Forensic Science International: Genetics
Volume 5, Issue 4 , Pages 358-360, August 2011

A comment on “The hare and the tortoise: One small step for four SNPs, one giant leap for SNP-kind”

  • António Amorim

      Affiliations

    • IPATIMUP, Instituto de Patologia e Imunologia Molecular da Universidade do Porto, Portugal
    • Faculdade de Ciências, Universidade do Porto, Portugal
    • Corresponding Author InformationCorrespondence address: IPATIMUP, R. Dr. Roberto Frias s/n, 4200-465 Porto, Portugal. Tel.: +351 225570700; fax: +351 225570799.

Received 22 January 2010; received in revised form 2 March 2010; accepted 14 April 2010. published online 10 May 2010.

Article Outline

Abstract 

A review recently published in this journal on the future of single nucleotide polymorphisms (SNPs) in the forensic field raised important points, some of which deserve further analysis. A contribution to the discussion of relevant theoretical, methodological and technological aspects, as well as of some legal constraints is presented.

Keywords: Single nucleotide polymorphisms (SNPs), Short tandem repeats (STRs), Forensic genetics

 

A review recently published in this journal [1] presented some important perspectives on the future use of non-STR based Y-chromosome information in forensics. Using a metaphorical approach, the authors forecast that the tortoise (SNPs) will eventually win the forensic race against the hare (STRs). Metaphors in science do have many virtues but entail no fewer dangers, and this one shows both of them. On the positive side, besides having undeniable appeal and a provocative character, it does call the attention of the forensic community to a type of genetic marker that will surely play an increasing role in the field. However, it is doubtful whether the image of a race is applicable or useful and if some of the suggested promises can be expected to be fulfilled. We will try to go deeper into the analysis of these prospects, contributing to the discussion of some theoretical, methodological and technological aspects, as well as some legal constraints.

First of all let us try to identify the runners (SNPs and STRs); otherwise the race (if indeed a race exists) makes no sense and we will not be able to declare the unambiguous winner. In the forensic context, STRs are a class of genetic marker in which the alleles are defined by length differences irrespectively of their sequences. In other words, although previous research has defined the molecular basis for the variation (i.e., the number of tandemly repeated motifs) the current genotyping method does not take into account other possible variations nor is able to discriminate between equally sized alleles (e.g., allele 10.2 at a tetranucleotide STR locus, potentially corresponds to multiple types of deletions or insertions, in both cases occurring either inside the repeat motif or in the flanking sequences).

Conversely, SNPs by definition require some sort of sequencing: the allelic state is determined by the presence/absence of a specific base at a definite location. Accepting these definitions, it becomes clear that the runners in the metaphorical race are not true distinct entities but just the result of the duality created by the instruments used to trace differences, so that a single hare may carry in disguise dozens of small tortoises.

A second major issue has to do with DNA replication processes. Even assuming that available sequencing technologies were error-free, in forensics it is mandatory to distinguish (or at least to be able to provide robust relative likelihoods) between ‘genuine’ non-identity of contributor and mutation. We surely agree that “it should be possible to find a SNP specific for almost any Y chromosome, distinguishing even between fathers and sons” [1], but the consequences of it are so challenging that we may reasonably have doubts about the forensic use of it [2]. Indeed, how to interpret the finding of such a difference between the sequences from a saliva reference sample and from a crime scene hair? How to calculate the relative likelihoods of the alternative hypotheses (a) different contributors, versus (b) same contributor and somatic mutation? In fact, such a situation echoes a long debate on the interpretation of genetic evidence in kinship involving ‘exclusions’. On this issue, the forensic community has moved from an initial naïve dichotomy (a certain minimum number of exclusions implying the exclusion of putative kinship) to reach a consensus in which genetic evidence is evaluated through a likelihood ratio [3] in an unbiased way. Therefore, it is implicit that, in order to be informative, a specific test must provide results which possess significantly different probabilities of being observed under the competing hypotheses. We fear that it is not the case in such a circumstance, and if so, the information will be useless.

