Essential Regression - a generalizable framework for inferring causal latent factors from multi-omic human datasets

Abstract

High-dimensional cellular and molecular profiling of human samples highlights the need for analytical approaches that can integrate multi-omic datasets to generate predictive biomarkers and prioritized causal inferences. Current methods are limited by high dimensionality of the combined datasets, the differences in their data distributions and their integration to infer causal relationships. Here we present Essential Regression (ER), an interpretable machine learning approach for high-dimensional multi-omic datasets, that addresses these problems by identifying latent factors and their likely cause-effect relationships with the system-wide outcome/properties of interest. ER is a novel data-distribution-free latent-factor regression model that integrates multi-omic datasets and identifies latent factors significantly associated with an outcome. ER outperforms a range of state-of-the-art methods in terms of prediction performance on simulated datasets. ER can be coupled with probabilistic graphical modeling thereby strengthening the causal inferences. ER generates novel cellular and molecular predictions, using multi-omic human systems immunology datasets, pertaining to immunosenescence and immune dysregulation.