Scientist, Structural Chemistry

Role Summary

Scientist, Structural Chemistry will lead cutting-edge AI/ML initiatives in the Modeling Group within Structural Biology & Chemistry at the Foster City, CA site. The role focuses on developing, evaluating, and integrating AI/ML-based technologies to advance Small Molecule discovery. Responsibilities include generating, validating, and implementing AI tools and ML models to increase productivity and improve efficiency of drug discovery programs.

Responsibilities

• Partner with project teams to identify opportunities where ML models can enhance design, prioritization, and hypothesis testing across target classes and discovery stages.
• Develop, evaluate, and benchmark ML models—including geometric deep learning, generative models, and co-folding architectures—for potency, selectivity, and ADMET prediction.
• Work cross-functionally with structural and medicinal chemists to translate computational insights into clear design recommendations.
• Track model performance on active discovery programs; identify failure modes, evaluate applicability domains, and propose improvements.
• Collaborate with Research Informatics & IT teams to deploy models into scalable production environments and maintain computational workflows.
• Communicate capabilities, limitations, and key experimental insights in discovery team meetings.
• Improve structure and potency prediction accuracy: Evaluate, develop, and deploy internal co-folding models on active and retrospective drug discovery programs.
• Enable virtual screening of ultra-large libraries: Assess AI/ML technologies and enhanced sampling methods on internal benchmarks; partner with modelers to apply these technologies on discovery projects.
• Bolster internal generative chemistry design for hit-to-lead and lead optimization: Evaluate multiple scoring paradigms for rapid assessment of chemical space; improve user interfaces to democratize generative workflows.
• Maintain state-of-the-art ADMET models: Train and deploy models at scale; collaborate with key stakeholders (MedChem, DMPK) to enhance adoption and analyze project-specific data.

Skills

• Strong knowledge of deep learning architectures relevant to chemistry and structural biology, including graph neural networks, geometric deep learning, diffusion or flow matching models, and multitask frameworks.
• Strong programming skills in Python and proficiency with ML frameworks (PyTorch, TensorFlow, or JAX).
• Ability to design, implement, and evaluate robust model validation strategies, including uncertainty quantification and applicability domain assessment.
• Expertise with cheminformatics toolkits such as RDKit, OpenEye, or Schrödinger.

Qualifications

• Required: PhD and 0+ years relevant research experience to the position such as postdoctoral roles, a proven track record of publications, or contributions to ML codebases.
• Required: Demonstrated expertise in developing and applying ML models to real-world problems in chemistry, computational chemistry, or materials science.
• Required: Hands-on experience with geometric deep learning, generative chemistry methods, or large-scale molecular modeling.
• Preferred: Background or strong interest in medicinal chemistry, ADMET modeling, or cheminformatics.
• Preferred: Knowledge of small-molecule drug discovery concepts (SAR development, hit-to-lead, lead optimization, ADMET, DMPK assays).
• Preferred: Experience developing software tools, libraries, or user-facing scientific interfaces.