Active Search for Computer-aided Drug Design.

Area of Science:

• Computational Chemistry
• Drug Discovery
• Bioinformatics

Background:

• Idiopathic pulmonary fibrosis (IPF) is a chronic lung disease with significant pharmaceutical interest.
• Alpha-v (αv) integrin receptors, specifically the Arg-Gly-Asp (RGD) integrin receptor family, are implicated in IPF pathogenesis.
• Targeting αv integrin receptors presents a promising therapeutic strategy for IPF.

Purpose of the Study:

• To develop and evaluate a data-driven adaptive Markov chain approach for lead discovery in drug design.
• To identify novel antagonists for αv integrin receptors as potential treatments for IPF.
• To assess the efficacy of the approach in generating structurally diverse and active molecules.

Main Methods:

• Utilized a data-driven adaptive Markov chain for active search in a labeled graph space.
• Employed molecular docking scores against an αvβ6 protein structure as a proxy for binding affinity.
• Used a probabilistic surrogate model to guide the search based on observed activity scores.

Main Results:

• The approach achieved a high structural variety of designed molecules with docking scores exceeding a desired threshold.
• Successfully identified 19 out of 24 previously known active compounds in biological assays.
• The surrogate model suggested novel molecules of significant medicinal chemistry interest, warranting further synthesis and prioritization.

Conclusions:

• The adaptive Markov chain approach is effective for lead discovery in drug design, particularly for targeting αv integrin receptors.
• The method demonstrates the ability to generate diverse and potentially active compounds with high efficiency.
• This approach holds promise for accelerating the development of novel therapeutics for diseases like idiopathic pulmonary fibrosis.