Computational Chemist, Machine Learning, Pfizer Inc. (18 February 2022)

Position Description

Pfizer’s Machine Learning and Computational Sciences group is looking for a computational chemist experienced in machine learning to develop predictive and generative models for ligand-based and structure-based small molecule drug design. This effort will leverage large datasets (including structure-activity relationship databases, and X-ray/cryo-EM structures) that are exclusively available at Pfizer. The successful candidate will explore physics-inspired deep learning approaches that can learn from these datasets and generate and accurately score new designs.

This role will collaborate closely with the broader Medicinal Sciences organization at Pfizer to prospectively test developed models on active discovery campaigns. This role will also collaborate with Pfizer’s hub for machine learning, and cheminformatics, MLOps, and HPC teams to train and deploy models. The incumbent is expected to join academic collaborations, publish in reputed journals, and present at high-profile conferences. Our group applies machine learning methods to multiple therapeutic modalities, so while the role initially focuses on small molecules, the incumbent will have the opportunity to branch out into vaccines, mRNA-based therapies, and biologics.

Required Qualifications

  • Formal training in Computational Chemistry, Computational Materials Science, Chemical Engineering, or a related technical discipline
  • Familiarity with small molecule drug discovery, and concepts such as receptors, ligands, binding affinity, and potency
  • Expertise in Python, PyTorch or TensorFlow, NumPy, scikit-learn and other elements of the scientific programming stack
  • Experience running code on HPC environments and schedulers
  • Published research on the application of deep learning methods to chemistry or materials science

Preferred Qualifications

  • MS/PhD in Computational Chemistry, Computational Materials Science, Chemical Engineering, or a related technical discipline
  • Expertise in rolling your own CNNs, MPNNs/GCNNs, and attention-based architectures
  • Familiarity with structure-based drug discovery, docking, scoring, quantum chemistry, and molecular dynamics
  • Experience with PyTorch Geometric or DGL
  • Experience collaborating on research codebases

Instructions on how to apply can be found HERE.