(Senior) Machine Learning Scientist

Company Overview:

Abiologics, Inc. is a privately held, early-stage biotechnology company on a mission to make biology better through an innovative therapeutic modality. We are pioneering the development of SynteinsTM, a transformational class of macromolecular medicines unlocked by novel synthetic and computational technologies. Abiologics is actively seeking exceptional molecular scientists who are passionate about driving therapeutic modality innovation in biotechnology.

Abiologics was founded in Flagship Pioneering’s venture creation engine, where companies such as Moderna Therapeutics (NASDAQ: MRNA) and Generate Biomedicines were conceived and created. Since Flagship’s founding in 2000, the firm has originated and fostered the development of more than 100 scientific ventures resulting in more than 500 issued patents and more than 50 clinical trials for novel therapeutic agents.

About the role:

We are hiring a Machine Learning Scientist to train, validate and ship the predictive models that power our molecular design platform. You will build models that score chemical molecules for developability (ADMET), potency, synthesizability and 3D structure, working alongside chemists, biologists and platform engineers.

In your first year, success means two or three models in production use by project teams, each with documented benchmarks, calibrated uncertainty and a clear applicability domain.

What you will do:

• Train and maintain ML models for ADMET and developability endpoints such as solubility, permeability, metabolic stability, toxicity and aggregation risk.
• Build potency and binding models (QSAR, ligand- and structure-based) to rank and prioritize molecules against targets.
• Develop synthesizability models: synthetic accessibility scoring, reaction-outcome prediction and retrosynthesis-informed filters.
• Apply and fine-tune structure-aware models (GNNs, equivariant networks, protein-ligand structure and docking-informed ML).
• Curate and clean assay data, define scaffold- or time-split benchmarks, and quantify uncertainty and applicability domain.
• Package models for the platform with engineers: versioned, reproducible and monitored for drift.
• Present results and limits of each model to chemists and project teams, and incorporate their feedback into retraining.

What you bring:

• PhD in computational chemistry, cheminformatics, machine learning or a related field, or an MS plus equivalent industry experience (about 3-6 years of hands-on modeling).
• Demonstrated experience training and validating ML models on chemical or molecular data, in industry or academia, with shipped or published results.
• Direct experience with at least two of: ADMET/developability, potency/binding, synthesizability/retrosynthesis, protein or ligand structure.
• Fluency with molecular representations (fingerprints, descriptors, graphs, 3D) and tools such as RDKit.
• Strong Python and a deep learning framework (PyTorch preferred); experience with GNNs or transformers on molecules.
• Rigorous evaluation habits: scaffold or temporal splits, baselines, noisy-label handling and uncertainty estimation.
• Clear communication with chemists and biologists.

Nice to have:

• Publications, open-source contributions or competition results in molecular ML.
• Experience with generative design, active learning or Bayesian optimization loops.
• Experience with large-scale training, MLOps and model serving.
• Familiarity with docking, free-energy methods or protein structure prediction tools.
• Experience working with biologics or non-small-molecule modalities.

The salary range for this role is $132,000 - $258,500. Compensation for the role will depend on a number of factors, including a candidate’s qualifications, skills, competencies, and experience. Abiologics, Inc currently offers healthcare coverage, annual incentive program, retirement benefits and a broad range of other benefits. Compensation and benefits information is based on Abiologics, Inc's good faith estimate as of the date of publication and may be modified in the future.