LCN – what now? (Published in Barrister Magazine)

A personal commentary published in Barrister Magazine #44 2010

Tweet

What the Court say

The stochastic threshold, consensus profiles, and safe range

Stochastic threshold and quantitation

The problem with LCN

Sub-threshold cases

Vindication and criticism

Introduction

The debate on the use of DNA profiles obtained from very small amounts of cellular material (Low Copy Number (LCN), or Low Template DNA (LTDNA)) took another turn in a judgement delivered by the Appeal Court in December 2009.

Although acknowledging that none of the cases in front of it were challenging the LCN technique that was used in them per se, the Court heard, de benne esse, evidence on behalf of the appellants on the reliability of the LCN process from Dr Bruce Budowle (former head of the FBI’s DNA laboratory) and me.

back to top >>

What the Court say

The Court stated with regard to LCN evidence,

“ i) Low Template DNA can be used to obtain profiles capable of reliable interpretation if the quantity of DNA that can be analysed is above the stochastic threshold – that is to say where the profile is unlikely to suffer from stochastic effects (such as allelic drop out mentioned at paragraph 48) which prevent proper interpretation of the alleles.

ii) There is no agreement among scientists as to the precise line where the stochastic threshold should be drawn, but it is between 100 and 200 picograms.

iii) Above that range, the LCN process used by the FSS can produce electrophoretograms which are capable of reliable interpretation. … However a challenge to the validity of the method of analysing Low Template DNA by the LCN process should no longer be permitted at trials where the quantity of DNA analysed is above the stochastic threshold of 100-200 picograms in the absence of new scientific evidence. …

iv) As we have mentioned, it is now the practice of the FSS to quantify the amount of DNA before testing. There should be no difficulty therefore in ascertaining the quantity and thus whether it is above the range where it is accepted that stochastic effects should not prevent proper interpretation of a profile.

v) There may be cases where reliance is placed on a profile obtained where the quantity of DNA analysed is within the range of 100-200 picograms where there is disagreement on the stochastic threshold on the present state of the science. We would anticipate that such cases would be rare and that, in any event, the scientific disagreement will be resolved as the science of DNA profiling develops. If such a case arises, expert evidence must be given as to whether in the particular case, a reliable interpretation can be made. …”

To the casual reader, the phrase,

“ Low Template DNA can be used to obtain profiles capable of reliable interpretation if the quantity of DNA that can be analysed is above the stochastic threshold”

may be construed as being an open door for the admissibility of LCN evidence. However, this decision needs careful dissection to see the potential impact on the presentation of such evidence because it is these stochastic effects, appearing more frequently as the amount of DNA reduces, that are the trouble. This of course was the central issue that critics, perhaps me being the most prominent in the UK, have been highlighting as the single major problem with the technique. Stochastic variation leads to unreliable interpretation. This view is explicit in the judgement which accepts that the presence of such variation, “prevents proper interpretation of the alleles”.

“ii) There is no agreement among scientists as to the precise line where the stochastic threshold should be drawn, but it is between 100 and 200 picograms.

iii) Above that range, the LCN process used by the FSS can produce electrophoretograms which are capable of reliable interpretation”

back to top >>

The stochastic threshold, consensus profiles, and safe range

This is a significant conclusion when considered in terms of the material offered as validation of the LCN technique: a technique specifically designed to analyse amounts of DNA below 100pg and to produce reliable profiles even in the presence of stochastic effects (allelic dropout in particular). A ‘binary’ or ‘biological’ model was introduced by the proponents ostensibly to account for these stochastic effects. It was recognised that this was an inferior approach to the more complex continuous version. This ‘consensus’ approach required that an allele would only be regarded as being associated with the sample if it occurred twice (in any number of runs).

The Court appears to be saying that the ‘safe’ range for DNA profiles is actually above the level for which the technique was designed, and have agreed that the very phenomenon claimed to have been overcome by the consensus approach is compromised in precisely the way that critics have been saying for some time. I have personally endorsed many LCN profiles because they did not feature stochastic effects. Indeed, in the case of Reed and Reed my statement said that there was no credible challenge to the identification of the alleles because of the lack of stochastic effect in the profiles – it was not an LCN case. The profiles in the Reed case actually showed other problematic features associated with there being too much DNA for the LCN technique; overamplification. Subsequent work in that case showed that a profile could actually be obtained from the samples using the standard approach.

