Integrating parental genomes to reduce reference bias and identify intramuscular fat genes in Qinchuan Black pigs

Traditional genomic analysis relies on a single reference genome, which struggles to effectively characterize the genetic diversity among populations. This is due to the substantial genetic differences between the genome of the studied species and the reference genome, potentially introducing reference bias. In this study, we focused on Guanzhong Black pigs (GZB), Danish Large White pigs (DLW), and their hybrid offspring, Qinchuan Black pigs (QCB). We provided two high-quality parental genomes at the chromosomal level and constructed a parental genomic reference panel to detect SNPs (single nucleotide polymorphisms), INDELs (insertions and deletions), and SVs (structural variations). Compared with the single-reference method, the integrated parental genomic strategy identified 5.48% more SNPs and 67.84% more INDELs. The uniformity of variant distribution and genome functional annotation remained consistent before and after integration, while the ratio of non-reference/non-reference genotypes was also improved. In population genetic structure analysis, principal component analysis (PCA) of the three variant types (SNPs, INDELs, and SVs) exhibited good clustering effects, and ADMIXTURE analysis demonstrated consistent stratification. Selection signal analysis based on the integrated parental genomic strategy successfully identified more differentiated windows and positively selected genes. By leveraging multiple variant types and employing two selection signal methods, we jointly identified several novel intramuscular fat candidate genes (MSMO1, SMC6, CCDC158, KIT, CCNC, etc.), which could not be identified by the single-reference method alone. Functional validation of the gene MSMO1 revealed its role in promoting intramuscular adipocyte proliferation and inhibiting adipogenic differentiation. This study is the first to construct a parental genomic reference panel specifically for pig hybrid populations, which significantly reduces reference bias and exhibits superior performance in downstream analyses. This strategy offers new possibilities for genomic selection breeding of livestock and establishes a methodological foundation for precisely dissecting complex traits in hybrid populations.