A comprehensive computational pipeline that bridges single-cell genomics and cheminformatics to accelerate Alzheimer's Disease research. This project integrates advanced bioinformatics and machine learning to create a seamless workflow from raw single-cell RNA sequencing data to predictive drug discovery models.

IntelliForm — Open-source Agentic AI platform for sustainable green chemistry formulations. Natural language → RDKit-powered optimization → ChemRich NJ pilot ready. Contributions warmly welcome from chemists, ML engineers & sustainability builders! 🚀 ChemeNova × ChemRich Global

AI-Powered Drug Discovery

Transferable Atom Equivalent Reconstruction (TAE-RECON) rapidly generates quantum-chemistry-derived molecular descriptors assembled from a pre-computed Transferable Atom Equivalent database of atomic electron-density fragments, skipping time-consuming ab initio calculations, without running a new quantum calculation for each molecule.

This repository introduces a proof-of-concept toolkit for generative virtual screening, combining seed-based molecule generation, QSAR predictions, retrosynthesis, and reaction yield estimation powered by Hugging Face transformer models.

Cheminformatics training covering an intro to drug discovery, molecular file formats and databases, aligning molecules, QSAR, screening and de novo design.

A web server for predicting inhibitors against bacterial target GlmU protein

QSAR modelling

Jupyter notebook and tabular data files to support the publication "Combined In vitro and In silico Workflow to Deliver Robust, Transparent, and Contextually Rigorous Models of Bioactivity" (Charest et al. 2025)

Explore data collection, analysis, and machine learning for drug discovery. Create predictive models, perform EDA, and deploy as web apps.

Post Drug repurposing analysis utilizing ADME screening, toxicity predictions with Tox21, and QSAR modeling to evaluate absorption, distribution, metabolism, and excretion profiles, predict toxicities, and assess molecular activity. This streamlined workflow accelerates drug discovery while ensuring safety and efficacy.

Modular cheminformatics ML pipeline for target-specific bioactivity prediction using ChEMBL data.

QSPR models for predicting density and refractive index of IL–water, IL–ethanol, and IL–isopropanol mixtures.

Some thoughts on building QSARModeling in Rust.