About The Position

The Director, Predictive and Agentic Target Biology is a hands-on technical individual-contributor role in GSK’s Discovery Data Sciences (DDS), within the Data, Automation, and Predictive Sciences function. The role applies machine learning and agentic AI across discovery data to advance target discovery from analysis toward prediction and to help build a Target Discovery Engine that surfaces, evaluates, and prioritises drug targets. The successful candidate will personally design, build, and ship agentic workflows, and work with scientific domain specialists and engineering teams to research, embed, and operationalise predictive and agentic methods. The work spans omics, functional genomics, chemogenomics, and quantitative pharmaceutical-science data. For candidates based at Stevenage, the position will temporarily be located there, with a planned office relocation to Cambridge, UK, proposed to take effect by early 2029.

Requirements

  • Master’s degree or higher in Computer Science, Computational Biology, Bioinformatics, or a related field.
  • 15 or more years of software engineering and applied computational science experience in bioinformatics, cheminformatics, or scientific computing.
  • Experience with pharmaceutical, biomedical, or life-sciences preclinical discovery data.
  • Experience applying machine learning, graph-based methods, or AI-agent methods to biological or chemical data.
  • Programming experience in Python and/or R, including machine-learning libraries such as scikit-learn, PyTorch, or scverse.
  • Experience with production AI/ML systems and MLOps practices, including CI/CD, containerisation, and cloud deployment.

Nice To Haves

  • Experience fine-tuning or applying biological foundation models to target-biology questions.
  • Experience with virtual-cell, in-silico perturbation, or perturbation-prediction methods, including understanding how to validate their reliability.
  • Experience building or deploying agentic workflows, including multi-agent orchestration, RAG, MCP-based tool access, or foundation-model integration.
  • Ability to match methods to scientific questions and assess whether predictions are credible enough to act on.
  • Knowledge of knowledge-graph or graph-learning methods for integrating heterogeneous evidence.
  • Experience productionising ML models, monitoring models, automated retraining, and hands-on MLOps, including versioning, experiment tracking, and containerisation.
  • Knowledge of deep-learning architectures for biological data, such as transformers, diffusion models, or GNNs, and data-engineering concepts such as ETL, databases, and data lakes.
  • Strong communication skills and ability to translate between biological and technical audiences.

Responsibilities

  • Adapt pre-trained biological foundation models to target-biology questions through fine-tuning, probing, and application, including questions about target-disease mechanisms.
  • Apply and evaluate virtual-cell and in-silico perturbation approaches to predict cellular responses and support mechanistic, in-silico target discovery.
  • Integrate heterogeneous target evidence—including omics, functional genomics, causal biology, chemogenomics, literature, and clinical evidence—into structured, traceable representations using knowledge-graph and graph-learning methods.
  • Design and build agentic scientific workflows that plan, retrieve, analyse, and synthesise information across target discovery, using approaches such as multi-agent systems, RAG, and MCP-based tool access, with grounding, provenance, and rigorous reliability evaluation.
  • Evaluate emerging AI/ML approaches for discovery data science, benchmark performance using appropriate validation strategies, and implement approaches for reliable production use.
  • Apply biological and chemical domain knowledge to ensure models are scientifically grounded, interpretable, and relevant to therapeutic questions; communicate evaluation and validation strategies and uncertainty.
  • Partner with scientific domain experts to ensure methods address portfolio-relevant questions and advocate for opportunities offered by predictive and agentic technologies.
  • Contribute with engineering teams to MLOps practices for reliable deployment, monitoring, and drift detection of predictive methods.

Benefits

  • Comprehensive benefits program for US employees; details are available in the GSK US Benefits Summary.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service