of the five identified traits is sufficient. Moreover, MLA's characterisation of the work involved in practising the invention in relation to other traits is debatable. As Branhaven says, it would not be necessary to "re-do" example 2 of the 253 Application for this purpose. Example 2 was part of the inventors' GWAS approach that resulted in the development of the panel of 2,510 SNPs associated with a variety of traits. The invention could be put into practice with far fewer SNPs. The method of claim 1 requires only three or more SNPs associated with a trait of interest. I am not satisfied that the skilled person could not identify a sufficient number of SNPs and their association with a trait in order to use the claimed method. 912 Fourth, MLA says that an association study would need to be conducted for non- specified SNPs to identify their association with a trait. But this again fails to apply Kimberly-Clark. 913 Finally, let me briefly deal with two other matters. First, the decision of Jagot J in Gilead Sciences Pty Ltd v Idenix Pharmaceuticals LLC does not assist MLA. Her Honour's acceptance that a "research project" would be required was quite a different context. But in the present case, 2,510 SNPs have been identified as being associated with one or more of five traits. The skilled person can elect to use them in the claimed method. It is a matter for the skilled person if he or she chooses instead to use other SNPs to carry out the method. The fact that this might involve some additional time and effort does not establish insufficiency. Second, MLA's reliance on correspondence from Branhaven's previous attorneys does not carry the matter far. It adds little to my assessment of whether the claimed invention has been sufficiently described to refer to representations made during the course of prosecution. Further, the letter does not in terms admit that, having been provided with the information in the 253 Application, the skilled person would be unable to validate the specified SNPs without undue experimentation. 914 I am not clearly satisfied that MLA's insufficiency ground has been made out.

Lack of fair basis 915 Section 40(3) requires a real and reasonably clear disclosure in the body of the specification of what is claimed. The language of s 40(3) points to a comparison between the claims and what is described in the body of the specification only. As stated by Barwick CJ in Olin Corporation v Super Cartridge Co Pty Ltd (1977) 180 CLR 236 at 240: the question is a narrow one, namely whether the claim to the product being new, useful and inventive, that is to say, the claim as expressed, travels beyond the matter disclosed in the specification. 916 Section 40(3) does not use the word "invention", but it requires that the claims "be fairly based on the matter in it that discusses the 'invention'"; Lockwood (No 1) at [53]. This is the embodiment(s) which is described and around which the claims are drawn, but it does not mean the inventive step taken by the inventor or the advance in the art made by the inventor. Lockwood (No 1) described the relevant test in the following terms (at [68] and [69]): Erroneous principles. The comparison which s 40(3) calls for is not analogous to that between a claim and an alleged anticipation or infringement. It is wrong to employ "an over meticulous verbal analysis". It is wrong to seek to isolate in the body of the specification "essential integers" or "essential features" of an alleged invention and to ask whether they correspond with the essential integers of the claim in question. "Real and reasonably clear disclosure". Section 40(3) requires, in Fullagar J's words, "a real and reasonably clear disclosure". But those words, when used in connection with s 40(3), do not limit disclosures to preferred embodiments. The circumstance that something is a requirement for the best method of performing an invention does not make it necessarily a requirement for all claims; likewise, the circumstance that material is part of the description of the invention does not mean that it must be included as an integer of each claim. Rather, the question is whether there is a real and reasonably clear disclosure in the body of the specification of what is then claimed, so that the alleged invention as claimed is broadly, that is to say in a general sense, described in the body of the specification. Fullagar J's phrase serves the function of compelling attention to the construction of the specification as a whole, putting aside particular parts which, although in isolation they might appear to point against the "real" disclosure, are in truth only loose or stray remarks. (footnotes omitted, save that the italicised phrase is drawn from Rehm Pty Ltd v Websters Security Systems (International) Pty Ltd (1988) 81 ALR 79 at 95 per Gummow J) 917 So, one should not use an "over meticulous verbal analysis". Further, the focus is not on an identity of language between the claims and the disclosure in the body of the specification. Rather, one is looking for a generalised disclosure in the body that provides support for the claims in substance. Moreover, it is inappropriate to isolate in the body of the specification essential integers or features of an alleged invention and to ask whether they correspond with the essential integers of the claim. 