Integrating GWAS and eQTL analysis to decipher genetic mechanisms of feed efficiency and feeding behaviors in pigs

Feed constitutes the largest cost in pig farming. However, the genetic mechanisms underlying feeding behavior (FB) and feed efficiency (FE), as well as their relationships, remain poorly understood. This study aims to: (1) identify genetic variants associated with FB, FE, and production traits, while improving genomic prediction (GP) accuracy; (2) investigate the relationships between FB, FE, and production traits; and (3) explore potential links between pig feeding-related traits and human health traits. A total of 358 lead SNPs associated with 28 feeding-related traits were identified through genome-wide association studies (GWAS) in 6938 genotyped pigs. In addition, GP was applied to an independent later-born population to validate the reliability of the GWAS summary statistics. By selection SNPs based on the P-values and linkage disequilibrium (LD), GP accuracy was improved by an average of 13.4%. Furthermore, a pleiotropic QTL located at SSC1:159,658,890–160,953,032 was identified for FB, FE, and production traits. Utilizing the cis-eQTL data from 34 tissues, seven trait-tissue-gene associations were identified. Additionally, mendelian randomization analysis indicated that increased feed intake per visit was genetically correlated with both FE and average daily gain, and eating more in the night might reduce lean meat percentage. Notably, heritability enrichment analysis revealed that GWAS signals for pig feeding behavior traits were significantly enriched in genomic regions associated with human fat intake–related traits. This study identified genetic variants associated with FE, FB, and production traits, and clarified their causal relationships. It represents the first elucidation of the connection between pig feeding traits and human traits. These insights are crucial for molecular breeding in pigs and for evaluating the suitability of pigs as model organisms in biomedical research.