In this context, it is certainly useful to recall the history of DNA evidence in forensics. The first DNA typing technique was described by the term genetic or DNA fingerprinting. Although still in use, this expression is now misleading: there is no logical basis for uniting the current use of DNA and the morphological comparisons used in fingerprints [4]. We must also learn from the comparison between the solid success of Y-chromosome use in forensics and the comparatively lagging and controversy-ridden counterpart of mtDNA analysis, including: ‘phantom’ mutations, heteroplasmy and so forth; for a recent example, see [5]. For illustration sake, suppose a specific 15 locus Y chromosome STR profile is found in a putative father and an almost identical profile in the child – with a single one-step difference at a locus. We can trivially deal with this result, provided a mutation rate estimate is available (statistical issues, such as the sufficiency of databases, since they are common to both the non-recombining Y-chromosome and mtDNA will not be addressed here). The situation turns out to be quite different if mtDNA sequence results were at stake in a maternity case. Depending on the specific difference, non-matching evidence could be attributed (but not easily quantitatively evaluated) to quite distinct explanatory hypotheses: (a) non-maternity or (b) true maternity and heteroplasmy, somatic or germinal, technical artifact and so on [6]. The forensic community is far from reaching a consensus on guidelines for such a situation, with approaches discredited in STR analysis proposed [7], [8]. The difficulties in estimating site-specific mutation rates seem to be here to stay, as the recently reported excess of triallelic SNPs in the human genome seems to indicate [9] (with approximately twice as many as would be expected by chance) and mutation recurrence has also been a matter of dispute (e.g. [10], [11]). It should be made clear that a substantial fraction of these challenging issues is not linked to the type of markers used, but is the result of statistically and evolutionarily distinct situations. Indeed, variation uncovered by sequencing includes both rare (which therefore should not be called Single Nucleotide Polymorphisms) and polymorphic mutations (with frequency of 1% or more: true SNPs). In contrast, forensic STRs were specifically selected because they harbor polymorphic alleles in all major human groups. But in any case, rare events require a different approach in interpretation compared with the analysis of common variation.

At this point it might be helpful to ask why forensic genetics enjoys a special status among sister disciplines [4]. This status is not due solely to the recent spectacular technological advances, but to a distinct methodological basis. Quoting [4], “Traditional forensic scientists seek to link crime scene evidence to a single person or object ‘to the exclusion of all others in the world”’, under the assumption of discernible uniqueness. In contrast, forensic genetics works with types of evidence: genotypes, phenotypes, haplotypes and so forth. This means that forensic genetics although relying also on observations, is able to predict the probability of types of occurrences under a solid theoretical basis, clearly outpacing the empirical level of the ‘expert opinion’.

Legal and ethical constraints also bring to bear important limitations upon forensic analyses. In this context, it must be recalled that most database legislations prescribe the use of non-coding DNA (cf. European Council Resolution of 9 June 1997, on the exchange of DNA analysis results (97/C 193/02)). The scientific bases of this restriction are at least debatable [12], but nonetheless it considerably limits the potential of SNPs (for a comprehensive discussion see [13]), particularly in non-recombining genomic segments, since some of them, although located outside coding regions can act as possible tag SNPs or proxies of phenotype determining variations [14], [15], [16]. Conversely, a negligible (or at least much weaker) association between STR alleles and physical variation is observed, due to the much higher mutation rate, which disrupts faster possible associations.

Nevertheless, it should be acknowledged that STRs have indeed strong limitations in forensics, besides the well known difficulty in distinguishing identity-by-state from identity-by-descent. The standard tetranucleotide-motif type STR requires an intact genomic DNA sequence length (well over 100bp) which precludes its application to highly degraded samples. A special class of polymorphisms (indels, bi- or multiallelic) seems particularly promising, as they combine advantages of both STRs and SNPs and their typing can be performed using already validated methodologies [17].

To conclude (and resuming the metaphoric approach), we have no money to bet, but if we had, we would put it equally on both tortoise and hare, not just for prudence but because they will both win this pseudorace.

Back to Article Outline

Acknowledgements 

I thank Leonor Gusmão and two anonymous reviewers for the critical review of the manuscript and helpful suggestions. This work was partially supported by Fundação para a Ciência e a Tecnologia through POCI 2010 (Programa Operacional Ciência e Inovação 2010).

