Causal reasoning over knowledge graphs leveraging drug-perturbed and disease-specific transcriptomic signatures for drug discovery

Abstract

Network-based approaches are becoming increasingly popular for drug discovery as they provide a systems-level overview of the mechanisms underlying disease pathophysiology. They have demonstrated significant early promise over other methods of biological data representation, such as in target discovery, side effect prediction and drug repurposing. In parallel, an explosion of -omics data for the deep characterization of biological systems routinely uncovers molecular signatures of disease for similar applications. Here, we present RPath, a novel algorithm that prioritizes drugs for a given disease by reasoning over causal paths in a knowledge graph (KG), guided by both drug-perturbed as well as disease-specific transcriptomic signatures. First, our approach identifies the causal paths that connect a drug to a particular disease. Next, it reasons over these paths to identify those that correlate with the transcriptional signatures observed in a drug-perturbation experiment, and anti-correlate to signatures observed in the disease of interest. The paths which match this signature profile are then proposed to represent the mechanism of action of the drug. We demonstrate how RPath consistently prioritizes clinically investigated drug-disease pairs on multiple datasets and KGs, achieving better performance over other similar methodologies. Furthermore, we present two case studies showing how one can deconvolute the predictions made by RPath as well as predict novel targets.

An Open Access Strategy for the Drug Repurposing Community

“…The predictions are always based on the metapaths that are known to be useful. Other methods do the opposite, evaluating every possible metapath [13], [14] that connects the compound to the disease. In the first case, the model is based on expert knowledge rather than data, and the second approach is computationally expensive.…”

Bioinformatics roadmap for therapy selection in cancer genomics

“…P rior CD [ 145 ] makes use of a network propagation algorithm and a drug–drug similarity network, along with pathway activity profiles to prioritise candidate drugs in cancer. Similarly, RP ath [ 146 ] relies on a knowledge graph built from disease, protein and drug causal relations along with disease and perturbed expression signatures to prioritise compounds for a given disease.…”

A knowledge graph approach to predict and interpret disease-causing gene interactions

“…Path-based approaches on KGs represent a promising avenue for inferring new relationships transparently, providing meaningful explanations for these predictions. For instance, the RPath method [ 29 ] reasons over paths within a knowledge graph, guided by transcriptomic information, to prioritise drugs for a given disease and reveal targeted proteins along these paths. In the context of machine learning, path information has also been employed for rule inference in KGs [ 38 – 40 ], enabling the prediction of new facts with a high degree of interpretability [ 41 , 42 ].…”