Applied AI Scientist, Cheminformatics

A healthier future. It’s what drives us to innovate. To continuously advance science and ensure everyone has access to the healthcare they need today and for generations to come. Creating a world where we all have more time with the people we love. That’s what makes us Roche.

The Opportunity:

Advances in AI, data, and computational sciences are transforming molecular design and development. Roche is leveraging these technologies to accelerate R&D, utilizing data and novel computational models to drive impact across our diagnostics and sequencing platforms. The "Gen-AI for SBX Chemistry" initiative is a strategic effort to harness the transformative power of generative AI to assist our scientists in exploring novel molecular structures and reducing design-to-test turnaround times. 

We are seeking an exceptional AI/ML scientist with a strong background in computational chemistry and a deep interest in molecular foundation models and targeted molecule generation. Ideal candidates are motivated builders who can take ideas from AI research papers and translate them into robust, scalable in-silico models that predict molecular performance. 

• Design and implement state-of-the-art generative AI pipelines to design novel small-molecule candidates optimized for specific performance metrics within our sequencing platforms.

• Design, train, and deploy advanced generative architectures for Computer-Aided Synthesis Planning (CASP), ensuring proposed molecules have highly feasible reaction pathways.

• Build automated machine learning models capable of predicting molecular performance phenotypes from 2D chemical structures, helping chemists prioritize or eliminate candidates prior to synthesis.

• Apply advanced few-shot learning techniques to combine molecular representations learned from massive public databases with Roche’s proprietary, high-quality datasets.

• Fine-tune public models on proprietary data for property prediction and to optimize relevant performance metrics.

• Work closely with experimental chemists and internal stakeholders to integrate in-silico predictions into applied AI frameworks used across our R&D pipeline.

Who you are: 

• You hold a PhD or equivalent advanced research experience in Computational Chemistry, Biophysics, Bioengineering, Computer Science, or a related technical field, and 3+ years of related experience (work experience can be prior or post-grad; relevant post-grad academic lab training will be considered).

• You demonstrate a deep understanding of AI/ML methods specifically applied to molecular modeling and cheminformatics.

• You have hands-on experience building and deploying generative AI architectures, specifically Transformers, Large Language Models (LLMs), Graph Neural Networks (GNNs), Diffusion models, Variational Autoencoders (VAEs), GFlowNets Reinforcement Learning Leraning (RL).

• You have a proven expertise and hands-on experience specifically in Property-Guided Molecule Generation.

• You demonstrate proficiency in Python, C/C++ and experience writing clean, modular, and testable code using standard ML and cheminformatics libraries (e.g., PyTorch, RDKit).

Relocation benefits are not available for this position.

We use artificial intelligence to screen, assess or select applicants for this role.

This posting is for an existing vacancy at Hoffmann-La Roche Ltd.