Scientist, Computational Chemistry

Led by Michael Antonov, a co-founder of Oculus, and well-funded by Formic Ventures, Deep Origin is poised to reinvent the way scientists work and life science innovations come to life. We see a future largely free of diseases, with a 150-year lifespan being the norm. To get there, we are building an operating system for science, enabling scientists to be more productive and to bring tomorrow's ideas to life quickly and at a reasonable cost.

Deep Origin is seeking a Scientist with strong expertise in small-molecule docking and benchmarking, molecular dynamics (MD) simulations, and free energy perturbation (FEP), machine learning, to support a transformative ARPA-H initiative. You'll lead the design of robust simulation workflows and analyze protein‑ligand structures across a large target panel to support predictive modeling for therapeutic discovery.

Applicants must be authorized to work for any employer in the U.S. We are unable to sponsor or take over sponsorship of an employment Visa at this time.

• Ph.D. in computational chemistry, structural biology, biophysics, or related field
• 2+ years of postdoctoral or industry experience in structure‑based modeling
• Hands‑on expertise with FEP (RBFE/ABFE), including best practices around setup, sampling, and analysis
• Proficiency with one or more simulation platforms (e.g., OpenFE, GROMACS, AMBER, NAMD)
• Hands‑on experience with RDKit and related cheminformatics tools, and with machine learning methods (RF, gradient boosting, SVM, linear models, Chemprop) for molecular property modeling
• Strong understanding of protein‑ligand binding, structure selection, and conformational variability
• Programming experience in Python, and familiarity with tools like MD Analysis, PyMOL APIs, or MDTraj

• Analyze tens to hundreds of protein targets relevant to ADMET and off‑targets, focusing on conformations, binding site flexibility, and ligand‑bound states to guide structure preparation and ensemble design
• Run and refine small‑molecule docking, MD, and FEP (RBFE and ABFE) simulations using state‑of‑the‑art tools
• Apply alchemical transformations and advanced sampling strategies to build robust, well‑converged, and reproducible FEP workflows for accurate binding free energy predictions
• Apply cheminformatics tools (e.g., RDKit, scikit‑learn, etc.) for molecular representation and descriptor generation, and with machine learning methods, including random forests, gradient‑boosted trees, SVM, linear/regularized regression, and Chemprop, for molecular property prediction and model validation
• Collaborate with ML and experimental teams to integrate structure‑based insights across discovery pipelines
• Communicate progress, technical findings, and challenges across internal and external teams
• Stay current with advances in structure‑based binding affinity prediction methods and best practices, and integrate relevant developments into ongoing work

• Experience benchmarking across multiple PDB entries or conformational states
• Prior work integrating structural modeling into machine learning pipelines
• Familiarity with MM/GBSA, docking scoring functions, or clustering methods
• Experience using Unix‑based HPC environments, workload managers (e.g., SLURM, etc.), and optionally AWS
• Comfort managing large‑scale simulation data for modeling or analysis

Deep Origin builds modern infrastructure for computational science at the interface of biology, chemistry, and AI. As part of our ARPA‑H program, you'll shape the future of structure‑based modeling for therapeutics.