Análise de SNPS em genes de pigmentação humana em indivíduos com alto ou baixo conteúdo de melanina

• Ano de defesa: 2014
• Autor(a) principal: Rodenbusch, Rodrigo
• Orientador(a): Alho, Clarice Sampaio
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul
• Programa de Pós-Graduação: Programa de Pós-Graduação em Biologia Celular e Molecular
• Departamento: Faculdade de Biociências
• País: Brasil
• Palavras-chave em Português: BIOLOGIA CELULAR; BIOLOGIA MOLECULAR; GENÉTICA MOLECULAR; DNA
• Área do conhecimento CNPq: CIENCIAS BIOLOGICAS::BIOLOGIA GERAL
• Link de acesso: http://tede2.pucrs.br/tede2/handle/tede/5997

The understanding of gene function in the externally visible characteristcs (EVC) expression has several uses in human population evolution studies or in forensic investigations. To this last, some effort has been done to discover an efficient and easy model for prediction of skin and eye color in humans. The obvious advantage of the prediction of such EVCs through the use of DNA is to be incorporated as routine in forensic labs and to be applied to police investigations. In our study we combined the genotyping of eight SNPs in pigment-related genes (rs4778138 - OCA2; rs12913832 - HERC2; rs16891982 - SLC45A2; rs8045560 - MC1R; rs1426654 - SLC24A5; rs2733832 - TYRP1; rs1042602 - TYR; rs916977 - HERC2) with different analytical approaches. Considering this SNP panel we evaluated allele frequencies from HAPMAP and ALFRED data obtained from subjects with High Melanin Content (HMC; from African populations), and Low Melanin Content (LMC; from European populations) and defined the alleles H (to predict HMC subjects) and alleles L (to predict LMC subjects). The cumulative distribution of alleles H and alleles L in two phenotypically different color groups of 134 South Brazilian subjects showed that 82% of HMC subjects (N = 61) had eight or more allele H and 100% of LMC subjects (N = 73) had less than eight allele H, with accuracy value of 96.3%. We performed other analyses using AUC (Area Under the Receiver Operating Characteristic Curve), PGL (Calculation of Pathway Genetic Load), and GP (Genetic Probability) approaches. The AUC was 0.99 in predicting both HMC and LMC phenotypes; PGL showed the eight SNPs panel had 93% of concordance between genotype and HMC or LMC phenotypes; and GP approach showed 91% of concordance between prediction and HMC or LMC phenotypes. Our high-throughput genotyping technology combined with different analytical approaches reached very high accuracy to predict the extreme phenotypes of human pigmentation. We believe this forensic DNA phenotyping (FDP) technique would be particularly useful in cases in which the genetic profiles of crime scenes were not found in the DNA data banks or to help classify degraded cadavers skeletons, or biological clues of dismissed people.