918 Now fair basis can be established by a comparison with the consistory clause(s). But even if a claim is based on and mirrors the form of the consistory clause(s), it will not be fairly based if other parts of the specification show that the invention is narrower than the consistory clause(s). What has been described as a "coincidence of language" between a claim and part of the body of a specification does not per se establish fair basing if that part of the language of the specification does not reflect the description of the invention in the light of the specification as a whole. 919 Further, it is appropriate to restate that the complete specification is not to be read in the abstract, but is to be construed in the light of common general knowledge and the relevant art before the priority date. 920 Further, where a feature included in a claim is a limiting feature, there is no need for it to be the subject of an explicit disclosure in the body of the specification, if the subject matter of the claim falls within the scope of what is more broadly described in the specification. As explained in DSI Australia (Holdings) Pty Ltd v Garford Pty Ltd (2013) 100 IPR 19; [2013] FCA 132 at [240] per Yates J: … the inquiry as to fair basis is directed to the question of claim width: see, for example, Olin Corporation at CLR 240 ... A claim may be fairly based for the purposes of s 40(3) of the Act where it adds a feature to a combination otherwise described in the specification and, by that addition, limits the described invention, as a matter of definition, to a more restrictive form than that to which the patentee might otherwise be entitled. In short, a claim may be fairly based for the purposes of s 40(3) of the Act even when all the characteristics by which the invention is defined in the claim are not described in the body of the specification itself, provided those characteristics are truly limiting ones in the sense that I have described. 921 Relatedly, the claims need not be restricted to precise embodiments described in the specification. As Gummow J said in Sartas No 1 Pty Ltd v Koukourou & Partners Pty Ltd (1994) 30 IPR 479; [1994] FCA 936 at 497: it is no objection to any particular claim that it claims a monopoly for less than every feature described in the body of the specification. It cannot be the case that, for example, a claim is restricted to the precise embodiment which is depicted in the body of the specification.

(a) MLA's arguments 922 MLA says that the claims essentially include methods of using any three SNPs from anywhere in the bovine genome that are associated with any trait to identify cattle with that trait. The only restrictions are that at least one of the SNPs is a specified SNP or a non-specified SNP i.e. within +/- about 500,000 nucleotides of a specified SNP, which MLA says covers two thirds of the bovine genome, and at least two SNPs occur on different genes. MLA says that the claims are extraordinarily wide and do not reflect what is described as the invention in the specification. MLA says that they travel beyond the matter described in the specification in at least the following two ways. 923 First, MLA says that there is no basis for a claim to any of the specified SNPs being associated with any trait other than those five traits examined in the association study, which are all carcass traits. It says that the 253 Application does not provide any basis for considering that the specified SNPs are associated with anything other than the five identified traits. Accordingly, insofar as the claims include SNPs that are associated with traits other than those five carcass traits, for example, disease resistance, they travel beyond the invention described in the specification. 924 Second, MLA says that there is no basis for any SNP other than those identified as associated with one of the five traits. It asserts that no association with any trait is shown with any non-specified SNP. The SNPs identified in Example 3 of the 253 Application concern specified SNPs. Further all of the SNPs in use in the LD analysis, and shown to be associated, are one of the specified SNPs. Further, in practice, the +/- about 500,000 nucleotide limitation is not a real limitation, as it covers more than two thirds of the bovine genome. MLA says that there is no basis for the claims to include any SNP that is not a specified SNP. Further, MLA asserts that Branhaven identified that the invention is in the panel of SNPs provided. MLA says that Branhaven's experts also understood the invention as being to a subset of SNPs. MLA says that it is apparent from Professor Plastow's characterisation of the 2,510 specified SNPs as a "windfall" and a "remarkable jump" that this is where he considered the true value of the invention to lie. Accordingly, MLA says that the claims lack fair basis in that they travel well beyond the panel of SNPs identified in the specification.

