Post by Halimat A.

Last weekend I had a great time at the Building an AI Scientist Hackathon, where scientists, engineers, chemists, and biologists teamed up to reimagine the future of drug discovery with AI. I'm honoured to have been selected by this talented group of peers to win the People's Choice Award 🏆 for my entry, Stratified Precision. (Watch the demo here! https://lnkd.in/e9TEp6Kx). My career path has been very non-traditional, starting as a doctor in the NHS before pivoting into ML research across academia and Pharma. Stratified Precision grew directly out of that journey and how I see the future of target identification unfolding. Most drug discovery efforts fail at the clinical trial stage. Traditional target identification is bottom-up: start with a target, then do years of research and hope it pays off downstream. I believe the future is top-down and patient-first. Since patients with the same clinical diagnosis can have very different underlying biology, Stratified Precision flips the paradigm, we start with the patient population, stratify them into their true biological subgroups, and then prioritise the best drug targets for each one. This is done via an agentic AI pipeline that turns messy clinical data into meaningful patient subgroups, surfaces the most promising targets for each, and ranks them by the attributes that matter most. It also comes with a personalised AI co-scientist you can chat with in real time, one that understands your data, draws links across diagnostic labels, and surfaces opportunities like cohort synergies and drug repurposing signals. Huge thanks to all the hosts and sponsors! Ternary Therapeutics, future.bio, pluto house, Anthropic, daphni, redalpine, Amazon Web Services (AWS), Boltz, Fifty Years, Advanced Research + Invention Agency (ARIA), Amass, Elicit, o2h technology. Excited for what's ahead in AI x drug discovery! 🧬 #DrugDiscovery #AI #MachineLearning #PrecisionMedicine #Biotech #AIinPharma