Expanding large-scale mechanistic models with machine learned associations and big datasets

Abstract

Computational models that can explain and predict complex sub-cellular, cellular, and tissue level drug response mechanisms could speed drug discovery and prioritize patient-specific treatments (i.e., precision medicine). Some models are mechanistic: detailed equations describing known (or supposed) physicochemical processes, while some models are statistical/machine learning-based: descriptive correlations that explain datasets but have no mechanistic or causal guarantees. These two types of modeling are rarely combined, missing the opportunity to explore possibly causal but data-driven new knowledge while explaining what is already known. Here, we explore a combination of machine learning with mechanistic modeling methods to develop computational models that could more fully represent cell-line-specific drug responses. In this proposed framework, machine learning/statistical models built using omics datasets provide high confidence predictions for new interactions between genes and proteins where there is physicochemical uncertainty. These possibly new interactions are used as new connections (edges) in a large-scale mechanistic model (called SPARCED) to better recapitulate the recently released NIH LINCS Consortium large-scale MCF10A dataset. As a test case, we focused on incorporating novel IFN{gamma}/PD-L1 related associations into the SPARCED model to enable description of the cellular response to checkpoint inhibitor immunotherapies. This work is a template for combining big data, machine-learning-inferred interactions with mechanistic models, which could be more broadly applicable towards building multi-scale precision medicine and whole cell models.