2026 Summer Intern - Digital and Spatial Pathology, Research Pathology

Roche • South San Francisco, CA

3d • $50 • Hybrid

About The Position

The Research Pathology Department is embedded within Genentech’s Research and Early Development Organization (gRED) and works to ensure that strategies for the treatment and cure of disease are based on accurate analyses of pathogenetic mechanisms. The department is a key driver in Genentech’s Digital and Spatial Pathology efforts, developing cutting-edge tissue technologies to support scientific discovery. This internship is within the Risom lab of the Spatial Omics (SO) group, which specializes in high-dimensional spatial proteomics assay and analytical tool development. This position is located in South San Francisco, on-site / hybrid (minimum 3 days a week on-site). The intern will investigate and deploy computational methods aimed at systematically profiling key histological structures in tissue specimens of the normal lung, lung cancer, and chronic inflammatory disease. The project will focus on leveraging high-dimensional spatial proteomics information (e.g. Lunaphore COMET, Phenocycler, MIBIscope, IMC) to train an H&E inference model to segment important tissue features.

Requirements

• You meet one of the following criteria: Must be pursuing a Master's Degree (enrolled student). Must have attained a Master's Degree. Must be pursuing a PhD (enrolled student).
• Required Majors: Computer Vision, Computational Biology, Artificial Intelligence / Machine Learning, Bioinformatics, Mathematics, Statistics, Physics, Engineering, Immunology or other related quantitative/scientific fields.
• Experience with training, validating, and refining image-based AI/ML models.
• Proficiency in Python programming.
• Strong problem-solving skills and critical thinking abilities.

Nice To Haves

• Excellent communication, collaboration, and interpersonal skills.
• Core Values: Complements our culture and the standards that guide our daily behavior & decisions: Integrity, Courage, and Passion.
• Technical Knowledge: Familiarity with computational pathology, tissue-based spatial technologies (e.g. spatial transcriptomics, spatial proteomics), and foundational digital pathology models.

Responsibilities

• Developing and evaluating computational approaches to classify tissue features in the spatial proteomics data through alternative approaches (ex. nnUNET vs ChannelNET).
• Deploying H&E vision transformer models to generate feature vectors of high resolution H&E patches (e.g. DINO-ViT, OmiCLIP).
• Evaluating approaches for multi-domain vector alignment and annotation of co-embedded clusters.
• Executing H&E inference model training.
• Benchmarking performance of alternative approaches using ground truth tissue feature datasets.
• Collaborating with upstream bench scientists to identify optimal experimental and computational strategies for improved dataset generation.
• Providing regular updates and technical reports to project stakeholders.

What This Job Offers

• Job Type: Full-time
• Career Level: Intern