98. It is important to note that the trace DNA from various contributors may or may not have been deposited by direct contact. Thus, the lack of evidence of direct contact with a person other than the assailant at the time of the attack does not warrant an assumption that there was no such other person contributing to trace DNA upon the clothing of either of the victims. It is, in my view, not possible, if the standard for evidence is proof beyond reasonable doubt, to exclude a reasonable hypothesis simply because that hypothesis is not established as a fact. If it had been the case that the mixed DNA could only have come from one person other than the victim then the likelihood of a match to the accused would to that extent support an inference of guilt. Where a third contributor can be identified then the extent of the contribution of that contributor, including a replication of alleles otherwise attributable to the accused, must be accounted for. It would clearly lower the likelihood ratio of the accused being a contributor. Again, this is an example of the reasoning process of the DNA analysis reversing the onus of proof. It was Dr Lim's assumption, in the case of more than one contributor apart from the victim, that any allele consistent with the accused be assessed as if it was contributed by him. In truth, if the question is what hypothesis exists consistent with the non-involvement of the accused, it should be assumed that he has not contributed any allele to the mix unless all other reasonable possibilities are excluded. Thus trace DNA from the victim, as well as a third person where identified, must be first excluded before the likelihood ratio of identification of the accused as a contributor can be calculated.