Live at HLTH: Revolutionizing Drug Discovery with Machine Learning—featuring Daphne Koller

Over time, drug development has become more and more challenging.

Success rates of clinical trials hover in the single digits, and the cost of developing a new treatment is greater than $2.5B.

So, what can we do to make drug discovery faster, less expensive and more successful? How might advancements in machine learning and the availability of biomedical data revolutionize the drug design process?

Daphne Koller is the CEO of Insitro, a company that is rethinking drug discovery using machine learning. She spent 18 years as a professor in the computer science department at Stanford before leaving to build the education platform Coursera. In 2016, Daphne returned to her passion for improving human health with machine learning, first as Chief Computing Officer at Calico Labs and then as the Founder of Insitro.

On this episode of The Beat, Daphne joins host Dr. Gautam Gulati to explain how her experience with her father’s autoimmune condition informs her work and why we need to rethink the fundamental categorization of disease. Daphne describes how releasing machines from our preconceptions of what’s important uncovers new science around the drivers of disease and serves as a critical starting point for developing new interventions. Listen in for Daphne’s insight on leveraging machine learning to improve clinical trials and learn why data collection should be part of the fabric of every biopharma company.

Topics Covered

Daphne’s background in machine learning, biology and medical dataHow Daphne’s experience with her father’s autoimmune disease informs her work at InsitroWhy we need to rethink the fundamental categorization of what disease isDaphne’s insight on the history of machine learning and the danger in overhyping what the technology can doThe pros and cons of using end-to-end learning to make predictionsHow releasing machines from preconceptions of what’s important helps uncover new scienceWhy understanding the drivers of disease is a critical starting point for developing new interventionsWhy data collection should be part of the fabric of every biopharma companyHow Daphne’s work in machine learning can be used to improve clinical trialsWhy now is the right time for a company like InsitroHow Daphne thinks about privacy and issues of informed consent

 

Connect with Daphne Koller

Insitro
 

Connect with Dr. Gautam Gulati

HLTH

Dr. G. on LinkedIn

Dr. G. on Twitter
 

Resources

Coursera

Art Levinson at Calico

Dr. Hal Barron at GSK

Aducanumab

Chasing My Cure: A Doctor’s Race to Turn Hope into Action by David Fajgenbaum

Insitro’s Partnership with Gilead

UK Biobank

 

Introductory Quote

[5:34] “What I really wanted to build was a company that rethought drug discovery and development from the ground up, using machine learning as a foundational tool.”