(b) Analysis 925 I would reject MLA's fair basis attack. 926 First, although MLA contends that there is no basis for a claim to any of the specified SNPs being associated with any trait other than those five traits examined in the association study, I agree with Branhaven that this misapplies the principles of fair basis. 927 Claim 1 is commensurate with the description in the body of the specification. As Branhaven points out, the broad form of the invention described at [0027] is not limited to the use of any particular SNP associated with any particular trait. And I also agree that the embodiment described at [0030] is also not so limited. Further, later embodiments are similarly broadly expressed. And as for paragraphs of the specification that refer to particular traits, including the five traits that were the subject of the association study, these are expressed to be non-limiting, preferred embodiments of the invention: see [0027] and [0090]. 928 In my view such disclosures provide ample fair basis for other traits. Further, I agree with Branhaven that there is no need for the description to establish, by experiment or otherwise, that any of the specified SNPs is in fact associated with any other trait. No authority discussing fair basis and binding upon me supports any such requirement. 929 Further, MLA's argument focuses on particular integers of claim 1, not the claim as a whole. It is the claim that must be fairly based. The claim is to a method for identifying a trait in a bovine subject that comprises, relevantly, identifying in a nucleic acid sample of the subject at least three SNPs associated with the trait. The claim does not require that any of the 2,510 specified SNPs will necessarily be associated with any particular trait whether within or outside the five traits the subject of the association study. They may or may not be. Moreover, the evidence does not establish that the 2,510 specified SNPs would not be useful to infer an association with other traits outside the five traits. 930 Second, as to MLA's contention that there is no basis for a claim to any SNP other than the 2,510 specified SNPs, I would also reject such a contention. There is support in the specification for a claim to a method that involves the use of SNPs other than the 2,510 specified SNPs; see, for example, [0030], [0035] and [0126] in the detailed description. The method described is not limited to the specified SNPs. The identified SNPs are examples of SNPs provided that can be used in the method. But the description teaches that the method can also be employed using other SNPs, provided that they satisfy the elements of claim 1, including being associated with a trait of interest, and for a limb (b) SNP being in relevant linkage disequilibrium (as I have explained elsewhere and as I will require by way of amendment) with the identified limb (a) SNP. 931 I reject MLA's fair basis attack.

Lack of clarity and definition 932 A valid claim is required to define with sufficient certainty the scope of the monopoly being claimed (s 40(3)). Given that a patent is a public instrument, the claim must be defined in such a way that it is not reasonably capable of being misunderstood so that others know the "exact boundaries of the area within which they will be trespassers": Electric & Musical Industries Ld v Lissen Ld (1939) 56 RPC 23 at 39 per Lord Russell of Killowen. A claim will lack clarity if a person skilled in the relevant art cannot ascertain whether what he proposes to do falls within the claim's ambit. 933 But lack of precise definition will not be fatal to the validity of a claim as long as it provides a workable standard suitable to the intended use. But as stated in Kauzal v Lee (1936) 58 CLR 670 at 685 per Dixon and McTiernan JJ: [v]agueness of description, want of particularity and evident indistinctiveness of thought may be the source of so much uncertainty as to the scope of the monopoly that the claim fails to fulfil the requirement of stating with definiteness to what the patentee is exclusively entitled. 934 A claim is clear if either there is no ambiguity or any ambiguity is resolvable by properly construing the claim applying the principles that I have set out previously. But I accept that a claim is bad if no reasonably certain construction can be given to it. And I also accept that I am not bound to find a meaning for a claim nor to approach a claim with the conviction that its language is capable of a reasonable construction when carefully examined. 935 In Flexible Steel Lacing Co v Beltreco Ltd (2000) 49 IPR 331; [2000] FCA 890, Hely J held that each of the method and product claims for pulling lagging used in conveyor systems was invalid for lack of clarity. In respect of the method claim, he said at [107]: The method claim is fairly open to more than one meaning not because of grammatical problems but because, even to a skilled reader, it would not be clear which of two methods claim 13 describes. 936 In respect of the product claim, he held at [131]: Thus, I conclude that the product claim is obscure; it is fairly and equally open to diverse meanings, namely that the sipes run at right angles across the strip, on the one hand, or that the sipes run along the length of the strip on the other. Another possibility is that the claim embraces both. Sometimes, ambiguity or insufficiency in description can be resolved by a skilled addressee through the application of commonsense and common knowledge: cf Innovative Agriculture Products Pty Ltd v Cranshaw (1996) 35 IPR 643 at 666. I do not think that this is such a case. 937 Finally, if a claim is clear, it is not to be made obscure or treated as obscure by taking elements of a preferred embodiment not referred to in the claim and artificially creating obscurity. 