Volume 2 • Issue 8 • 1000139
J Forensic Res
ISSN: 2157-7145 JFR, an open access journal
Open Access Research Article
Forensic
Research
Valenzuela et al., J Forensic Res 2011, 2:8
http://dx.doi.org/10.4172/2157-7145.1000139
Prediction Model Validation: Normal Human Pigmentation Variation
Robert K. Valenzuela
1,2
, Shosuke Ito
3
, Kazumasa Wakamatsu
3
and Murray H. Brilliant
1,2
*
1
Department of Pediatrics, College of Medicine, University of Arizona, Tucson, AZ 85724, USA
2
Center for Human Genetics, Marshield Clinic Research Foundation, Marshield, WI 54449, USA
3
Department of Chemistry, Fujita Health University School of Health Sciences, Toyoake, Aichi, Japan
*Corresponding author: Center for Human Genetics, Marshield Clinic Research
Foundation, Marshield, WI 54449, USA, Tel: +1 715-207-9493; Fax: +1 715-389-
4950; E-mail: Brilliant.Murray@mcrf.mldclin.edu
Received September 24, 2011; Accepted October 28, 2011; Published October
29, 2011
Citation: Valenzuela RK, Ito S, Wakamatsu K, Brilliant MH (2011) Prediction
Model Validation: Normal Human Pigmentation Variation. J Forensic Res 2:139.
doi:10.4172/2157-7145.1000139
Copyright: © 2011 Valenzuela RK, et al. This is an open-access article distributed
under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the
original author and source are credited.
Abstract
In a past study, we developed multiple linear regression (MLR) models that employed three single nucleotide
polymorphisms (SNPs) that predicted a signiicant proportion of variation in pigmentation phenotypes from a large
population cohort (n=789, training sample). Multiple linear regression models were developed for skin relectance,
eye color, and two aspects of hair color (log of the ratio of eumelanin-to-pheomelanin and total melanin). In this re-
port, using an independent cohort (n=242 , test sample), we 1) externally cross-validated the prediction models, and
2) tested and reined the algorithm presented in the study by Valenzuela and colleagues, (2010). Relative shrinkage
was moderate for skin relectance (23.4%), eye color (19.4%), and the log of the ratio of eumelanin-to-pheomelanin
in hair (37.3%), and largest for total melanin (67%) in hair. Independent construction of predictive models using our
algorithm for the test sample set yielded the same or similar models as the training sample set. Two of the three
SNPs composing the models were the same, with some variability in the third SNP of the model.
Keywords: Forensic Science; Genetics; Human; Pigmentation;
Prediction Models; QTL
Background
According to the Federal Bureau of Investigation (FBI) Laboratory’s
Combined DNA Index System (CODIS) – National DNA Index
System (NDIS) statistics (http://www.bi.gov/hq/lab/codis/clickmap.
htm), there are signiicantly more unmatched proiles than there are
matched proiles. Ancestry informative markers (AIMs) can be helpful
in reducing the pool of suspects. However, a more eicient means
of reducing a pool of suspects is to predict an unmatched proile’s
phenotype based on their genetic information. Forensically informative
phenotypes include skin, eye, and hair color. he appearances of these
traits are largely inluenced by pigmentation, which is a quantitative
trait controlled by many genetic loci.
In developing prediction models, interpretation of correlated
genetic variants can be confounded by population stratiication. When
population stratiication is not accounted for, erroneous inferences of
a gene’s involvement, and therefore false inferences of the biology of
a trait, may be made. Clearly, accounting for population stratiication
is important in determining the biology of a trait. Correlation of a
genetic marker to a trait may result if the marker is the causal variant
that presumably afects the expression/function of a gene, if the
marker is closely linked to a causal variant, or as a result of population
stratiication. Confounding genetic associations are markers that co-
segregate with a trait that varies between populations, allele frequency
diferences are haphazardly associated with a trait due to unique
evolutionary histories of each population. herefore, by deinition,
ancestry informative markers (AIMs) are confounding associations
with respect to a given trait in most instances. However, multiple
studies have demonstrated that speciic AIMs that are associated with
melanin pigmentation are functional [1-3].
Melanin is the main pigment responsible for skin, eye, and hair
color. Variation in a number of genes, including the melanocortin 1
receptor (MC1R), agouti-signaling protein (ASIP), oculocutaneous
albinism 2 (OCA2), solute-carrier transport protein 45A2 (SLC45A2),
and solute-carrier transport protein 24A5 (SLC24A5), have been
associated with pigmentation. Functional and bioinformatics
analyses support the biological role of variants (rs1805007, rs2424984,
rs12913832, rs16891982, and rs1426654) associated with these genes.
he melanocortin 1 receptor (MC1R), a seven transmembrane G-
protein coupled receptor located in the membrane of epidermal and
follicular melanocytes, is a key protein involved in the regulation of
melanin production (reviewed in [4]). Ligands of MC1R include the
paracrine hormones, alpha-melanocyte stimulating hormone (α-MSH)
and adrenocorticotropic hormone; both are produced in the keratino-
cytes associated with the melanocyte. hey are derived from the pre-
cursor protein, proopiomelanocortin (POMC) (reviewed in [5]). he
binding of α-MSH to MC1R causes a cAMP signal cascade resulting
in an increased production of eumelanin. Non-synonymous SNPs, in-
cluding rs1805007, used in this study, have been associated with red
hair and fair skin [4,6]. Functional and bioinformatics studies have
demonstrated that SNP rs1805007 alters the function of MC1R [7-9],
and hence, melanin production.
he protein antagonist to signaling through MC1R is agouti-
signaling protein (ASIP). he antagonistic action of ASIP results in
a relative decrease in the production of eumelanin to pheomelanin
(reviewed in [10]). Hence, MC1R acts as a switch between the two types of
melanin for skin and hair melanoctyes. Several polymorphisms, within
ASIP, including rs6058017 and rs2424984, have been associated with
pigmentation variation. he rs6058017 polymorphism located within
the 3’ un-translated region (UTR) of the ASIP has been associated with
normal human pigmentation variation of the skin [10-12], hair [11],
and eyes [11,13]. In particular, the G allele is associated with increased