Associate Principal Scientist, AI for Chemical Toxicology at AstraZeneca in Beijing Yizhuang

Job description

• In this role you will lead the next‑generation predictive safety modelling at scale, applying chemoinformatics, bioinformatics and AI/ML to deliver decision-shaping safety insights and cross-functional scientific leadership. Provide scientific leadership for predictive safety modelling, integrating chemical and biological data (e.g., Safety Omics, Cell Painting, NAMs) to inform risk assessment and progression decisions. This is a fully hands-on, computational role with cross-functional influence.

Responsibilities

• · Develop, optimise and deploy predictive safety models using chemoinformatics, bioinformatics and ML/AI.
• · Create reproducible workflows, data QC procedures and model validation frameworks.
• · Integrate multimodal data: chemical descriptors, omics, imaging and NAM datasets.
• · Serve as scientific lead for project teams and influence program strategy.
• · Drive innovation and identify strategic modelling opportunities across R&D.
• · Lead collaborations with academia, consortia and technology partners.
• · Mentor junior scientists and contribute to capability building.
• · Communicate complex modelling concepts to diverse scientific and governance audiences.

Requirements

• · PhD (or equivalent years of experience) in mathematics, computer science, engineering, physics, statistics, computational sciences, chemoinformatics, bioinformatics, computational toxicology or a related field.
• · Demonstrated experience in implementing machine learning/AI workflows to automate bioinformatics or cheminformatics analysis pipelines, ideally in the Pharma and/or Healthcare space
• · Advanced Python programming skills.
• · Familiarity with GitHub, CI/CD pipeline, and best DevOps and MLOps practices
• · Familiarity with existing machine learning models for chemical structure modelling: GNNs (graph neural networks), directed message-passing neural network (D-MPNN) as well as foundation models (e.g., CheMeleon). Experience with multimodal biological + chemical datasets.
• · Proven leadership in delivering impactful modelling work.
• · Excellent communication and stakeholder engagement skills.
• · Acts as discipline leader and shapes scientific strategy.
• · Solves complex problems using scientific judgement.
• · Builds strong cross-functional relationships and networks.
• · Coaches and mentors others, enhancing team capability.
• · Experience with Safety Omics, Cell Painting or imaging data.
• · Background in toxicology, pharmacology or ADMET.
• · Experience with cloud computing or workflow automation.
• · Track record of publications in top AI conferences or journals in pharmaceutical research (e.g., NeurIPS, ICML, Nature Machine Intelligence, Nature Communications, NEJM AI, etc.)
• · Experience supervising scientists or managing collaborations.

Skills

Required

• Machine learning/AI
• Python
• GitHub
• CI/CD
• DevOps
• MLOps
• GNNs
• D-MPNN
• Foundation models (e.g., CheMeleon)
• Multimodal data handling
• Leadership
• Communication

Preferred

• Safety Omics
• Cell Painting
• Imaging data
• Toxicology
• Pharmacology
• ADMET
• Cloud computing
• Workflow automation