Fenotipagem por DNA - variantes em genes humanos que regulam a pigmentação de olhos e de pele : análise fenotípica e genotípica de indivíduos sulbrasileiros para fins forenses

• Autor(a) principal: Sawitzki, Fernanda Rosa
• Data de Publicação: 2017
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da PUC_RS
• Texto Completo: http://tede2.pucrs.br/tede2/handle/tede/7868
• topic: Forense; Pigmentação Humana; Fenotipagem; SNP; CIENCIAS BIOLOGICAS::BIOLOGIA GERAL

Genes related to externally visible characteristics (EVC) present a great importance for the prediction of skin and eye colors in humans. In this study, we tested the ability of a set of SNPs in pigment-related genes to predict skin and eye colors in an admixed south Brazilian population. First, we used the Cinderella - Tiana - Snow White model to analyze the allelic distribution of eight biallelic SNPs in pigment-related. Cinderella and Tiana patterns have respectively low and high melanin content in skin and eyes; Cinderella has white skin and blue eyes (low melanin content; LMC) and Tiana has dark skin and eyes (high melanin content; HMC). Comparative investigation between frequencies of genetic variants in Cinderella-Like versus Tiana-Like subjects may indicate which polymorphic variant is associated with melanin synthesis in skin and eyes. Coordinately, studies with Snow White-Like subjects may be informative to reveal any tissue-specific expression, since these individuals have both white skin (LMC) and dark eyes (HMC). Based on allele frequencies of different human popullations, allele “L” was used for the alleles associated with low melanin content populations (Cinderela-like subjects), and allele “H” was used for the alleles associated with high melanin content populations (Tiana-like subjects). Allelic distribution of eight SNPs showed that 100% of Cinderella- like subjects (N= 73) had less than eight H alleles, and 82% of Tiana-Like subjects (N= 61) had eight or more H alleles. The AUC (Area Under the Receiver Operating Characteristic Curve) value was 0.99, and the calculation of PGL (Pathway Genetic Load) and GP (Genetic Probability) showed that the SNP set presented 93% and 91% concordance between DNA genotype and phenotypes, respectively. Factorial discriminant analyses (FDA) performed in the Snow White group (light skin and dark eyes; N= 116) showed an association between SNPs rs16891982 (SLC45A2), rs8045560 (MC1R), rs1426654 (SLC24A5), rs2733832 (TYRP1), and rs1042602 (TYR) and the Cinderella cluster for skin phenotype, and an association between SNPS rs4778138 (OCA2), rs12913832 (HERC2), and rs916977 (HERC2) and the Tiana cluster for eye phenotype. Lastly, we studied 436 South Brazilian subjects with different skin and eye colors to construct a multinomial logistic regression model able to predict phenotype. The model was tested in 40 random subjects to measure the efficiency of the prediction. The data presented 93% concordance between the predicted and the observed phenotype, showing the successful phenotype prediction of the model in South Brazilian individuals. This is important since Brazil has an ethnically mixed population in which genomic structure may favor epistatic and pleiotropic effects not observed in populations that are more homogeneous. The analyses presented here are an important contribution to forensic DNA phenotyping scenario.