Target-specific compound selectivity for multi-target drug discovery and repurposing.

Area of Science:

• Medicinal Chemistry
• Computational Biology
• Pharmacology

Background:

• Drug molecules often interact with multiple protein targets, leading to both therapeutic effects and side effects.
• Target promiscuity contributes to high attrition rates in drug discovery, increasing costs and time.
• Existing selectivity metrics do not adequately quantify selectivity against a specific target protein.

Purpose of the Study:

• To extend the concept of compound selectivity to target-specific selectivity.
• To develop a method for identifying maximally selective compound-target pairs.
• To provide a systematic approach for discovering potent and selective kinase inhibitors.

Main Methods:

• Decomposing target-specific selectivity into absolute and relative potency.
• Formulating the identification of selective compounds as a bi-objective optimization problem.
• Utilizing a large-scale kinase inhibitor dataset for computational experiments.

Main Results:

• The optimization-based selectivity scoring systematically identifies potent and selective compounds against kinase targets.
• Target-specific selectivity provides deeper insights into kinase inhibitor selectivity and promiscuity compared to existing metrics.
• The developed method is robust against missing bioactivity data and dataset size variations.

Conclusions:

• Target-specific selectivity offers a novel framework for assessing compound selectivity against specific targets.
• The method facilitates the discovery and repurposing of multi-targeting drugs by distinguishing selective from broadly-active inhibitors.
• Empirical p-values were developed to assess the statistical significance of observed selectivity.