Protein Design Scientist, Machine Learning

As a Protein Design Scientist, you leverage an AI-first approach, utilizing protein language models (pLMs) and generative sequence design to explore sequence-function relationships and pioneer next-generation agricultural traits. As an Applied ML Scientist, you are a hypothesis-driven scientist who leverages and adapts open-source machine learning models to biological data, addressing complex biological questions where data may be sparse and expensive to generate. You are also a collaborative team player who thrives in the dry-to-wet lab loop by turning agricultural and trait challenges into practical machine learning hypotheses and projects, while translating complex ML concepts and outputs into clear, practical suggestions for diverse stakeholders. Position will be located at Durham, North Carolina with an opportunity for remote work.
Accountabilities:
Design & Optimize: Formulate biological hypotheses and design computational workflows for large scale variant design and property prediction to accelerate trait discovery
Deploy ML Models: Implement, adapt, and tune state-of-the-art biomolecular ML models—including single-sequence LMs, generative models, co-evolutionary aware architectures, and 3D structural prediction models—to drive innovative projects for the trait pipeline
Collaborate Cross-Functionally: Partner closely with wet-lab research teams to design variant libraries, leveraging active learning and Bayesian optimization to iteratively integrate experimental screening data into design loops
Communicate Insights: Communicate complex deep learning concepts, protocols, and project progress clearly to technical and non-technical stakeholders
Innovate: Monitor the rapidly changing protein design literature and bring promising new tools and project ideas to the team
PhD with +1-year experience in bioinformatics, computational biology, biochemistry & biophysics, or a related field with a focus on protein sequence or structure modeling
ML and pLM Experience: Demonstrated hands-on experience developing, adapting, or fine-tuning machine learning models for protein sequence design, variant library generation, protein property prediction, or related biomolecular engineering tasks
Protein Science Expertise: Deep understanding of protein sequence-structure-function relationships
Scientific Problem Solving: Proven ability to transform complex biological challenges into actionable, ML/DL computational hypotheses to support the trait pipeline
Collaborative Communication: Strong communication skills with a track record of working effectively alongside experimental/wet-lab scientists
ML & Deep Learning Engineering: Experience working with Python and deep learning libraries (like PyTorch or JAX) to adapt or fine-tune open-source models
Desired Qualifications:
Structural Integration and biophysics: Experience integrating protein structural information, including structure prediction models (AlphaFold, Boltz), structure-aware design methods (ProteinMPNN, GNNs, structure-conditioned pLMs, Foldseek 3Di), or physics-based approaches (Rosetta, molecular dynamics) to augment sequence-based design and prediction workflows.
Wet-Lab History: Strong familiarity of or experience in wet-lab workflows—either in library design or generating screening data for model training or validating designed variants (e.g., protein characterization)—to facilitate seamless communication with experimental partners
Compute and data: Familiarity querying large-scale biological datasets, working in cloud environments (AWS, HPC), and utilizing containerized workflows (Docker, Singularity, Nextflow)
Industry Experience: Prior experience in agricultural biotechnology, plant biology, or a strong interest in translating computational protein design to crop science

Employment type
Full-time

Work arrangement
No