Unraveling the Complexity of Genetic Interactions: A Critical Review of Methods for High-Order Epistasis Detection in Genomic Studies

High-order epistasis – referring to complex interactions involving more than two genetic loci – plays a pivotal role in unraveling the genetic architecture underlying complex diseases and traits. However, detecting such interactions remains a formidable challenge due to the combinatorial explosion of possible loci combinations and the inherent limitations of existing computational and statistical methodologies. This review offers a comprehensive examination of current approaches to high-order epistasis detection, spanning from traditional statistical models to emerging machine learning techniques. We systematically classify and compare these methods based on their underlying principles, strengths, and weaknesses, with particular attention to scalability, interpretability, and effectiveness on large-scale genomic datasets. In addition, we explore how recent advancements – especially in deep learning and network-based frameworks – are shaping the future of this field. By highlighting critical challenges and proposing future research directions, this review aims to support researchers in selecting appropriate tools for their investigations and to inspire novel strategies for deciphering the complexities of high-order genetic interactions.