Machine Learning Scientist/Sr Scientist - Small Molecule Property Prediction and Generative Design

Role Summary

Machine Learning Scientist/Sr Scientist - Small Molecule Property Prediction and Generative Design plays an essential leadership role within the TuneLab platform, specializing in small molecule drug discovery. This role requires deep expertise in medicinal chemistry, ADME/Tox prediction, and small molecule optimization, combined with advanced data science capabilities in generative modeling and property prediction. You will help develop predictive and generative models that accelerate lead optimization and candidate selection across the TuneLab federated network.

Responsibilities

• Small Molecule Property Prediction: Architect and implement advanced multi-task learning models for small molecule properties including ADMET endpoints, solubility, permeability, metabolic stability, and off-target liabilities, handling SMILES, graphs, and 3D conformations.
• Generative Chemistry Models: Design and deploy state-of-the-art generative models for de novo small molecule design, lead optimization, and scaffold hopping with synthetic accessibility and drug-likeness constraints.
• ADMET-Driven Design: Develop integrated prediction-generation pipelines that optimize molecules across multiple ADMET properties while maintaining potency, using multi-objective optimization and Pareto front exploration.
• Chemical Space Navigation: Implement algorithms for exploring synthetically accessible chemical space, including reaction-aware generation, retrosynthetic planning, and fragment-based design.
• Structure-Activity Learning: Build models that learn structure-activity relationships from sparse, noisy bioactivity data across federated partners, including matched molecular pair analysis and activity cliff prediction.
• Molecular Representation Learning: Develop self-supervised and semi-supervised methods to learn robust molecular representations from unlabeled compounds for better generalization.
• Lead Optimization Workflows: Create AI-driven workflows for medicinal chemistry tasks including bioisosteric replacement, metabolic site prediction, toxicophore removal, and property optimization while protecting IP.
• Synthetic Feasibility Integration: Collaborate with synthetic chemists to ensure generated molecules are synthesizable, incorporating reaction prediction and building block availability into generation.
• Cross-Partner Chemical Diversity: Leverage chemical diversity across federated partners while respecting competitive boundaries, identifying complementary regions of chemical space for collaboration.
• Small Molecule Benchmarking: Establish benchmarks for small molecule property prediction and generation using public datasets and Lilly data.

Qualifications

• PhD in Computational Chemistry, Cheminformatics, Medicinal Chemistry, Chemical Engineering, or related field.
• Minimum of 2 years of experience in small molecule drug discovery.
• Strong experience with molecular property prediction and QSAR/QSPR methods.
• Deep understanding of medicinal chemistry principles and ADMET optimization.

Additional Preferences

• Experience with federated learning and distributed optimization in chemical applications
• Publications on molecular generation or property prediction
• Expertise in graph neural networks and geometric deep learning for molecules
• Background in organic chemistry and synthetic feasibility assessment
• Experience with fragment-based and structure-based drug design
• Knowledge of PK/PD modeling and clinical translation
• Proven track record in developing generative models for molecular design
• Proficiency in cheminformatics tools (RDKit, DeepChem)
• Understanding of IP considerations in generative molecular design
• Experience with active learning and design-make-test-analyze cycles
• Portfolio mindset ensuring decisions align with TuneLab ecosystem goals

Education

• PhD in a relevant field (as listed under Basic Qualifications)