Machine Learning-Driven SNP Identification: Enhancing Genomic Selection in Beef Cattle

Genomic selection has revolutionized livestock breeding by enabling precise genetic predictions based on Single Nucleotide Polymorphisms (SNPs). However, high-dimensional genomic datasets pose challenges such as multicollinearity and overfitting. This study explores the application of penalized regression techniques—LASSO, Ridge, and Elastic Net—to SNP data from chromosome 1 of beef cattle, aiming to identify key genetic markers associated with growth traits. Each method’s strengths and limitations were evaluated using cross-validation and performance metrics like Mean Squared Error (MSE) and Area Under the Curve (AUC). Results demonstrate that Elastic Net outperforms LASSO and Ridge by balancing variable selection and model stability, effectively managing correlated predictors, and achieving superior prediction accuracy. These findings underscore the potential of machine learning-driven genomic analysis to enhance breeding strategies, paving the way for efficient genetic improvements in livestock. Future work will expand the methodology to additional traits, chromosomes, and species to further refine genomic prediction models.