Machine Learning Research Scientist/Senior Machine Learning Research Scientist, Structure and Simulation – AI for Drug Discovery

Roche

Machine Learning Scientist/Senior Machine Learning Research Scientist focusing on AI in Drug Discovery. Engaging in high quality research and innovative model design for transformative medicines.

Posted 6/21/2026full-timeNew York City • California, New York • 🇺🇸 United StatesSenior💰 $141,100 - $312,300 per yearWebsite

ATS Keywords

Tailor your resume

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills

Pythondeep learningPyTorchJAXmolecular dynamics simulationsclassical force fieldsAMBERCHARMMOpenFFgeometric deep learning

Soft Skills

communicationcollaboration

Tools & Technologies

OpenMMRosettaGitHub

Certifications & Qualifications

PhD in Computational BiologyPhD in Computer SciencePhD in ChemistryPhD in Physics

Industry Keywords

drug discoveryprotein engineeringstructural biologybiophysical datasetsdeep learning modelsmultimodal representation learningdiffusion modelsmolecular modelingresearch experiencefirst author publication

Tech Stack

Tools & technologies

PythonPyTorch

About the role

Key responsibilities & impact

Design and train foundation models at scale to answer challenging research questions in large-molecule drug discovery and protein engineering
Leverage massive structural biology and biophysical datasets, building novel architectures that capture complex geometric and physical priors
Contribute to publications and present scientific findings at internal and external venues
Solve real, pressing problems in drug discovery that enable new portfolio capabilities

Requirements

What you’ll need

PhD degree in Computational Biology, Computer Science, Chemistry, Physics or related disciplines, with up to 2 years of industry research experience (Scientist) or 2+ years of industry research experience (Senior Scientist)
Demonstrated experience with Python and deep learning libraries such as PyTorch and/or JAX
Demonstrated experience architecting and training deep learning models, particularly utilizing modern approaches (e.g., multimodal representation learning, geometric deep learning, and diffusion models)
Expertise in molecular dynamics simulations and classical force fields (e.g., AMBER, CHARMM, OpenFF), as well as hands-on experience with molecular modeling tools (e.g., OpenMM, Rosetta)
Demonstrated research experience, including at least one first author publication (or equivalent)
Strong communication and collaboration skills
Public portfolio of computational projects (available on e.g. GitHub)

Benefits

Comp & perks

Relocation benefits are NOT available for this job posting
A discretionary annual bonus may be available based on individual and Company performance