Machine Learning Scientist, Synthesis‑Aware Drug Discovery

Roche; San Francisco (CA); On-site; USD 148,000 - 274,000; Full time

Benefits offered by this job

Discretionary annual bonus

Benefits described by Roche/Genentech

Job summary

Roche Genentech is seeking a Machine Learning Scientist to advance synthesis-aware molecular design. You will build ML methods connecting design with practical synthesis and integrate proprietary data to drive hit finding and optimization.

The role emphasizes collaboration across computational and experimental teams, publishing work, and contributing to scalable pipelines that interface with automated synthesis platforms.

Qualifications

• Deep machine-learning expertise with a strong foundation in linear algebra and optimization.
• Familiarity with synthesis planning, cheminformatics, and small-molecule data.
• Fluent Python and experience with ML frameworks; strong scientific software development.

Responsibilities

• Develop ML methods for synthesis-aware molecular design across retrosynthesis and generation.
• Integrate proprietary reaction data to design next-gen synthesis-aware models.
• Build scalable pipelines for active-learning with synthesis platforms.
• Design batch synthesis-planning algorithms to maximize chemical-space coverage.
• Publish findings and contribute to open-source releases and talks.
• Collaborate with Roche Genentech computational and experimental teams.

Skills

Machine learning expertise

Linear algebra

Graph neural networks

Sequence models

Reinforcement learning

Python

PyTorch/JAX

Cheminformatics familiarity

Education

PhD or equivalent in ML/comp chem/related field

Tools

RDKit

OpenEye

PyTorch

JAX

Job description

Roche Genentech is seeking a Machine Learning Scientist to advance synthesis-aware molecular design. You will build ML methods connecting design with practical synthesis and integrate proprietary data to drive hit finding and optimization.

The role emphasizes collaboration across computational and experimental teams, publishing work, and contributing to scalable pipelines that interface with automated synthesis platforms.