Machine Learning Scientist, Biomolecular Structure Prediction

Build and improve all-atom co-folding models for biomolecular structure prediction, focusing on antibody-antigen interfaces and protein-nucleic acid assemblies that current models handle poorly. The scientist will diagnose failure modes such as CDR-H3 and epitope placement, interface geometry, nucleic acid conformations and confidence calibration, curate structural and experimental training data, and integrate structure models with generative protein models for antibody design and genome editing.

PhD (or equivalent industry experience) in computer science, computational biology, biophysics or a related field, with hands-on experience training AlphaFold- or Protenix-style structure prediction models. Publications at major ML conferences or science journals, or widely used open-source models, and PyTorch or JAX multi-GPU training experience. Antibody docking, protein-DNA/RNA modelling, structural data pipelines and diffusion or flow-matching models are preferred.