Casework - Problem with low template DNA and hits on the DNA Database

The F Word

Tweet

When does an uncertain result become certain? When you repeat it once? Twice? Ten times? This was an issue at a recent trial involving Low Copy Number (or Low Template) DNA.

Each person inherits two types of DNA (allele) at each area (locus). The same type may be inherited from each parent and therefore only one type detected. When the types are different they are present in equal amounts in the body and the individual is termed a heterozygote at that area. When the same type is inherited from both parents at a specific area then the individual is termed a homozygote. In the former instance, the two types will appear with approximately the same amounts in a ‘standard’ DNA profile; when one type is there the other is usually there. In the latter (homozygote) situation only one type is seen , but at a higher amount.

That is when there are no problems arising from what is called sampling error; errors that arise from the fact that you have so few things to sample that sometimes you miss them. The statistical term for this is ‘stochastic effect’.

The stochastic effect means that as the amount of DNA reduces, one of the types may not be detected because of a phenomenon called 'dropout'. In standard profiling it has been found that if there is a high amount of one type present then it is almost certain that the other type simply isn’t there. This is one way of establishing that there is a contribution from more than one person in a sample – two types at the same area but present in widely different amounts. The technical term is heterozygote balance (or imbalance).

In Low Copy Number or Low Template DNA profiling the relationship between the amount of DNA and the amount apparently detected (usually shown by the height of the peaks in the profile graph – or electropherogram, epg) breaks down because of stochastic effects which cause large variations in peak heights. This means that one type (allele) can appear without the other. This is called allele, or allelic, dropout.

The practical effect of this is that when only one type (allele) appears in Low Template profiles it is not possible to say whether it is because there IS only one, or that the other type has dropped out. As stated by the originators of the technique, “Because of stochastic variation any apparent homozygote is considered to be a potential heterozygote.” The terminology used is that the potentially missing allele is described by the letter F, so the description of that is now ‘n, F’, where n is the number of the detected allele.

Professor Jamieson gave evidence in a case involving LCN in June 2012 in which one locus (D3) produced only a 17 allele in 5 analyses. The National Database was searched with the 17, F (i.e. 17 with any other D3 allele) and produced 9 ‘hits’, including the defendant. The scientist for the Crown excluded the other 8 people by forming the opinion that he was ‘certain’ that it was impossible that another D3 allele had not been detected. This was on the basis of the peak heights and the number of analyses.

Professor Jamieson gave evidence that there was no scientific support for the certainty of that opinion, given the problems associated with Low Template analysis. In response to Crown Counsel’s question as to how many times the analysis would need to produce that result to count as certain that the correct designation was 17 17, he responded that when each test is uncertain then certainty is logically impossible; the designation was a reasonable inference, but not a certainty.

At another locus, only the 28 allele had been detected twice in five analyses, but the Crown scientist did not consider that this was evidence that another allele had not been detected. The question remains; when does uncertain convert to certain in these circumstances?

If the Crown scientist was wrong about the 17 allele then at least another 8 persons matched on the DNA database. If the 28 was the only allele in the sample, then the defendant would have been excluded as the source of the DNA.

2012