Abstract: Identifying genetic variants associated with complex diseases/traits via genome-wide single nucleotide polymorphisms (SNPs) has proved to be a new and efficient method for studying genetics. With a large number of achievements of genome-wide association study (GWAS), researchers have focused on performing genome-wide SNPs interaction analysis. The search for interaction effects is marked by an exponential growth, not only in terms of methodological development, practical applications and translation of statistical interaction to biological interaction, but also in terms of integration of omics information sources. Many strategies and methods have been applied in detecting interaction analysis, which pro-vides new insights into genetics basis of complex diseases/traits. In this review based on the theory and algorithm realiza-tions, the statistical methods have been sorted into regression, machine learning, Bayesian model, SNP filtering methods and parallel processing methods. Especially, the principle, efficiency and difference of the methods are summaried to offer references to the researchers in this field.

Key words: genome-wide SNPs interaction analysis, complex diseases/traits, statistical methods