Machine Learning in Genome-Wide Association Studies

The primary goal of genome-wide association studies (GWAS) is to detect genetic variants that explain common human diseases. The initial GWAS have achieved an impressive success by identifying thousands of genes associated with a variety of genetic disorders. However, these identified genes mostly contribute to diseases individually and are only able to explain a small portion of the heritability of diseases. It has been realized that complex diseases are more likely attributed to multiple interacting genetic variations. Such as non-linear, non-additive gene-gene interaction effects, i.e. epistasis, render traditional one-gene-at-a-time analysis methods ineffective for GWAS, and thus call for powerful machine learning algorithms that can detect and characterize high-order interactions among multiple genetic variants.

The focus of this Research Topic is on the novel design and application of machine learning algorithms in detecting interacting genetic variants for GWAS. We welcome original research and review articles that address the challenges of detecting gene-gene interactions for large-scale, high-dimension GWAS data, or propose innovative design and application of machine learning techniques for GWAS.
**Please note** Abstracts submission is Optional**