AI · Cheminformatics · Drug Discovery

Designing molecules with artificial intelligence.

I'm Suneel Kumar BVS, Director of AI & Drug Design at Molecular Forecaster. I build machine learning and deep learning systems that turn complex chemistry into faster, smarter decisions across the drug discovery pipeline.

Read the blog View résumé

What I work on

Research & focus areas

Bridging deep learning and medicinal chemistry to design better molecules, faster.

AI-driven molecular design

Generative models (RNN-LSTM, transformers) for de novo design, fragment expansion and hit optimization.

Explainable AI

Interpretable models that tell chemists why a prediction was made — trust you can act on.

Structure-based discovery

Deep learning fused with docking and free-energy methods for structure-based drug discovery.

Toxicity forecasting

Data-driven AI/ML toxicity models — from theory to reproducible, practical pipelines.

10+

Years in AI & cheminformatics

AI-designed candidates proposed (CACHE)

764 nM

IC₅₀ of lead FLT-3 inhibitor

Book chapter on AI toxicity models

Selected projects

Featured work

A few recent projects at the intersection of machine learning and drug design.

Book chapter

Toxicity Forecasts: Data-Driven AI/ML Models

A practical guide to AI/ML toxicity modeling with RDKit, DeepChem and scikit-learn — from theory to working code. Published by Apple Academic Press (Taylor & Francis).

CACHE Challenge

Generative models for fragment-based design

An RNN-LSTM approach for fragment expansion against the SARS-CoV-2 NSP3 Mac1 domain — 91 novel candidates proposed and tested in an open-science effort.

Published

De novo design of a selective FLT-3 inhibitor

AI-assisted de novo design produced PCW-A1001, a highly selective FLT-3 (D835Y) inhibitor with an IC₅₀ of 764 nM. Published in Frontiers in Molecular Biosciences.

From the blog