Whole Genome Association Analysis for Detecting QTLs Related to Fat and Protein Production in Buffaloes

Abstract Text:

Whole genome association studies are important for the livestock industry because they allow to incorporate the QTL detected in genetic evaluations, thus enabling greater selection accuracy and faster genetic progress. Therefore, this study aims at identifying loci associated with fat and protein production in river buffaloes. A total of 452 animals (57 males and 395 females) were genotyped using the 90K panel Axiom® Buffalo Genotyping (Affymetrix). For sample quality control, we established the threshold values for: call rate 0.95 and heterozygosity ± 3 standard deviations. For the marker, we adopted call rate>0.98, MAF>0.05 , HWE up to 10^-6, correlation between markers up to 0.998, plus the elimination of coinciding SNPs and with possible errors of physical positioning in relation to the reference map. The number of SNPs left after quality control was 56,716. Statistical analyses were performed using R scripts and the GenABEL software (AULCHENKO et al., 2007). The information used in this studied were the de-regressed breeding values to traits production of fat (FY) and protein (PY), according to Garrick et al., (2009). These data were corrected for population substructure using the principal components obtained by multidimensional scaling of genomic similarity matrix, with residuals weighted by c+(1-r²)/r². The significance level of 0.05 was corrected for Bonferroni. The five SNPs with the highest P-value (candidate SNPs) were selected for each trait, and through their genomic coordinates (BTAU_4.0 assembly), the annotation of the closest genes was taken out using the NCBI database (http://www.ncbi.nlm.nih.gov). After population structure corrections, the inflation factors (lambda) were estimated as 1.0014 and 1.0078 for FY and protein PY; within the acceptable range. At the significance level corrected by Bonferroni, only 2 SNPs were deemed significant for FY and PY. The significant SNPs for both traits were present on chromosome 10. Deiodinase type 2 (DIO2) was the closest gene (~150 Kb). This is the main enzyme that converts Thyroxine (T4) to the active Triiodothyronine (T3) (active form). From the prior knowledge that thyroid hormones directly influence lactation, this gene may explain the greater buffalo hardiness during this phase, adapted to feeding conditions poor in protein.

Keywords: milk quality, buffalo, markers