Posted May 26, 2020

Drug design is fundamentally a data science problem. Naturally, data science starts with the data, with a key challenge being the creation of accurate and highly reproducible data at scale. But data on its own are useless; it has to be effectively interpreted to advance the drug design process. insitro has built a world-class team to push the boundaries of both of these essential elements–the generation of high quality data and the methods for its interpretation–in order to minimize or eliminate key bottlenecks encumbering drug design today.

Biology is in the midst of a reproducibility crisis, with many scientists questioning results and received wisdom. To circumvent this, insitro generates its own massive and in-depth datasets. Critically, insitro brings an engineering approach to biological and chemical data generation, embracing advanced technologies such as robotics, CRISPR, stem cells, and cellular engineering. This yields cell-based disease models that are accurate reflections of the disease phenotype, allowing for rapid yet predictive assays of potential therapeutics, mutations, and other aspects of the disease biology.

This huge wealth of data yields a great opportunity: to marry these datasets with advanced machine learning. We stand at an inspiring nexus where rapid advances in engineering biology are combined with a huge uprising in the power of data science, machine learning (ML), and artificial intelligence (AI). insitro incorporates novel ML and AI methods with these massive data sets, yielding predictions for drug candidates as well as novel biological insights. Moreover, modern ML/AI methods can yield interpretable results that scientists can monitor, understand, and even learn and build upon. 

A final major challenge is not technical, but human. A critical part of insitro’s foundation was the construction of a team that incorporates expertise in both the experimental and computational sides, not prioritizing one over the other. They realize the deep value of integrating large scale biological data and data science throughout the entire process. The team looks dramatically different than most drug development teams (which typically silo biology, chemistry, and computation into separate groups) and instead blend these specialties, starting with the leadership and continuing all the way down. With its visionary founder, Daphne Koller, MacArthur Award winner, co-founder and former co-CEO of Coursera, and scientific pioneer at the interface of machine learning, biology, and medicine, insitro combines an engineering focus, a deep dedication to advancing biology experiments, and an eye to fully harness the power of machine learning.  

By combining these major leaps forward, insitro is poised to be a leader in a new iteration of biopharma, ready to make the transition to engineering the drug development process. insitro is more than a data company and more than a technology company, it is a new hybrid–a new way to build a pharmaceutical company–that will deliver novel medicines to patients via its innovative approach to the problem. I’m delighted to announce we’re leading insitro’s Series B Investment and am excited to join the board and partner with Dr. Koller and her team to revolutionize drug development.