Multidimensional Boolean Patterns in Multi-omics Data

Motivation: Virtually all biological systems are governed by a set of complex relations between their components. Identification of relations within biological systems involves a rigorous search for patterns among variables/parameters. Two-dimensional (involving two variables) patterns are identified using correlation, covariation, and mutual information approaches. However, these approaches are not suited to identify more complicated multidimensional relations, which simultaneously include 3, 4, and more variables. Results: We present a novel pattern-specific method to quantify the strength and estimate the statistical significance of multidimensional Boolean patterns in multiomics data. In contrast with dimensionality reduction and AI solutions, patterns identified by the proposed approach may provide a better background for meaningful mechanistic interpretation of the biological processes. Our preliminary analysis suggests that multidimensional patterns may dominate the landscape of multi-omics data, which is not surprising because complex interactions between components of biological systems are unlikely to be reduced to simple pairwise interactions.