The conclusion of the Court that,

“ … a challenge to the validity of the method of analysing Low Template DNA by the LCN process should no longer be permitted at trials where the quantity of DNA analysed is above the stochastic threshold”

would therefore appear not to alter the pre-Appeal position, at least insofar as challenges presented by me (and I am unaware of any others).

back to top >>

Stochastic threshold and quantitation

The judgment continues,

“ it is now the practice of the FSS to quantify the amount of DNA before testing.”

The entire thrust of my evidence at Omagh, and in many trials since, was that without such quantitation it was impossible for analysts to take reasoned account of stochastic effects. At the Omagh Bomb trial I said in a statement,

“We know that all of these effects are dependent on the amount of starting DNA, yet the LCN technique by definition does not know how much DNA it is testing, and drop-in rates are variable. Therefore it is impossible to know how to reliably incorporate data on allelic drop-out or drop-in.” [drop out is a stochastic effect]

The Court now appears to have accepted this argument, and the FSS now routinely quantify the amount of DNA. The need to quantify to ascertain how little material is there should not be confused with the need, identified in the Caddy Review, to quantify to avoid putting too much DNA into the process to avoid the phenomenon of over-amplification.

“ There should be no difficulty therefore in ascertaining the quantity and thus whether it is above the range where it is accepted that stochastic effects should not prevent proper interpretation of a profile”

back to top >>

The problem with LCN

So, stochastic effects are now accepted by the Court as a problem. This has always been the position of the critics (at least this one!) and may have repercussions in other cases where Courts have admitted LCN evidence despite the presence of significant stochastic effects in the profiles.

In another Appeal case that was originally linked to the Reed case, my written advice read,

“I have examined the profiles produced in this case. The LCN technique frequently suffers from what are termed ‘stochastic effects’. These are the consequence of the extremely low amounts of DNA that the technique is claimed to amplify. These effects include the presence of alleles (pieces of DNA) that are not associated with the sample (termed ‘drop-in’), and the absence of alleles that are associated with the sample (termed ‘drop-out’), both of which are thought to lead to the frequently observed very poor reproducibility in some samples subjected to the technique.

In this case, there was a consistency and completeness [i.e. no stochastic effects] in the profiles and the conclusions of [the scientist] were reasoned and reasonable. I can see no credible challenge to the DNA profiles.”

This clearly demonstrates that the presence of stochastic effects, regardless of the means by which the profile is produced, is the key problem.

The Court goes on,

“There may be cases where reliance is placed on a profile obtained where the quantity of DNA analysed is within the range of 100-200 picograms where there is disagreement on the stochastic threshold on the present state of the science. We would anticipate that such cases would be rare and that, in any event, the scientific disagreement will be resolved as the science of DNA profiling develops. If such a case arises, expert evidence must be given as to whether in the particular case, a reliable interpretation can be made”

back to top >>

Sub-threshold cases

The issue that now arises is what happens to all of the cases with sub 200pg (or unmeasured) levels of DNA presented in previous cases?

Does the Appeal Court judgment now provide ‘new evidence’ regarding the reliability of LCN evidence in past cases where the DNA was never quantified and may show evidence of stochastic effects? If so, how will the expert evidence be prepared and by whom? It is clear that in many cases the defence scientist has simply accepted the LCN technique in line with the dominant UK view of the LCN providers and advised instructing solicitors accordingly.

back to top >>

Vindication and criticism

Despite the criticism levelled by the Court, it would appear that my views have been supported by, among others, the former head of the FBI DNA laboratory and the Appeal Court itself. The court accept that I have,

“ given evidence in so many Low Template DNA cases since then on the strength of the observations in R v Hoey that he has acquired a degree of experience from these cases, his discussion with others and his reading of papers. ”

The ‘papers’ include all of the experimental and validation material disclosed at my request from the laboratories offering LCN. To my knowledge, we are the only organisation to have achieved this depth of assessment of the data, not only for the claimed validation of the technique, but for the experiments done by the same laboratories on transfer of low amounts of DNA between items.

This judgement appears to be an acknowledgement and acceptance of the main criticisms levelled at the use of the LCN technique at the levels of DNA for which it was designed. It remains to be seen what reaction will follow in terms of future and past prosecutions based on the technique. Following the Appeal Court decision in Doheny & Adams, two eminent FSS scientists wrote,

“The recent judgements from the Court of Appeal have provided guidance on the presentation of DNA evidence. This guidance has however, as far as the forensic scientist is concerned, actually added to the confusion” (Lambert, JA, and Evett, IA., The impact of recent judgments on the presentation of DNA evidence. Science & Justice 1998: 38(4), 260-270).

Is history repeating itself?

back to top >>

Professor Allan Jamieson

Tweet