CLEAR-IT supplementary data: models, embeddings, and analysis outputs for tumor microenvironment cell phenotyping

This dataset provides supplementary research data for the manuscript "CLEAR-IT, a framework for Contrastive Learning to Capture the Immune Composition of Tumor Microenvironments".

The archive contains derived and supporting artifacts used to reproduce the analyses, including precomputed embeddings, pretrained model weights, and precomputed output tables/results for benchmarking, ranking, SHAP analysis, survival analysis, and hyperparameter evaluation workflows.

The supplementary artifacts cover analyses on multiple cohorts/modalities used in the study:
TNBC1-MxIF8 (Hammerl et al., 2021, https://doi.org/10.1038/s41467-021-25962-0; dataset record: https://doi.org/10.4121/126d8103-6de5-4493-a48e-5d529fef471e),
TNBC2-MIBI44 (Keren et al., 2018, https://doi.org/10.1016/j.cell.2018.08.039; data resources: https://www.weizmann.ac.il/mcb/Keren/resources and https://www.angelolab.com/mibi-data),
CRC-CODEX26 (Schurch et al., 2020, https://doi.org/10.1016/j.cell.2020.07.005; image resource: https://doi.org/10.7937/TCIA.2020.FQN0-0326), and TONSIL-IMC41 (Hunter et al., 2024, https://doi.org/10.1002/cyto.a.24803; image/segmentation resource: https://www.ebi.ac.uk/biostudies/bioimages/studies/S-BSST1047).

A README.MD file is included in the ZIP archive and provides the detailed folder inventory, file-level organization, and usage notes.

This record is intended to distribute reproducibility assets for the CLEAR-IT study. Dataset-specific source data provenance remains documented in the referenced original dataset records/publications.

https://github.com/qnano/CLEAR-IT/tree/main/clearit