Our data portfolio

Structured, research-derived oncology datasets designed to power biomarker discovery, model development, and translational research.

Connect with us to discuss our datasets 

Scientist looking at computer screen

What we can offer you

Explore cancer research datasets that combine imaging, multi-omics, pathology, and clinical outcomes - designed for researchers, AI teams, and translational scientists. Our datasets are generated in controlled research settings, often with greater depth and consistency than typical hospital data.

All data is ethically consented and made available through transparent, patient-reviewed partnerships. We offer flexible, non-exclusive licensing models tailored to different research and development needs for companies at all stages.

Read how we govern ethical, secure data access

Current dataset portfolio

NameCancer indicationData typeNumber of patientsDataset and cohort descriptionDownload summary
OPTIMAM mammography datasetBreast
  • Mammography images (2D and 3D)
  • Clinical outcomes
  • 740,000+
Mammograms collected from the national breast screening programme, with annotated regions of interest and associated clinical informationClick here

Use potential: 

Suitable for training and validating diagnostic imaging models at national screening scale.

Why it matters: One of the largest annotated breast imaging datasets available. Mirrors the scale, variability, and complexity of real-world screening. Ideal for training, validating, or stress-testing diagnostic imaging models at population scale.

NameCancer indicationData typeNumber of patientsDataset and cohort descriptionDownload summary

 

S:CORT                                                                   

Colorectal
  • Genomics
  • Transcriptomics
  • Digital pathology images
  • Epigenetics
  • Clinical outcomes
  • 1,700+
Mostly late-stage CRC patients recruited to 8 clinical trials testing radiotherapy, chemotherapy and targeted drugsClick here

Use potential:

Ideal for late-stage CRC model development, treatment response prediction, and pathway exploration.

Why it matters: A rare, well-structured multi-modal dataset linking tumour genomics to therapy response across radiotherapy, chemotherapy, and targeted agents. Ideal for biomarker discovery, predictive model development, or trial simulation in advanced colorectal cancer.

NameCancer indicationData typeNumber of patientsDataset and cohort descriptionDownload summary
Digital pathology images from the BEST 2/3 trials                           Oesophageal cancer and pre-cancerous states
  • Digital pathology images
  • 3,300+
Digital pathology images collated from the BEST 2 and 3 trials that looked to test the clinical suitability of the EndoSign clinical sampling technique, contains associated ground truth labels 
Click here

Use potential:

Enables the development of histopathology AI tools and benchmark evaluation.

Why it matters: This dataset includes gold-standard histology images with expert-verified ground truth labels - uncommon in open datasets. Suitable for developing and benchmarking AI models for early detection, dysplasia classification, or novel scoring systems in pre-cancer detection.

NameCancer indicationData typeNumber of patientsDataset and cohort descriptionDownload summary

NSCLC 

  • Real-world clinical data 
  • CT images 
  • 730+ 

Detailed clinical data including treatment outcomes and longitudinal follow-up from patients treated with first line immunotherapy. 

Use potential:

This dataset is ideal for studying real-world outcomes of immunotherapy-treated patients as well as training and validation of prediction models utilising CT images.

Why it matters: This rare dataset has been carefully curated to include high-quality, structured clinical data from a real-world treatment setting and is paired to baseline and longitudinal CT images. The dataset will be routinely updated with existing and new patient data and relevant future NHS-approved treatments. 

NameCancer indicationData typeNumber of patientsDataset and cohort descriptionDownload summary

Oesophageal

  • Clinical
  • Genomics
  • Transcriptomics
  • Methylation array
  • 4,300

Multi-omics data from 500 oesophageal cancer patients, encompassing demographic details, treatment regimens, responses and outcomes.

Use potential:

Enables prognosis prediction and patient stratification, as well as descriptive epidemiological research.
Why is matters: One of the world’s largest and most comprehensive collections of clinically linked, multi-omics data from oesophageal cancer patients.

Data assets coming soon

Cancer indicationData descriptionEstimated availability
Bladder 
  • Multi-omic profiling and images from >1250 patients with longitudinal clinical data
3 months
Breast
  • Imaging Mass Cytometry
  • Associated clinical outcomes (response to immune checkpoint inhibitors treatment)
6 months
Liver (Hepatocellular Carcinoma, HCC)
  • Sample collection database from >1300 patients
6-12 months

 

Interested in a specific dataset?

Contact us for dataset summaries, licensing terms, or to scope a collaboration. Alternatively, if you have specific data needs for your project that aren't listed here, get in touch, and we can discuss opportunities arising from our early-stage pipeline.

Contact us