关键词: 剩余采食量, 随机回归模型, 遗传力, 重复力, 全基因组关联分析

Abstract:

【Objective】In this study, the polynomial coefficients of the random regression model were analyzed with genome wide association study (GWAS) to dissect the genetic architecture of the residual feed intake (RFI) in the lay, so as to lay the foundation for the selection of the RFI in laying hens. 【Method】 Data were collected from an F₂ segregation population of Dongxiang blue-shelled chickens and White Leghorn. The phenotypic data set included body weight, egg size, egg production, and feed intake. The residual feed intake at 40 and 60 weeks of age were calculated with Koch’s method, respectively. Blood samples were collected in F₂ generation. Genomic DNA was extracted with the phenol chloroform method. 1 534 hens were genotyped using 600 K gene microarray. The SNPs data were quality controlled. Missing data was imputed by Beagle software. The principal components were determined by using Plink software. The random regression model was used to analyze the variance components and genetic parameters of the segregation population. Gemma software was used to analyze the polynomial coefficients with a mixed linear model to obtain the significant P-value associated with the RFI in the lay. Gene annotation was carried out on the significant SNP. Chromosome heritability and haplotype block were calculated for RFI polynomial coefficients. 【Result】The second-order Legendre polynomials should be embedded in fixed, additive genetic, and permanent environmental effects. The homogeneous residual variance was harbored in the random regression model. The RFI in the lay showed moderate levels of heritability. The heritability on the RFI ranged from 0.23 to 0.32, and increased with hen age. The repeatabilities were 0.63-0.68, so this indicated that the RFI phenotype value could be accurately estimated by 3-4 repeated measurements. The genetic correlation coefficients in the RFI gradually decreased as the week intervals increased, as did the phenotypic correlation coefficients. The first principal component of the additive genetic matrix accounts for 96.12%. The inflation factor of the GWAS analysis was 1.026, indicating that the principal components of the linear mixed model had eliminated the potential stratified structure. There was a locus close to NAV2 on chromosome 5 that was significantly associated with the RFI, which was supported by the additional 11 SNPs. This QTL explained 1.73% of the phenotypic variance, and the haplotype block size reached 286 kb. Furthermore, potential SNPs were also detected on chromosomes 12 and 27. The potential SNP markers on chromosome 27 were located the upstream of NGFR. The chromosome heritability was directly proportional to the chromosome size, indicating that the small-effect polygenes controlled the RFI. 【Conclusion】The random regression model was used to estimate the genetic parameters on the RFI in the lay. Using polynomial coefficients as GWAS pseudo-phenotypic data, it was found that the NAV2 gene was significantly associated with the RFI.

Key words: RFI, random regression, heritability, repeatability, GWAS