Prediction of Eye, Hair and Skin Color in Admixed Populations of Latin America

Abstract

We report an evaluation of prediction accuracy for eye, hair and skin pigmentation based on genomic and phenotypic data for over 6,500 admixed Latin Americans (the CANDELA dataset). We examined the impact on prediction accuracy of three main factors: (i) The methods of prediction, including classical statistical methods and machine learning approaches, (ii) The inclusion of non-genetic predictors, continental genetic ancestry and pigmentation SNPs in the prediction models, and (iii) Compared two sets of pigmentation SNPs: the commonly-used HIrisPlex-S set (developed in Europeans) and novel SNP sets we defined here based on genome-wide association results in the CANDELA sample. We find that Random Forest or regression are globally the best performing methods. Although continental genetic ancestry has substantial power for prediction of pigmentation in Latin Americans, the inclusion of pigmentation SNPs increases prediction accuracy considerably, particularly for skin color. For hair and eye color, HIrisPlex-S has a similar performance to the CANDELA-specific prediction SNP sets. However, for skin pigmentation the performance of HIrisPlex-S is markedly lower than the SNP set defined here, including predictions in an independent dataset of Native American data. These results reflect the relatively high variation in hair and eye color among Europeans for whom HIrisPlex-S was developed, whereas their variation in skin pigmentation is comparatively lower. Furthermore, we show that the dataset used in the training of prediction models strongly impacts on the portability of these models across Europeans and Native Americans.

Applicability of the IrisPlex system for eye color prediction in an admixed population from Argentina

“…When analyzing eye color prediction made with the IrisPlex system, the prediction of intermediate colors in our population was inaccurate. In fact, none of the samples were predicted to be of intermediate color, showing concordance with other studies of mixed populations (Carratto et al., 2021; Marano et al., 2019; Palmal et al., 2021). It can be observed that in our predictions, intermediate colors show a considerable lack of definition when we use the markers included in the IrisPlex system, making this method more appropriate for a classification into two categories.…”

“…We extracted DNA using Proteinase K (Promega, USA) and LiCl (Gemmell & Akiyama, 1996), and quantified it by using the Nanodrop spectrophotometer (Thermofisher). Then, we genotyped the samples for the six SNPs of the IrisPlex: rs12913832:A>G, rs1800407:C>T, rs12896399:G>T, rs16891982:C>G, rs1393350:G>A, and rs12203592:C>T, and additionally, another four SNPs: rs1129038:C>T, rs7183877:C>A, rs1800410:G>A and rs4778232:T>C. These polymorphisms are located in genes associated with eye color ( HERC2, OCA2, LOC105370627, SLC45A2, TYR , and IRF4 ) (Duffy et al., 2007; Eiberg et al., 2008; Eriksson et al., 2010; Frudakis et al., 2003; Han et al., 2008; Lippert et al., 2017; Liu et al., 2009; Palmal et al., 2021; Ruiz et al., 2013; Visser et al., 2012; Wollstein et al., 2017). We designed oligonucleotides with Primer‐BLAST (Ye et al., 2012) and performed polymerase chain reaction (PCR) using recombinant Taq polymerase (T‐Plus, Inbio‐Highway, Tandil, Argentina), as well as the thermocyclers MPI (La Plata, Argentina) and Biometra T3000 (Biosciences, Dublin, Ireland).…”

“…This way of prediction can be helpful for individual identification in certain criminal investigations, and it is known as forensic DNA phenotyping (FDP) (Koops & Schellekens, 2008). However, the bulk of available studies on EVC genetics is mainly focused on European populations (Liu et al., 2009; Mengel‐From et al., 2009; Shapturenko et al., 2019; Sulem et al., 2007; Walsh et al., 2012), and only more recently, the scientific community began to analyze this topic in admixed populations (Beleza et al., 2013; Edwards et al., 2015; Lloyd‐Jones et al., 2017), especially from Latin America (Carratto et al., 2019, 2020, 2021; Hohl et al., 2018; Marano et al., 2019; Palmal et al., 2021; Silva de Cerqueira et al., 2014). Considering that differences in the genetic pool of a population can influence phenotype predictions based on genotype information (Beleza et al., 2013; Carratto et al., 2019, 2020, 2021; Frudakis et al., 2003; Popejoy & Fullerton, 2016; Sirugo et al., 2019), there is a need for these comprehensive studies in admixed populations.…”