How Chemoproteomics & ML for Drug Discovery

View profile for Dr. Ankita Dutta

Enabling development of accessible & equitable health care products via implementation of innovative strategies |Public Health Professional |Health for all | PhD| AMR | EIDs| Translational Research | Product development

An innovative approach for accelerating Drug discovery applying Chemoproteomics & Machine learning (ML)!! 🧩 The Missing Puzzle Pieces: Most of the proteins in our bodies lack small-molecule partners, making it hard to understand how they work or find treatments for diseases. ➡️ Application of a novel Search Strategy: A method was developed to find these missing partners by testing hundreds of small molecules and developing predictive models using ML classifiers (computer based) to predict how they interact with proteins. 💡Findings: Thousands of interactions between small molecules & proteins were found, including some promising candidates for new treatments targeting specific proteins involved in diseases. 🧑🔬 ML Framework: ML models were developed to predict fragment interactions with native proteins on a proteome-wide scale, including quantitative & qualitative interactome signatures. These models based on user-defined input of target proteins enable prediction of fragment binding to custom protein sets… Or simply put they could predict how these molecules interact with proteins. ❓Impact: The work helps us understand how proteins work & enables accelerating drug discovery efforts.

View profile for Gagan Kukreja

Associate Director, Medicinal Chemistry, SynVent IDD, Syngene

Screening reveals thousands of 'undrugged, yet druggable' proteins Collaboration involving Pfizer develops freely accessible AI tool to probe protein–compound interactions Screening around 400 compounds directly in cells, unearthed around 2300 proteins that are bound including a large fraction of targets for which there are no prior known ligands. https://lnkd.in/ggpB88qX

  • No alternative text description for this image

To view or add a comment, sign in

Explore topics