Developmental validation of the HIrisPlex system: DNA-based eye and hair colour prediction for forensic and anthropological usage
Introduction
When investigating a case that involves DNA, current forensic practice calls for short tandem repeat (STR) DNA analysis to identify the donor of a biological sample found at a crime scene. However, in certain circumstances an STR profile obtained from evidence material does not match that of a known suspect including any from a criminal DNA database, which often leads to progress in the investigation being halted (so-called cold cases). Due to this, and by taking advantage of scientific progress in the genetic understanding of human appearance, advancement has been made in the DNA prediction of externally visible characteristics (EVCs) [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18]. This concept of DNA intelligence, also termed Forensic DNA Phenotyping (FDP), adds a new dimension to forensic DNA analysis, and provides a promising alternative to help with future investigations in cases where conventional STR profiling fails to identify a crime scene sample donor [19], [20], [21]. At this moment, only eye and hair colour DNA prediction is feasible with accuracies high enough to base police investigations on, while other group-specific EVCs such as skin colour, male baldness, hair morphology are under investigation to identify the underlying DNA variants and to estimate their phenotype predictive value (for recent reviews see [20], [21]).
DNA prediction of eye colour is now achievable when it comes to broad categories such as blue or brown, with systems such as IrisPlex that uses model-based probability prediction [22], [23], [24] and that have been extensively tested [24] and forensically validated [25]. Liklihood association ratio methods [9], a bayesian classifier approach based on likelihood ratios [13] or lastly a prediction guide process [11], amongst others also currently exist for eye colour DNA prediction. Furthermore, DNA prediction of hair colour is already feasible via the HIrisPlex system [26] for parallel prediction of hair and eye colour from DNA, generating leading intelligence for two of the most obvious externally visible traits. Combining our previous eye colour prediction system IrisPlex with knowledge from early systems used for hair colour prediction [1], [10], [27], and considering knowledge from our previous study where we established the hair colour predictive value of a larger number of hair colour associated DNA variants [28], the HIrisPlex system was finally developed [26]. The HIrisPlex system consists of a single multiplex genotyping assay targeting 24 DNA variants identified in European population studies to be highly informative for eye and hair colour prediction [22], [28], as well as two statistical prediction models, one for eye colour and one for hair colour and shade. These prediction models were previously developed from thousands (eye colour) [22] and hundreds (hair colour) [26] of individual genotype and phenotype datasets from European populations. Combined in this highly effective multiplex genotyping assay are the six most eye colour informative SNPs previously used in the IrisPlex system [23], [24], [25] and the twenty-two most hair colour informative DNA variants previously identified as carrying the most hair colour predictive information from a larger number of hair colour associated DNA variants [22], [23], [26], [28]. According to previously established knowledge [22], [23], [24], the use of the IrisPlex eye colour prediction model within the IrisPlex or the HIrisPlex system can correctly predict human blue and brown eye colour with >90% precision, and the use of the HIrisPlex hair colour and shade prediction model and guide can predict on average hair colour accuracies of 79% [26].
In the present study, we performed the developmental validation of the HIrisPlex genotyping assay following the Scientific Working Group on DNA Analysis Methods (SWGDAM) guidelines [29]. These guidelines allow an assessment of the assay for use in certified forensic laboratories. It examines the assay's performance quality and limitations under differing conditions such as sensitivity, reproducibility and concordance as well as non-human species amplification, mixtures, degraded DNA and simulated casework samples. In completing these evaluations, and by demonstrating that the HIrisplex assay fully meets all SWGDAM requirements, we provide a necessary step towards the implementation of the HIrisPlex system in forensic (and other) laboratories to be used for enhancing police investigations in suitable cases. We also introduce here an enhanced prediction model for hair colour, which includes additional individuals to the underlying prediction databases relative to our previously introduced eye and hair colour prediction models. Furthermore, we developed an online web-based tool freely available under www.erasmusmc.nl/fmb/resources for eye and hair colour DNA prediction from complete and partial IrisPlex/HIrisPlex genotype profiles.
