Genome Scans for Population Genomics of Small Ruminants: Which SNP for Which Species and Which SNP Density for Which Study?

NEXTGEN is an on-going project based on a comparative analysis of Whole Genome Sequences (WGS at 10X coverage) at the intraspecific level to optimise genetic management of livestock diversity. Adaptation is investigated in sheep and goats by studying for each species 162 individuals sampled in places representative of the contrasted environmental conditions found all over Morocco. Evaluation of the potential of wild ancestors and traditional breeds to act as reservoirs of neutral and adaptive genetic diversity is conducted by analysing the genomes of 80 individuals (asiatic mouflons, bezoars and local Iranian domestic sheep and goats). This dataset was used to address the key question of the optimal sample size and genotyping strategy to choose for optimizing the time and money dedicated to a population genomic study. We showed how the sample size (10-30 individuals) and the number of SNPs (1000 - 5 Millions and WGS) affected the estimation of population genomic parameters and the detection of signature of selection in small ruminants (O. aries, C. hircus and their wild ancestors O. orientalis and C. aegagrus). We found that 1K SNPs was not sufficient for such estimation but random 10K SNPs allowed a good estimation of heterozygosity, inbreeding coefficient and nucleotide diversity. We showed also that a reliable estimation of LD required at least 500K SNPs. We demonstrated that the Ovis exome capture and commercial 50K-SNP Chips (Ovine and Caprine Illumina® SNP50 Beadchips designed to describe the diversity of industrial breeds) biased the estimations when studying traditional breeds and wild animals.