Postdoctoral Fellow, Ai/ml Applications for Vaccine

Pfizer Pfizer · Pharma · New York, NY

This role focuses on developing and validating AI/ML models for vaccine strain selection, specifically using deep learning on viral sequences and integrating various biological data sources. The goal is to predict and evaluate prospective vaccine strains.

What you'd actually do

  1. Develop sequence-based deep learning models for rapidly evolving virus, including: transformer or language-model-based architectures for viral protein sequences and graph neural networks that predict time-dependent changes in strain dominance.
  2. Integrate multi-source surveillance, immunogenicity, and vaccine efficacy data to compute and evaluate prospective coverage scores for candidate vaccine strains.
  3. Utilize interpretation frameworks to identify key features for virus evolutional advantage related to infectious disease burden and vaccine antigen design.
  4. Conduct rigorous retrospective and prospective benchmarking validation. iterative fine-tuning to improve model performance
  5. Communicate complex data and results clearly to both technical and non-technical stakeholders. Collaborate extensively with those from other scientific disciplines within the group, from other subdivisions of Pfizer, and potentially from external partners.

Skills

Required

  • Ph.D. in Computational Biology, Bioinformatics, Computer Science, Machine Learning, or a closely related field.
  • Demonstrated ability to independently design and implement complex ML models, evidenced by first‑author publications or equivalent open-source research contributions.
  • Strong hands-on experience with deep learning for sequence data, including transformer or language-model architectures, and model training, validation, and benchmarking on large biological datasets
  • Proficiency in Python and modern ML frameworks (e.g., PyTorch, TensorFlow, scikit-learn), with experience managing full modelling pipelines and statistical modeling, including regression analysis and mixed-effects models.
  • Experience working with viral or microbial sequence data, including alignment, curation, and longitudinal analysis across time.
  • Less than 2 years of post-degree experience.
  • Two letters of recommendation must be provided prior to interview.
  • Willingness to make a minimum 2-year commitment.
  • Strong communication and collaboration skills with the ability to work effectively in a hybrid team environment.
  • Strong organizational skills and attention to detail in managing deadlines and documentation.
  • Ability to clearly communicate complex modeling concepts and results to both technical and biological audiences.

Nice to have

  • Direct experience with viral evolution modeling, fitness/dominance prediction, or time-resolved sequence forecasting.
  • Experience building or extending protein language models or MSA-based neural networks for biological inference.
  • Familiarity with antigenicity data or related experimental measurements, and how such data can be integrated into machine learning models
  • Knowledge of SHAP or similar model interpretation frameworks for feature attribution in complex models.
  • Prior work on influenza, SARS‑CoV‑2, or other rapidly evolving viruses, particularly in the context of immune escape, antigenic drift, or vaccine design.

What the JD emphasized

  • Demonstrated ability to independently design and implement complex ML models, evidenced by first‑author publications or equivalent open-source research contributions.
  • Two letters of recommendation must be provided prior to interview.
  • Willingness to make a minimum 2-year commitment.

Other signals

  • AI-driven models for prospective vaccine strain selection
  • Develop sequence-based deep learning models
  • Integrate multi-source surveillance, immunogenicity, and vaccine efficacy data
  • Utilize interpretation frameworks
  • Conduct rigorous retrospective and prospective benchmarking validation