Postdoctoral Scientist – Multimodal Representation Learning Predictive Biology

Postdoctoral Scientist – Multimodal Representation Learning for Predictive Biology

Johnson & Johnson Innovative Medicine is recruiting for a Postdoctoral Scientist – Multimodal Representation Learning for Predictive Biology to join the Data, Data Science & Artificial Intelligence (DDSAI) organization for a two-year fixed term position, helping advance AI/ML analytics and multimodal modeling for drug discovery. This position will be based at any of the following locations:
Cambridge, MA (preferred);
Spring House, PA;
Beerse, Belgium; or Madrid, Spain. (No fully remote option.)

Within DDSAI, our teams develop innovative solutions using a variety of data sources across multiple therapeutic areas. We are looking for a highly motivated and innovative Postdoctoral Scientist to work at the intersection of advanced AI/ML modeling at scale, multimodal representation learning, drug discovery and predictive biology. The successful candidate is self-motivated, creative, and an effective communicator, with a strong interest in deep learning for imaging microscopy and multi-omics.

The role focuses on developing advanced models that analyze and quantify the heterogeneity of perturbed cellular systems and integrate multi-scale biological data (e.g., high-content imaging, phenomics, transcriptomics, and proteomics) to derive new biological insights that support the next-generation portfolio for drug discovery. A successful candidate should demonstrate a strong capacity to build, evaluate, and benchmark AI/ML methods, and publish findings in top-tier peer-reviewed publications.

• Conceive, design, develop, implement and validate innovative AI/ML solutions for drug discovery problems using multimodal data of perturbed cells.
• Analyze and extract novel biological insights from large-scale, heterogeneous high-dimensional data (e.g., imaging microscopy, phenomics, transcriptomics, proteomics).
• Develop and evaluate innovative computer vision solutions to quantify heterogeneity of perturbed cells.
• Collaborate closely with both internal and external cross-functional teams of scientists and engineers to advance algorithms and product development to improve project outcomes.
• Clearly communicate complex technical methodologies and present findings to diverse audiences and stakeholders to support informed decision-making.
• Follow standard processes for documentation and maintaining an up-to-date code repository.
• Draft manuscripts and disseminate research findings internally and externally (e.g., publishing in peer-reviewed conferences/journals).

• Ph.D.
  * in Electrical Engineering, Biomedical Engineering, Computer Science, Applied Mathematics, or a related major. (
  * PhD completed within the last 3 years or to be completed within the next 6 months.)
• Strong technical expertise in AI/ML for biological applications, with excellent analytical skills in at least one domain area (representation learning, multimodal integration, computational biology, imaging microscopy) and a proven track record (publications in top-tier AI/ML conferences or journals).
• Experience developing foundation models and/or multimodal modeling of biological data in related applications, such as Cell Painting, imaging microscopy, and multi-omics.
• Solid working knowledge of deep-learning architectures and techniques (e.g., Transformers, CNNs, graph networks, self-supervised and multi-instance learning, etc.) and their application to multimodal representation learning, alignment, and fusion.
• Proficiency with one or more programming languages, preferably Python, and proficiency in one or more AI frameworks such as PyTorch or Tensor Flow.
• Strong communication skills, with the ability to present analytical results on behalf of the project team in a multi-functional setting, contextualizing how the work fits within the broader project.