A collection of 1,276 spatial transcriptomic profiles, each linked and aligned to a Whole Slide Image (with pixel size < 1.15 µm/px) and metadata.
HEST-1k was assembled from 180 public and internal cohorts encompassing:
26 organs
2 species (Homo Sapiens and Mus Musculus)
398 cancer samples from 25 cancer types.
HEST-1k processing enabled the identification of >1.5 million expression/morphology pairs and >76 million nuclei… See the full description on the dataset page: https://huggingface.co/datasets/MahmoodLab/hest.