Daphne Koller, insitro | Stanford Women in Data Science (WiDS) Conference 2020
summary
insitro builds an end-to-end, data-first drug discovery stack that couples half life scientists, half ML teams to design experiments, generate massive biological datasets (high‑throughput assays, sequencing, image-based readouts) and train models that surface nonintuitive targets and biomarkers. The concrete claim is measurable: increase drugs approved per R&D dollar to bend Eroom’s Law downward. Workflows include experiment design, model training, and lead selection — the company is framing success in faster, cheaper validated preclinical candidates.
AI-written summary · about
chapters (11)
- 0:00Introduction and welcome
- 0:43Founding insitro and its mission
- 1:47Eroom's law in drug development
- 2:27Interdisciplinary culture at insitro
- 3:08Founding Coursera and its impact
- 4:51Transition from education to biomedicine
- 6:18Advice for aspiring data scientists
- 7:42Role of technical background in leadership
- 8:45Future of insitro and NASH treatment
- 9:49Digital biology in 2020
- 12:15Closing remarks
related talks
21mShaping the future of health: Boehringer Ingelheim's Data & AI Journey
34mFrom Data to Discovery: How AI Is Finding Hidden Cures
43mDomain-driven Data Science
08mEngineering the Future of Pharma: AI-Driven Drug Discovery
42mApplications in ML Drug Discovery pipelines
39mData Challenges and AI Impact in Research-Driven Organizations
39mThe Translation Gap
16mBeyond Patterns: The Case for Physics-Grounded AI in Drug Discovery