Bayesian and Machine Learning Models for Genomic Prediction of Anterior Cruciate Ligament Rupture in the Canine Model

Area of Science:

• Genetics
• Veterinary Medicine
• Bioinformatics

Background:

• Anterior cruciate ligament (ACL) rupture is a common condition leading to osteoarthritis and reduced quality of life.
• Both genetic and environmental factors contribute to ACL rupture risk.
• Dogs, particularly Labrador Retrievers, serve as a valuable genomic model for human ACL rupture due to similar clinical presentation.

Purpose of the Study:

• To explore Bayesian and machine learning models for genomic prediction of ACL rupture in Labrador Retrievers.
• To assess the feasibility of predicting ACL rupture using single nucleotide polymorphisms (SNPs).
• To investigate the impact of non-genetic risk factors on ACL rupture prediction.

Main Methods:

• Genome-wide association studies (GWAS) were leveraged to identify candidate genetic variants.
• Bayesian and machine learning models were applied for genomic prediction.
• Analyses were conducted with and without the inclusion of non-genetic risk factors.

Main Results:

• The study demonstrated the feasibility of predicting ACL rupture from SNPs in Labrador Retrievers.
• Genomic prediction models incorporating non-genetic risk factors achieved near clinical relevance.
• Multiple linear Bayesian and non-linear models showed promising predictive capabilities.

Conclusions:

• This research represents a foundational step towards developing a predictive algorithm for ACL rupture in Labrador Retrievers.
• Accurate prediction of high-risk individuals can facilitate targeted clinical trials.
• Findings have potential benefits for both veterinary and human medicine, aiding in preventative management.