GWAS & Selection signatures

Temporal dynamics of major SNPs in large broiler chicken populations

---

Abstract

Genome-wide association studies (GWAS) identify genomic regions linked to traits by detecting significant associations between single nucleotide polymorphisms (SNPs) and phenotypes. Major SNPs are often in linkage disequilibrium with quantitative trait loci (QTL), which can change across generations due to selection, migration, and drift. Detection power for QTL effects is influenced by the trait's heritability, population size, and number of genotyped animals. Single-step GWAS (ssGWAS) are computationally demanding because obtaining p-values for SNP effects requires inverting the genomic relationship matrix, which is expensive for large datasets. To overcome this challenge, the Algorithm for Proven and Young (APY) provide an efficient sparse approximation of the inverse genomic relationship matrix , where "core" genotyped animals capture most of the genomic variation. This study aimed to evaluate the consistency of major SNPs across generations of selection in broiler chickens using a large genotyped dataset, identify significant associations based on p-values, and assess the effect of different core sizes for APY on ssGWAS. Data from Cobb Vantress Inc. included 1,515,952 broiler chickens, of which 721,041 were genotyped for 46,199 SNPs. Traits analyzed included body weight, breast meat percentage, and leg score. A multi-trait model was fitted with sex and contemporary group as fixed effects, additive genetic effects for all traits, and a maternal permanent environment effect for body weight. SNP effects were backsolved from genomic estimated breeding values using single-step genomic best linear unbiased prediction (ssGBLUP) and applied in ssGWAS, with p-values computed from using four core sizes, corresponding to eigenvalues explaining 99%, 98%, 49.5% and 33% of the variance of G across generations 1-3, 2-4, 3-5, and 4-6. Additionally, marker effects and their percentage of variance explained were analyzed for the same time periods. Analyses were performed with the BLUPF90 software suite. Using this large dataset, clear association peaks were detected for all traits, with significant SNPs associated with body weight located on GGA1, near KPNA3, CAB39L, and SETDB2. These genes are potentially involved in growth and metabolic regulation. Significant SNPs remained stable across generations, indicating a robust and consistent genetic architecture. Compared with previous studies, the large sample size and the use of p-values allowed the detection of additional signals. Moreover, results obtained using core sizes explaining 99% and 98% of the genomic variation were similar. When using smaller cores explaining 49.5% and 33% of the variance of G, the resolution was reduced for some SNPs. However, some main chromosomal regions were consistently detected, indicating that reduced core sizes still preserve some of the genomic information. Therefore, these results demonstrate that major SNPs remained stable over generations for the evaluated traits. The large sample size provides better genome-wide association resolution, avoiding spurious associations and dubious fluctuations over time.

Keywords: 2026

How to Cite:

Moreno, B., Bermann, M., Bussiman, F., Likness, V. & Lourenco, D., (2026) “Temporal dynamics of major SNPs in large broiler chicken populations”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2280168. doi: https://doi.org/10.31274/wcgalp.23440

Rights: 1