Section snippets
Human samples
A selection of human body fluid and tissue samples were collected from donors with informed consent. The hair and eye colour phenotype of these sample donors was also recorded. Test samples included single and multiple source samples, and simulated casework samples (blood, saliva, semen, hair with roots, touched items). DNA was extracted from the samples using the QIAamp DNA Mini kit (Qiagen, Hagen, Germany) according to the manufacturer's guidelines or an in-house extraction protocol
Multiplex design and protocol
The multiplex design of the HIrisPlex assay presented and used here was slightly altered from its previously published version [26] with the aim to improve the peak balance of the several SNPs at the lower concentration levels of input DNA. This involved some adjustment of primer concentrations and a redesign of the N29insA (DNA variant 1) SBE primer (see Table 1). Overall, these modifications enhanced the HIrisPlex assay's performance at the lower DNA input levels, as seen in Fig. 1.
Sensitivity, peak height balance and consistency
HIrisPlex
Conclusions
The developmental validation of the HIrisPlex assay matches all SWGDAM guidelines in terms of sensitivity, stability, reproducibility, precision and accuracy, species specificity, casework samples and population studies. The observed problems with mixture detection are not typical to the HIrisPlex assay but concern all bi-allelic DNA marker systems. HIrisPlex surpasses many genotyping assays currently used in a forensic environment in terms of sensitivity, with only 63 pg minimum DNA input
Acknowledgements
We are grateful to all volunteers who provided DNA samples for this study. We are additionally grateful to S. van Baal and A. Verkerk for help in establishing the HIrisPlex online prediction tool. This work was funded in part by the Netherlands Forensic Institute (NFI), Erasmus MC University Medical Center Rotterdam, and by a grant from the Netherlands Genomics Initiative (NGI)/Netherlands Organization for Scientific Research (NWO) within the framework of the Forensic Genomics Consortium
References (41)
- et al.
A three single-nucleotide polymorphism haplotype in intron 1 of OCA2 explains most human eye-color variation
Am. J. Hum. Genet.
(2007) - et al.
Three genome-wide association studies and a linkage analysis identify HERC2 as a human iris color gene
Am. J. Hum. Genet.
(2008) - et al.
Prediction of eye and skin color in diverse populations using seven SNPs
Forensic Sci. Int. Genet.
(2011) - et al.
Further development of forensic eye color predictive tests
Forensic Sci. Int. Genet.
(2013) - et al.
A polymorphism in the agouti signaling protein gene is associated with human pigmentation
Am. J. Hum. Genet.
(2002) - et al.
DNA-based prediction of human externally visible characteristics in forensics: motivations, scientific challenges, and ethical considerations
Forensic Sci. Int. Genet.
(2009) Forensic DNA Phenotyping: DNA testing for externally visible characteristics
- et al.
IrisPlex: a sensitive DNA tool for accurate prediction of blue and brown eye colour in the absence of ancestry information
Forensic Sci. Int. Genet.
(2011) - et al.
DNA-based eye colour prediction across Europe with the IrisPlex system
Forensic Sci. Int. Genet.
(2012) - et al.
Developmental validation of the IrisPlex system: determination of blue and brown iris colour for forensic intelligence
Forensic Sci. Int. Genet.
(2011)
The HIrisPlex system for simultaneous prediction of hair and eye colour from DNA
Forensic Sci. Int. Genet.
Sequence polymorphism in the human melanocortin 1 receptor gene as an indicator of the red hair phenotype
Forensic Sci. Int.
A new SNP assay for identification of highly degraded human DNA
Forensic Sci. Int. Genet.
Developmental validation of the PowerPlex® ESX 16 and PowerPlex® ESX 17 Systems
Forensic Sci. Int. Genet.
Validation of a single nucleotide polymorphism (SNP) typing assay with 49 SNPs for forensic genetic testing in a laboratory accredited according to the ISO 17025 standard
Forensic Sci. Int. Genet.
SNPs in forensic genetics: a review on SNP typing methodologies
Forensic Sci. Int.
Genetic determinants of hair, eye and skin pigmentation in Europeans
Nat. Genet.
A genome-wide association study identifies novel alleles associated with hair color and skin pigmentation
PLoS Genet.
Association of polymorphic sites in the OCA2 gene with eye colour using the tree scanning method
Ann. Hum. Genet.
Blue eye color in humans may be caused by a perfectly associated founder mutation in a regulatory element located within the HERC2 gene inhibiting OCA2 expression
Hum. Genet.
Cited by (0)
- 1
Current address: Yale Molecular Anthropology Laboratory, Department of Anthropology, Yale University, 10 Sachem Street, New Haven, CT 06511, USA.