MohamedKhattab-pharma/Drug-Discovery-AI-in-CADD

This is a 109-day-old repository with zero stars, forks, or commit velocity—all markers of an unused personal project or educational exercise. The README describes a broad survey of standard CADD techniques (QSAR, docking, MD, ADME) that are well-established in cheminformatics and drug discovery. No evidence of novel methodology, novel combination, or meaningful differentiation from existing tools (RDKit, AutoDock, GROMACS, DeepChem, etc.). The project appears to be a learning portfolio combining commodity libraries and established workflows. Defensibility is minimal: the approach is standard, reproducible by any chemoinformatics practitioner, and lacks adoption, specialization, or network effects. Frontier labs (OpenAI, Anthropic, Google) have no interest in this; they are more likely to integrate specialized pre-trained models or partner with domain platforms (e.g., DeepMind's AlphaFold for structure prediction) than build a generalist CADD reference implementation. The low frontier risk reflects that this is educational scaffolding, not a tool or dataset that would be subsumed into a frontier platform.