Accelerating BRPF1b hit identification with BioPhysical and Active Learning Screening (BioPALS)

Area of Science:

• Drug discovery and development
• Computational chemistry and cheminformatics
• Biophysics and structural biology

Background:

• Efficient hit identification is crucial for drug discovery.
• Traditional screening methods can be time-consuming and costly.
• Integrating artificial intelligence with biophysical assays offers a promising alternative.

Purpose of the Study:

• To develop and validate a novel, rapid hit identification protocol named BioPhysical and Active Learning Screening (BioPALS).
• To apply BioPALS to identify novel binders for the BRPF1b bromodomain.
• To assess the efficiency, versatility, and data richness of the BioPALS workflow.

Main Methods:

• Development of the BioPALS protocol, combining AI-powered virtual screening with GCI-driven biophysical confirmation.
• Application of BioPALS to screen for binders of the BRPF1b bromodomain.
• Determination of binding kinetics and prediction of binding topologies for identified hits.

Main Results:

• Successful identification of novel micromolar binders for the BRPF1b bromodomain.
• Demonstration of a high in silico/in vitro confirmation rate for BioPALS.
• Characterization of binding kinetics and predicted binding topologies for all identified hits.
• Identification of binders with favorable ADMET properties.

Conclusions:

• BioPALS is a rapid, versatile, and data-rich protocol for hit identification.
• The protocol integrates AI virtual screening with biophysical confirmation effectively.
• BioPALS is applicable to a wide range of biological targets, accelerating drug discovery efforts.