Learning high-order interactions for polygenic risk prediction

Abstract

Within the framework of precision medicine, the stratification of individual genetic susceptibility based on inherited DNA variation has paramount relevance. However, one of the most relevant pitfalls of traditional Polygenic Risk Scores (PRS) approaches is their inability to model complex high-order non-linear SNP-SNP interactions and their effect on the phenotype (e.g. epistasis). Indeed, they incur in a computational challenge as the number of possible interactions grows exponentially with the number of SNPs considered, affecting the statistical reliability of the model parameters as well. In this work, we address this issue by proposing a novel PRS approach, called High-order Interactions-aware Polygenic Risk Score (hiPRS), that incorporates high-order interactions in modeling polygenic risk. The latter combines an interaction search routine based on frequent itemsets mining and a novel interaction selection algorithm based on Mutual Information, to construct a simple and interpretable weighted model of user-specified dimensionality that can predict a given binary phenotype. Compared to traditional PRSs methods, hiPRS does not rely on GWAS summary statistics nor any external information. Moreover, hiPRS differs from Machine Learning-based approaches that can include complex interactions in that it provides a readable and interpretable model and it is able to control overfitting, even on small samples. In the present work we demonstrate through a comprehensive simulation study the superior performance of hiPRS w.r.t. state of the art methods, both in terms of scoring performance and interpretability of the resulting model. We also test hiPRS against small sample size, class imbalance and the presence of noise, showcasing its robustness to extreme experimental settings. Finally, we apply hiPRS to a case study on real data from DACHS cohort, defining an interaction-aware scoring model to predict mortality of stage II-III Colon-Rectal Cancer patients treated with oxaliplatin.

A perspective on genetic and polygenic risk scores—advances and limitations and overview of associated tools

“…Tools belonging to this category: AFA-Recur [ 82 ], CanRisk [ 75 ], GDM [ 77 ], SCFA [ 74 ], PXS [ 66 ].…”

“…A more realistic scenario suggests the existence of numerous independent marginal effects alongside a vast array of interaction effects. Current research offers extensions to the PRS methodology [ 81 , 82 ] to address these problematic effects, and showcases a notable role of gene–gene interactions in bipolar disorder.…”

“…Most review articles on PRS present a comprehensive view of different algorithms or point out directly specific challenges [ 14 , 28 , 29 ]. However, a comprehensive overview of the available tools is in need of an update, which we aim to provide with this article.…”

Integrating genetic markers and adiabatic quantum machine learning to improve disease resistance-based marker assisted plant selection

“…Linear mixed models (LMMs) are commonly used for predicting quantitative plant variables, accounting for the polygenic nature of complex traits by modeling the relationship between the observed phenotype and both fixed and random effects (Zhou et al, 2013). Machine learning algorithms, such as support vector machines (SVMs) and random forests (RFs), can also handle large datasets and complex interactions between SNPs and other factors affecting the trait of interest (Massi et al, 2023). Despite challenges like correlation structure among SNP markers and potential bias in the selection of SNP markers, predicting quantitative plant variables using SNP markers holds great promise for improving crop yields, disease resistance, and stress tolerance through marker-assisted selection (MAS) (Difabachew et al, 2023).…”