Genomic Landscape and Prediction of Udder Traits in Saanen Dairy Goats

Abstract

This study focused on advancing genetic selection and productivity in Saanen dairy goats, a key species in the dairy industry of Shaanxi, China, by exploring genomic prediction and the genetic architecture of udder traits. Using genome-wide association studies (GWAS) and phenotypic data from 635 Saanen dairy goats, heritability of udder traits was estimated to range from 0.13 to 0.32, and four candidate genes associated with these traits were identified. Three approaches—GBLUP, Kernel Ridge Regression (KRR), and Adaboost.RT—were compared to evaluate the accuracy and reliability of genomic prediction models for udder size traits. The machine learning models (KRR and Adaboost.RT) demonstrated superior performance, achieving up to 20% and 11% higher prediction accuracy than GBLUP, respectively, indicating their greater stability and reliability. These findings highlight the potential of combining GWAS and machine learning techniques to enhance genetic selection and productivity in dairy goats.