Skip to main content
An official website of the United States government
NIH Websites Are Changing
Per Executive Order 14434, federal agencies are now using "Super Intelligence" (SI) in place of "Artificial Intelligence" (AI). You may see both terms across NIH content during this transition.

Human Cancer Models Initiative (HCMI)

HCMI banner collage of cancer cells, DNA and pills with a male doctor speaking to an older patient in a hospital bed holding spouse's hand

HCMI is an international collaboration between NCI, Cancer Research UK, Wellcome Sanger Institute, and Hubrecht Organoid Technology.

The goal of HCMI is to create up to 1,000 patient-derived next-generation cancer models (NGCMs), such as organoids, conditionally reprogrammed cells, neurospheres, or optimal growth condition models, as a community resource. Additionally, HCMI models are broadly available through  American Type Culture Collection (ATCC). 

Background about HCMI

Addressing a Challenge

Many of the cancer cell lines that are commonly used in cancer research were established decades ago. These cell lines have been useful for in vitro experiments to study cancer biology, biochemistry, and drug targets. However, drawing conclusions about how in vitro observations may relate to clinical biology is challenging because cancer cell lines: 

  • Lack the cellular complexity and architecture of human tumors.
  • Are not associated with clinical information from the patient. 
  • Lack genomic and molecular characterization data and how it relates to the parent tumor.
  • Limited representation of rare cancers and certain human populations. 

HCMI addresses the deficiencies in current models by collecting patients’ clinical data and assessing (as much as possible) the genomes and transcriptomes of the parent tumor, case-matched normal tissue, and the derived next-generation cancer model. 

Tumor Types

The goal of HCMI is to generate models from many human cancer subtypes, including breast, colorectal, glioblastoma, gastroesophageal, lung, melanoma, and pancreas. 

Models from rare adult and pediatric cancers have also been developed by HCMI, including neuroblastoma, osteosarcoma, Wilms tumor, rhabdomyosarcoma, and Ewing sarcoma. 

Model generation from new cancer types is ongoing, including head and neck, kidney, ovarian, and endometrial cancers.

HCMI Searchable Catalog

The HCMI Searchable Catalog is a continuously updated resource for querying available next-generation HCMI models. 

Within the catalog, you can search by patient demographics, tumor, and model elements (e.g., age at diagnosis, sex, treatment information, clinical tumor diagnosis, primary site, and clinical stage, type of model). 

HCMI Data and Computational Tools

  • Genomic Data Commons (GDC) Portal - Shares HCMI data and publications
  • cBioPortal - Shares processed molecular datasets (i.e., mutation calls, copy-number profiles, and clinical annotations) related to HCMI models
  • HCMI Explorer Suite - Enables exploration and analysis of HCMI data (e.g., treatment timelines, model-specific genomic data, and transcriptional similarities between cancer models, paired tumors, and reference datasets)

HCMI Resources

HCMI Publication Guidance

Acknowledgement of HCMI resources in a publication: "The models and the data are from the Human Cancer Models Initiative (HCMI);  dbGaP accession number phs001486."

Contacts for HCMI

For additional information about the HCMI, please contact Dr. Justin Benavidez or Dr. Cindy Kyi. 

Email