Research Advances from the Human Tumor Atlas Network (HTAN)
Human tumor atlases, data, and discoveries from HTAN are revealing insights into cancer evolution and enabling the development of predictive models for cancer patients.
Human Tumor Atlases
Strand et al. with the Duke University HTAN Center generated a spatially resolved atlas of breast precancers, which revealed biomarkers of clinical outcome and recurrence in ductal carcinoma in situ (DCIS). This resource and related data provide insights into the disease states of precancerous breast lesions.
Heiser et al. with the Vanderbilt University HTAN Center created a spatial-omic atlas of colorectal cancer, which revealed how these types of tumors evade the immune system. This resource may also inform patient stratification and the development of new treatments.
Nirmal et al. with the Harvard University HTAN Center developed spatial maps of melanoma, which revealed cellular
interactions and mechanisms of immunosuppression that promote cancer progression. Along with the study, they also released the largest imaging-based melanoma dataset at that time to support investigations of skin cancer prevention and treatment.
Sussman et al. with the Children’s Hospital of Philadelphia HTAN Center generated a longitudinal single-cell and spatial multi-omic atlas of pediatric high-grade glioma, which shows mechanisms of therapeutic response and reveals potential therapeutic strategies for pediatric brain tumors.
Johnson et al. with the OHSU HTAN Center developed an omic and multidimensional spatial atlas of an evolving metastatic breast cancer. This resource also shows potential mechanisms of response and resistance to treatment.
New Computational Tools and Data Standards
HTAN investigators developed a Minimum Information about highly multiplexed Tissue Imaging (MITI) standard for highly multiplexed tissue images and traditional histology.
The computational pipelines used for analyzing the Human Tumor Atlas Pilot Project (HTAPP) data are available to the community. Some of these workflows were used in the generation of a single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors.
Kunes et al. with the MSKCC HTAN Center developed Spectra, an algorithm for the supervised discovery of interpretable gene programs from single-cell data. This open-source software scales to large tumor atlases for determining molecular factors linked to clinical outcomes.