SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Lila Sciences is seeking a Machine Learning Scientist to develop next-generation cofolding models for drug discovery. This role focuses on training and improving models that reason over proteins, ligands, binding context, and experimental data using contrastive learning and representation-learning approaches.
You will train and evaluate cofolding models for protein-ligand and molecular discovery applications, designing training objectives that connect ligands, proteins, structures, assays, simulations, and experimental data. Key responsibilities include developing modeling approaches that leverage DEL (display element library) data for learning binding, enrichment, selectivity, and structure-activity signals. You'll build and evaluate models informed by Boltz, AlphaFold-style cofolding, equivariant GNNs, and related structure-aware ML methods.
You will design rigorous evaluation frameworks that distinguish meaningful molecular learning from dataset artifacts, leakage, or spurious correlations. Collaboration is central: you'll work with ML researchers, computational chemists, computational biophysicists, data engineers, and drug discovery teams. You'll partner with computational chemistry and biophysics teams to connect model outputs to physically and chemically meaningful hypotheses, and help expose trained models as tools for scientists and AI agents.
Required qualifications include a PhD or equivalent in machine learning, computational biology, computational chemistry, bioinformatics, or computer science. You need hands-on experience training deep learning models for molecular, protein, or structural biology applications, with expertise in contrastive learning, representation learning, self-supervised learning, or multimodal learning. Practical experience with PyTorch, JAX, or equivalent ML frameworks is essential, along with strong understanding of data quality, leakage risks, and benchmark design.
Bonus experience includes hands-on DEL data work, drug discovery background, familiarity with Boltz/AlphaFold/equivariant GNNs, distributed model training, active learning, or integration of ML into agentic scientific workflows.
