2P2I HUNTER: a tool for filtering orthosteric protein-protein interaction modulators via a dedicated support vector...

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Protein-protein interactions (PPIs) represent a significant and growing class of therapeutic targets.
• High-throughput screening (HTS) is extensively used to identify modulators of PPIs, necessitating effective chemical libraries.
• The chemical space for PPI inhibitors remains an area of active investigation.

Purpose of the Study:

• To define the general profile of orthosteric PPI inhibitors.
• To develop and validate a computational protocol for filtering screening libraries to enrich for PPI inhibitors.

Main Methods:

• Analysis of 40 non-redundant small molecules from the 2P2I database.
• Development of a support vector machine (SVM) model using 11 Dragon molecular descriptors.
• Validation of the SVM model with external datasets from PubChem BioAssay and in-house screening campaigns.

Main Results:

• The SVM model effectively filters chemical libraries, reducing compound numbers by up to 96%.
• The filtering protocol enhances the proportion of active compounds by a factor of up to 8.
• A defined chemical space for PPI inhibitors was identified.

Conclusions:

• The developed SVM-based filtering protocol provides a valuable tool for the scientific community.
• This approach supports and enhances the efficiency of HTS campaigns for identifying PPI inhibitors.
• The identified chemical space aids in the targeted search for novel PPI-modulating therapeutics.