Genomic characterization and breed identification of Wannan Yellow-feathered chicken using whole-genome resequencing and machine learning

Accurate identification of indigenous chicken breeds is fundamental for the conservation and utilization of poultry genetic resources. To enable reliable identification of Wannan Yellow-feathered (WN) chickens, we developed a genomic framework based on whole-genome resequencing data. Comparative analyses of WN (n = 10) and 12 reference populations (n = 60) revealed clear population structure and identified 119 candidate selective regions containing 34,036 SNPs and 41 genes related to metabolism, reproduction, muscle development and behavior. Using these loci, a Random Forest model was constructed to prioritize informative markers, resulting in a compact panel of 41 SNPs. Validation on an independent cohort (n = 40) achieved 92.5% accuracy and an AUC of 0.995. Our results demonstrate that reliable breed identification requires integration of multiple frequency-variable loci rather than reliance on single markers. This study provides both biological insights into WN genetic architecture and a practical tool for breed authentication and conservation.