Exploring Landscape of Drug-Target-Pathway-Side Effect Associations

Area of Science:

• Pharmacology and Bioinformatics
• Computational Drug Discovery
• Network Biology

Background:

• Adverse drug reactions are a major cause of drug attrition and mortality.
• Accurate prediction of drug side effects and their mechanisms is vital for advancing drug discovery and clinical practice.

Purpose of the Study:

• To develop and evaluate novel computational algorithms for predicting drug-side effect associations and inferring underlying biological pathways.
• To assess the performance of these algorithms, particularly for rare side effects, compared to existing methods.

Main Methods:

• REMAP (neighborhood-regularized weighted and imputed one-class collaborative filtering) was used to predict drug-side effect associations from networks.
• FASCINATE, an extension of REMAP for multi-layered networks, was applied to infer side effect-drug target associations.
• Statistical methods including random permutation analysis and gene overrepresentation tests were employed to identify side effect-pathway associations.

Main Results:

• REMAP significantly outperformed state-of-the-art methods in predicting drug-side effect associations, especially for rare side effects.
• FASCINATE successfully inferred associations among side effects and drug targets.
• Identified statistically significant associations between side effects and biological pathways, consistent with clinical evidence.

Conclusions:

• The developed algorithms, REMAP and FASCINATE, are effective tools for predicting drug-side effect associations and elucidating mechanisms.
• These computational approaches hold promise for identifying novel drug-side effect relationships and pathways in large-scale biological networks.
• The findings support the potential for improved drug safety and efficacy through advanced computational predictions.