Utilizing machine learning in genomic selection of Holstein dairy cattle’s feed efficiency complex

Feed efficiency (FE) complex plays a pivotal role in the economic viability of most livestock production systems and in dairy cattle much progress has been made in enhancing feed efficiency to optimize production outputs. Advancements in genomic selection have the potential to accelerate breeding programs and help tackle the current economic crisis and minimize the environmental impacts of dairy production. However, FE is a complex trait that can be represented by multiple metrics and is influenced by genetic, environmental, and other quantifiable features. Machine learning (ML)-based approaches would help to better integrate these large and complex datasets, aiding the identification of genetic markers associated with FE complex, increasing the accuracy of genomic selection models.
Utilizing high-precision animal phenotype and genotype data available to AFBI, this project will seek to employ a variety of ML algorithms to identify features that influence the feeding efficiency complex. Subsequently, findings will be used to develop feed efficiency complex prediction models based on ML and deep learning techniques. Genetic markers will be identified and integrated across the genotype associated with feeding efficiency metrics using GWAS and ML to improve our understanding of genomic influence on FE complex. The selected cattle population would be included in simulations to model breeding programs and determine the mean genetic gain per generation. The work of this project will aim to provide a data-driven framework that can detect, evaluate and select for cattle which can produce more feed efficient progeny.