Two-Stage Machine Learning-Based GWAS for Wool Traits in Central Anatolian Merino Sheep

This study utilized a two-stage machine learning-based GWAS framework to investigate the genetic basis of wool traits in Central Anatolian Merino sheep, using data from 228 animals. The first stage involved feature selection with LASSO, Ridge Regression, and Elastic Net, while the second stage used Random Forest and Support Vector Regression for association modeling, achieving predictive models with R2 up to 0.86. Key candidate genes identified include MTHFD2L and EPGN for fiber diameter, COL5A2 and COL3A1 for staple length, and FAP and DPP4 for greasy fleece yield. The study highlights the effectiveness of machine learning-enhanced GWAS in identifying significant genetic markers and suggests new targets for improving wool quality and yield in sheep breeding programs.