938 Let me turn to the question of definition. 939 Section 40(2)(b) requires the complete specification "to end with a claim or claims defining the invention". In General Tire & Rubber Co v Firestone Tyre and Rubber Co Ltd (1971) 1A IPR 121, the Court stated at 167 that "the issue of definition is to be considered as a practical matter and little weight is to be given to puzzles set out at the edge of the claim which would not as a practical matter cause difficulty to a manufacturer wishing to satisfy himself that he is not infringing the patent". The Court also observed that allowances should be made for any difficulties of the case, so that an alleged issue of want of definition should always be considered in relation to the particular facts. It concluded (at 167 and 168): It is clear in our judgment that the question whether the patentee has sufficiently defined the scope of his claims is to be considered in relation to the facts of each case, that allowance is to be made for any difficulties to which the circumstances give rise, and that all that is required of the patentee is to give as clear a definition as the subject matter admits of. It is also clear in our judgment that, while the court is to have regard to all the relevant facts, the issue of definition is to be considered as a practical matter and little weight is to be given to puzzles set out at the edge of the claim which would not as a practical matter cause difficulty to a manufacturer wishing to satisfy himself that he is not infringing the patent. We accept also that definition of the scope of a claim is not necessarily insufficient because cases may arise in which it is difficult to decide whether there has been infringement or not provided the question can be formulated which the court has to answer in the issue of infringement. 940 A claim will be bad if it fails to define the monopoly claimed so that the skilled addressee of the patent can know the exact boundaries of the area within which they will be trespassers. 941 MLA contends that there is no basis for a claim to any of the specified SNPs being associated with any trait other than those five traits examined in the association study, which are all carcass traits. The 253 Application does not provide any basis for considering that the specified SNPs are associated with anything other than the five identified traits. Insofar as the claims include SNPs that are associated with traits other than those five carcass traits, for example, disease resistance, they travel beyond the invention described in the specification. 942 Further, MLA contends that there is no basis for any SNP other than those identified as associated with one of the five traits. No association with any trait is shown with any non-specified SNP. In practice, the +/- about 500,000 nucleotide limitation is not a real limitation, as it covers more than two thirds of the bovine genome. There is no basis at all for the claims to include any SNP that is not a specified SNP. At the priority date, many such SNPs were yet to be discovered (but they have since been discovered). 943 MLA says that as a consequence of the above two matters, each of the claims fails to define the invention described in the specification. As is apparent from my reasons I would reject these contentions. 944 In terms of lack of clarity and for the reasons that I have already expressed in the construction section, claim 1 (and analogous claims) will require amendment to: (a) define "associated" in terms of statistical significance at the p value of equal to or less than 0.01 (or such other measure as I decide after hearing from counsel further); (b) require each of the 3 SNPs to satisfy that level of statistical significance (to be discussed further with counsel); and (c) require the limb (b) SNP to be in LD with the relevant limb (a) SNP and to the requisite degree (to be discussed further with counsel). 945 As presently formulated, claim 1 and analogous claims fail for lack of clarity and proper definition, and also give rise to aspects of inutility. But if such amendments are made, these matters may be rectified. 946 I should note that I have not made any final decision on whether amendment will be permitted and if so its form. I will hear further from the parties on that aspect.

Conclusion 947 As I have indicated in these reasons, aspects of the claims in their present form are deficient in terms of: (a) lack of clarity; (b) a failure to define the invention ; and (c) related to some questions of construction, lack of utility. 948 Accordingly I would uphold MLA's appeal to this extent. But otherwise I would dismiss MLA's appeal. I will not, however, make any final orders until Branhaven has been given the opportunity to consider whether to apply to amend any of the claims to address the concerns that I have expressed in these reasons. 949 I will hear further from counsel as to the course that they suggest should be taken and the orders that they suggest should be made consequential upon these reasons. I certify that the preceding nine hundred and forty-nine (949) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Beach.