Senior Machine Learning Scientist

(Senior) Scientist, Machine Learning About Quotient Therapeutics Quotient Therapeutics is a Flagship Pioneering company that uses somatic genomics to discover drug targets. We use high-accuracy sequencing to find rare somatic mutations in cells from non-cancerous diseased human tissue. Mutations that help cells survive or expand in a disease environment point to causal biology. We use that evidence to nominate and assess targets, and to decide whether a target should be copied (protective mutations) or opposed (disease-driving mutations). The Role Quotient is seeking a (Senior) Scientist, Machine Learning to build models and AI agents that turn our somatic mutation data into evidence for target discovery. A central question is what a given mutation does to a protein, and whether that effect explains why cells carrying it are selected in disease. You will build models of variant effects on protein function and connect them to genomic, single-cell, perturbation and phenotype data. You will also develop agentic AI workflows that help scientists analyse data, test hypotheses and make decisions. This is a hands-on role on a cross-functional team. You will write clear, tested code, use modern AI tools to work faster, and work closely with the computational scientists, experimental biologists, engineers and target discovery leads who act on your results. The role is based in Cambridge, UK, and is hybrid, with a minimum of 3 days per week in the office. Key Responsibilities Build and evaluate models that predict how coding and non-coding variants affect protein function, stability, interactions and pathway activity, using protein language models, structure-based methods and related approaches. Use variant-effect predictions to interpret somatic selection signals, including whether mutations are likely loss-of-function, gain-of-function or neutral, and where they cluster in protein structure. Design, build and evaluate AI agents and agentic workflows that retrieve evidence, run analyses and support hypothesis testing and target assessment. Train, fine-tune or adapt foundation models, language models and representation-learning methods for target discovery. Use AI coding tools to speed up development while keeping code readable, tested and reproducible. Work as a core member of cross-functional project teams, share ownership of team goals, and contribute to shared codebases, reviews and tools. Communicate results clearly to scientists and target discovery leadership, including model assumptions, limitations and next steps. Qualifications PhD in machine learning, computer science, computational biology, genomics, bioinformatics, statistics, engineering or a related field, or an MSc with equivalent experience. Level (Scientist or Senior Scientist) will be set based on experience. Strong Python skills and experience building practical deep learning systems (e.g. PyTorch, JAX). Experience training, evaluating or applying ML models to complex real-world biological data. Understanding of how genetic variants affect protein structure and function, and experience with variant-effect prediction methods (e.g. protein language models such as ESM, structure-based models, or tools such as AlphaMissense). Regular use of AI coding assistants, LLM-based workflows or agentic development tools in your own work. Depth in one or more of the following: biomedical foundation models, AI agents, Perturb-seq or single-cell genomics, genotype–phenotype modelling, causal inference, perturbation modelling, or large-scale biological datasets. Proven ability to work in cross-functional teams and to explain technical results to non-specialists. Evidence of technical depth through papers, thesis work, open-source projects, industry projects or substantial applied research. Values and Behaviours Encourage respectful disagreement and cultivate open-minded, ego-free interactions to continuously pu