Differential admixture in Latin American populations and its impact on the study of colorectal cancer

Abstract

Genome-wide association studies focused on searching genes responsible for several diseases. Admixture mapping studies proposed a more efficient alternative capable of detecting polymorphisms contributing with a small effect on the disease risk. This method focuses on the higher values of linkage disequilibrium in admixed populations. To test this, we analyzed 10 genomic regions previously defined as related with colorectal cancer among nine populations and studied the variation pattern of haplotypic structures and heterozygosity values on seven categories of SNPs. Both analyses showed differences among chromosomal regions and studied populations. Admixed Latin-American samples generally show intermediate values. Heterozygosity of the SNPs grouped in categories varies more in each gene than in each population. African related populations have more blocks per chromosomal region, coherently with their antiquity. In sum, some similarities were found among Latin American populations, but each chromosomal region showed a particular behavior, despite the fact that the study refers to genes and regions related with one particular complex disease. This study strongly suggests the necessity of developing statistical methods to deal with di- or tri-hybrid populations, as well as to carefully analyze the different historic and demographic scenarios, and the different characteristics of particular chromosomal regions and evolutionary forces.

Figure 1. Global admixture analysis performed in ADMIXTURE, with k=3 representing the 3 ancestral components of the Admixed American populations. The barplots show each individual as a vertical line, and the ancestries are indicated by different color (NAM= Native American ancestry, AFR= African ancestry and EUR= European ancestry). a) Parental populations, b) Admixed populations from 1000G and c) Mexican unpublished samples.

Figure 2. Admixture analysis by region performed with ADMIXTURE, with k=3 representing the 3 ancestral components of the Admixed American populations. The barplots show the mean ancestry of each population, and the ancestry proportions are indicated by different colors.

Figure 3. Mean heterozygosity values in four SNPs categories by gene and population. Each bar corresponds to a SNP category in a certain gene and population. SNPs categories are: intronic, non-synonymous, synonymous and UTR.

Figure 4. The graphic shows the number of phased haplotype blocks estimated for the 7 genes detailed per population.

Figure 5. Block characteristics size (in kb) distribution of all haplotype blocks found in the analysis. Summary of haplotype diversity across all blocks.