Computational Biologist, Spatial Transcriptomics

Origin; San Francisco (CA); Hybrid; USD 120,000 - 170,000; Full time

Job summary

Origin Bio seeks a computational biologist to own post-run QC and secondary analysis of Xenium data, from transcript-level outputs to datasets aligned with post-run H&E images.

You will develop reproducible QC workflows, align histology images, and work with pathologists to annotate tissue regions, while exploring cell states, spatial neighborhoods, and tumour-immune interactions across samples.

Qualifications

• PhD or postdoctoral research in spatial biology or related field.
• Experience with Xenium or CosMx; other spatial platforms welcome.
• Experience analyzing oncology tissue, H&E slides, or multiple imaging modalities.
• Strong Python or R skills and reproducible analysis workflows.
• Experience with scRNA-seq or bulk RNA-seq, with reference data for spatial interpretation.
• Good statistical judgement: noise, batch effects, artefacts, confounding factors.

Responsibilities

• Develop reproducible QC workflows for transcript detection, background signal, cell segmentation, transcript assignment, tissue and imaging artefacts, and sample-to-sample variation.
• Align post-run H&E whole-slide images to Xenium DAPI images and assess registration accuracy across tissue sections.
• Collaborate with pathologists to bring tumour regions and histological annotations into the spatial data.
• Analyze data beyond cell typing: cell states, spatial neighbourhoods, tumour-immune interactions, differential expression, and other patterns.
• Identify meaningful questions and derive reference results that test AI model capabilities from the data.

Skills

Python

R

Reproducible workflows

Single-cell RNA-seq

Bulk RNA-seq

Statistics

Spatial biology

Education

PhD in spatial biology or related field

Tools

Xenium

CosMx

Spatial platforms

Job description

Origin Bio is producing matched spatial transcriptomics, histopathology images and clinical datasets from real patient tumours. We’re looking for a computational biologist to own post-run QC and secondary analysis of our Xenium data, from transcript-level and cell-feature outputs through to datasets aligned with post-run H&E images.

What you’ll work on

• Develop reproducible QC workflows for transcript detection, background signal, cell segmentation, transcript assignment, tissue and imaging artefacts, and variation across samples and runs.
• Align post-run H&E whole-slide images to Xenium DAPI images and assess registration accuracy across each tissue section.
• Work with pathologists to bring tumour regions and other histological annotations into the spatial data.
• Analyse the data beyond cell typing: investigate cell states, spatial neighbourhoods, tumour-immune interactions, differential expression and other patterns that emerge from the tissue. Use relevant single-cell reference datasets where helpful.
• Identify meaningful questions that require analysis of the data to answer. For selected questions, derive well-supported reference results and turn the work into reproducible tasks that evaluate whether AI models can reach those results from the underlying data.

Helpful past experience

• PhD or postdoctoral research in spatial biology or equivalent hands-on industry experience. Xenium or CosMx experience is especially valuable; experience with other spatial platforms is welcome.
• Analysis of oncology tissue, H&E whole-slide images, or multiple imaging modalities such as IHC and immunofluorescence.
• Strong Python or R skills and experience building reproducible analysis workflows.
• Experience with scRNA-seq or bulk RNA-seq analysis, including the use of reference data to interpret spatial measurements.
• Sound statistical judgement: recognising noise, batch effects, experimental artefacts and confounding variables, and knowing when an apparent biological finding needs further validation.

This is a full-time role. We prefer someone who can work with us in person in San Francisco, but we’re open to the right person working remotely.