Machine Learning Engineer We are looking for talented Machine Learning Engineers to join Prescient Design, a division devoted to developing structural and machine learning based methods for molecular design within Genentech's Research and Early Development (gRED) organization. The successful candidate will manage projects deploying new techniques for machine learning based molecular optimization for the analysis and design of small and large molecule drugs within target-driven design campaigns. Special focus will be given to engineering pipelines for probabilistic molecular property prediction and Bayesian acquisition for active learning based drug discovery. Additional activities may extend to include engineering pipelines for molecular generative modeling.
The Role:
You will join Prescient Design within the Computational Sciences organization in gRED.Your peers will be machine learning scientists, engineers, computational chemists, and computational biologists.
You will closely collaborate with scientists within Prescient and across gRED.You will develop machine learning and Bayesian optimization workflows to analyze existing, and design new, small and large molecules.
You will be expected to form close working relationships with small molecule and protein therapeutic development efforts across the gRED organization.
You will be expected to work on existing projects and generate new project ideas.
Qualifications:- PhD degree in a quantitative field (e.g., Computer Science, Chemistry, Chemical Engineering, Computational Biology, Physics), or MS degree and 3+ years of industry experience.
- Demonstrated experience with machine learning libraries in production-ready workflows (e.g., PyTorch + Lightning + Weights and Biases)
- Record of achievement, including at least one high-impact first author publication or equivalent.
- Excellent written, visual, and oral communication and collaboration skills.
- Additional desired qualifications:
- Experience with physical modeling methods (e.g., molecular dynamics) and cheminformatics toolkits (e.g., rdkit)
- Previous focus on one or more of these areas: molecular property prediction, computational chemistry, de novo drug design, medicinal chemistry, small molecule design, self-supervised learning, geometric deep learning, Bayesian optimization, probabilistic modeling, statistical methods.
- Public portfolio of computational projects (available on e.g. GitHub).
