Identification of Candidate Genes for Body Weight in Broilers Using Extreme-Phenotype Genome-Wide Association Study

Abstract

Traditionally, genome-wide association studies (GWAS) require maximum numbers of genotyped and phenotyped animals to efficiently detect marker-trait associations. Under financial constraints, alternative solutions should be envisaged such that of performing GWAS with fractioned samples of the population. In the present study, we investigated the potential of using random and extreme phenotype samples of a population including 6,700 broilers in detecting significant markers and candidate genes for a typical complex trait (body weight at 35 days). We also explored the utility of using continuous vs. dichotomized phenotypes to detect marker-trait associations. Present results revealed that extreme phenotype samples were superior to random samples while detection efficacy was higher on the continuous over the dichotomous phenotype scale. Furthermore, the use of 50% extreme phenotype samples resulted in detection of 8 out of the 10 markers identified in whole population sampling. Putative causative variants identified in 50% extreme phenotype samples resided in genomic regions harboring 10 growth-related QTLs (e.g. breast muscle percentage, abdominal fat weight etc.) and 6 growth related genes (CACNB1, MYOM2, SLC20A1, ANXA4, FBXO32, SLAIN2). Current findings proposed the use of 50% extreme phenotype sampling as the optimal sampling strategy when performing a cost-effective GWAS.