Back to Article Outline

References 

  1. Xue Y, Tyler-Smith C. The hare and the tortoise: one small step for four SNPs, one giant leap for SNP-kind. Forensic Sci. Int. Genet. 2010;4:59–61
  2. Amorim A, Pereira L. Pros and cons in the use of SNPs in forensic kinship investigation: a comparative analysis with STRs. Forensic Sci. Int. 2005;150:17–21
  3. Gjertson DW, Brenner CH, Baur MP, Carracedo A, Guidet F, Luque JA, et al. ISFG: recommendations on biostatistics in paternity testing. Forensic Sci. Int. Genet. 2007;1:223–231
  4. Saks MJ, Koehler JJ. The coming paradigm shift in forensic identification science. Science. 2005;309:892–895
  5. Yao YG, Salas A, Logan I, Bandelt HJ. mtDNA data mining in GenBank needs surveying. Am. J. Hum. Genet. 2009;85:929–933
  6. Parson W, Bandelt HJ. Extended guidelines for mtDNA typing of population data in forensic science. Forensic Sci. Int. Genet. 2007;1:13–19
  7. Carracedo A, Bär W, Lincoln P, Mayr W, Morling N, Olaisen B, et al. DNA commission of the International Society for Forensic Genetics: guidelines for mitochondrial DNA typing. Forensic Sci. Int. 2000;110:79–85
  8. Tully G, Bär W, Brinkmann B, Carracedo A, Gill P, Morling N, et al. Considerations by the European DNA profiling (EDNAP) group on the working practices, nomenclature and interpretation of mitochondrial DNA profiles. Forensic Sci. Int. 2001;124:83–91
  9. Hodgkinson A, Eyre-Walker A. Human triallelic sites: evidence for a new mutational mechanism?. Genetics. 2010;184:233–241
  10. Adams SM, King TE, Bosch E, Jobling MA. The case of the unreliable SNP: recurrent back-mutation of Y-chromosomal marker P25 through gene conversion. Forensic Sci. Int. 2006;25:14–20
  11. Cruciani F, Trombetta B, Novelletto A, Scozzari R. Recurrent mutation in SNPs within Y chromosome E3b (E-M215) haplogroup: a rebuttal. Am. J. Hum. Biol. 2008;20:614–616
  12. Budowle B, van Daal A. Forensically relevant SNP classes. Biotechniques. 2008;44:603–608
  13. Kayser M, Schneider PM. DNA-based prediction of human externally visible characteristics in forensics: motivations, scientific challenges, and ethical considerations. Forensic Sci. Int. Genet. 2009;3:154–161
  14. Zivković M, Stanković A, Dincić E, Popović M, Popović S, Raicević R, et al. The tag SNP for HLA-DRB1*1501, rs3135388, is significantly associated with multiple sclerosis susceptibility: cost-effective high-throughput detection by real-time PCR. Clin. Chim. Acta. 2009;406:27–30
  15. Sudandiradoss C, Sethumadhavan R. In silico investigations on functional and haplotype tag SNPs associated with congenital long QT syndromes (LQTSs). Genomic Med. 2008;2:55–67
  16. Ouyang C, Smith DD, Krontiris TG. Evolutionary signatures of common human cis-regulatory haplotypes. PLoS One. 2008;3:e3362
  17. Pereira R, Phillips C, Alves C, Amorim A, Carracedo A, Gusmão L. A new multiplex for human identification using insertion/deletion polymorphisms. Electrophoresis. 2009;30:3682–3690

PII: S1872-4973(10)00076-1

doi:10.1016/j.fsigen.2010.04.002

Refers to article:

  • The hare and the tortoise: One small step for four SNPs, one giant leap for SNP-kind , 03 September 2009

    Yali Xue, Chris Tyler-Smith
    Forensic Science International: Genetics February 2010 (Vol. 4, Issue 2, Pages 59-61)

Forensic Science International: Genetics
Volume 5, Issue 4 , Pages 358-360, August 2011

Article Tools