ML Expert / Senior Scientist, II
AbbVie
5 hours ago
Remote friendly (North Chicago, IL)
United States
IT
Purpose
- Leverage AI-enabled drug discovery across the small molecule portfolio to identify, develop, optimize, validate, and deploy machine learning approaches that guide molecular optimization.
Responsibilities
- Consult with project teams to establish quantitative optimization targets.
- Identify models to predict small molecule binding affinities and pharmacological properties (e.g., PK).
- Build and benchmark predictive models for each project endpoint; evaluate performance and domain of applicability.
- Select the best model for each query molecule using data-driven approaches that account for model uncertainty and heteroscedasticity.
- Communicate model selection rationale to project teams, explaining why chosen models fit the task.
- Deploy models to ensure project-team accessibility and support collaborative ideation and optimization.
Qualifications
- Bachelor’s (12+ years), Master’s (10+ years), or PhD (4+ years).
- Expertise developing/implementing/deploying ML/deep learning and cheminformatics computational solutions.
- Expertise in AI/ML for molecular generation, pose prediction, affinity prediction, and/or pharmacological property prediction (e.g., PK).
- Strong Python programming skills; experience with numpy, pandas, scikit-learn, and related scientific libraries.
- Ability to implement, debug, and maintain computational tools; proficiency with cloud computing.
- Familiarity with deep learning architectures (GNN, CNN, RNN, Transformer, GCNN, MPNN) and paradigms (generative models, GAN, active learning).
- Strong analytical/problem-solving skills; ability to work independently and collaboratively.
- Clear communication skills, including explaining complex ideas to non-specialists.
- Leverage AI-enabled drug discovery across the small molecule portfolio to identify, develop, optimize, validate, and deploy machine learning approaches that guide molecular optimization.
Responsibilities
- Consult with project teams to establish quantitative optimization targets.
- Identify models to predict small molecule binding affinities and pharmacological properties (e.g., PK).
- Build and benchmark predictive models for each project endpoint; evaluate performance and domain of applicability.
- Select the best model for each query molecule using data-driven approaches that account for model uncertainty and heteroscedasticity.
- Communicate model selection rationale to project teams, explaining why chosen models fit the task.
- Deploy models to ensure project-team accessibility and support collaborative ideation and optimization.
Qualifications
- Bachelor’s (12+ years), Master’s (10+ years), or PhD (4+ years).
- Expertise developing/implementing/deploying ML/deep learning and cheminformatics computational solutions.
- Expertise in AI/ML for molecular generation, pose prediction, affinity prediction, and/or pharmacological property prediction (e.g., PK).
- Strong Python programming skills; experience with numpy, pandas, scikit-learn, and related scientific libraries.
- Ability to implement, debug, and maintain computational tools; proficiency with cloud computing.
- Familiarity with deep learning architectures (GNN, CNN, RNN, Transformer, GCNN, MPNN) and paradigms (generative models, GAN, active learning).
- Strong analytical/problem-solving skills; ability to work independently and collaboratively.
- Clear communication skills, including explaining complex ideas to non-specialists.