Scientists / Our Scientists

Stephen Ra

Director of Frontier Research, Prescient Design, Computational Sciences

Postdoc Mentor

Years at Genentech

I joined Genentech in 2021 as part of the acquisition of Prescient Design. Prior to joining Prescient, I led Machine Learning at Pfizer Worldwide Research & Development from 2017 to 2020, working on developing methods for a variety of challenging problems in the medicinal sciences and across broader R&D. My background is in statistical machine learning, having worked on my Ph.D. in Computer Science and Statistics. I finished my graduate studies at Columbia University and undergraduate studies at Cornell University.

Featured Publication

Function-guided protein design by deep manifold sampling.

(2021) biorXiv:2021.12.22.473759

Gligorijević, V., Berenberg, D., Ra, S., Watkins, A., Kelow, S., Cho, K., & Bonneau, R.

View Abstract

Black box recursive translations for molecular optimization.

(2019) arXiv:1912.10156

Damani, F., Sresht, V., & Ra.S.

View Abstract

FAAH genetic variation enhances fronto-amygdala function in mouse and human.

(2015) Nature Communications, 6(1)

Dincheva, I., Drysdale, A., Hartley, C., Johnson, D., Jing, D., King, E., Ra, S., Gray, M., Yang, R., DeGruccio A., Huang, C., Cravatt, B., Glatt, C., Hill, M., Casey, B.J., & Lee, F.

View Abstract

Broadly, the Frontier Research group is focused on fundamental concepts, algorithms, and theories in machine learning and their application in enabling scientific discovery. As part of Prescient, we are interested in developing methods in a number of different areas including: adaptive and controllable generation in generative models, uncertainty quantification, robust representation learning, and optimizing the interface between in silico frameworks and in vitro / in vivo experiments.

Attention: This page may not render properly. The browser you are using, Internet Explorer 10, is no longer supported.

Stephen Ra

Director of Frontier Research, Prescient Design, Computational Sciences

Featured Publication

Function-guided protein design by deep manifold sampling.

Black box recursive translations for molecular optimization.

FAAH genetic variation enhances fronto-amygdala function in mouse and human.