Detection of epistatic interactions with Random Forest
In order to elucidate the influence of genetic factors on phenotype variation, non-additive genetic interactions (i.e., epistasis) have to be taken into account. However, there is a lack of methods that can reliably detect such interactions, especially for quantitative traits. Random Forest was previously recognized as a powerful tool to identify the genetic variants that regulate trait variation, mainly due to its ability to take epistasis into account. However, although it can account for interactions, it does not specifically detect them. Therefore, we propose three approaches that extract interactions from a Random Forest by testing for specific signatures that arise from interactions, which we termed paired selection frequency, split asymmetry, and selection asymmetry. Since they complement each other for different epistasis types, an ensemble method that combines the three approaches was also created. We evaluated our approaches on multiple simulated scenarios and two different real datasets from different Saccharomyces cerevisiae crosses. We compared them to the commonly used exhaustive pair-wise linear model approach, as well as several two-stage approaches, where loci are pre-selected prior to interaction testing. The Random Forest-based methods presented here generally outperformed the other methods at identifying meaningful genetic interactions both in simulated and real data. Further examination of the results for the simulated and real datasets established how interactions are extracted from the Random Forest, and explained the performance differences between the methods. Thus, the approaches presented here extend the applicability of Random Forest for the genetic mapping of biological traits.\n\nAuthor summaryThe genetic mechanisms underlying biological traits are often complex, involving the effects of multiple genetic variants. Interactions between these variants, also called epistasis, are also common. The machine learning algorithm Random Forest can be used to study genotype-phenotype relationships, by using genetic variants to predict the phenotype. One of Random Forests strengths is its ability to implicitly model interactions. However, Random Forest does not give any information about which predictors specifically interact, i.e. which variants are in epistasis.\n\nHere, we developed three approaches that identify interactions in a Random Forest. We demonstrated their ability to detect genetic interactions using simulations and real data from Saccharomyces cerevisiae. Our Random Forest-based methods generally outperformed several other commonly used approaches at detecting epistasis.\n\nThis study contributes to the long-standing problem of extracting information about the underlying model from a Random Forest. Since Random Forest has many applications outside of genetic association, this work represents a valuable contribution to not only genotype-phenotype mapping research, but also other scientific applications where interactions between predictors in a Random Forest might be of interest.