Welcome to HistoCell
HistoCell is a weakly-supervised framework for accurate super-resolution cell spatial profiles at the pixel-level only from histology images. This website allows users to systematically investigate the hierarchical cellular information, including cell types/states and their spatial network, and discover the spatial and cell biomarkers associated with diverse clinical outcomes, on user provided histology images.
Website features:
· Visualization and explore HistoCell results.
· Run HistoCell to infer the cell spatial profiles based on the user provided histology images.
· Run HistoCell to discover clinically-relevant spatial and/or cell biomarkers based on user provided data.
· Download data associated with the HistoCell publication.
Note:
The website does not allow users to infer new cell types and cell states with their own data. New cell types and cell states inference model should be pre-trained based on
spatial transcriptomics data and single-cell data. These functionalities are provided in the HistoCell source code, which will be available at https://github.com/recolyce/HistoCell after publication.
Cell spatial profile
inference
inference
Tissue architecture
annotation
annotation
Clinical indicator
identification
identification
Image-based
biomarker discovery
biomarker discovery