--- viewer: true license: cc-by-nc-sa-4.0 language: - en tags: - spatial-transcriptomics - histology - pathology - benchmark task_categories: - image-classification - feature-extraction - image-segmentation size_categories: - 100B- - This dataset and associated code are released under the [CC-BY-NC-ND 4.0 license](https://creativecommons.org/licenses/by-nc-nd/4.0/) and may only be used for non-commercial, academic research purposes with proper attribution. - Any commercial use, sale, or other monetization of the hescape dataset and its derivatives, which include models trained on outputs from the hescape datasets, is prohibited and requires prior approval. - By downloading the dataset, you attest that all information (affiliation, research use) is correct and up-to-date. Downloading the dataset requires prior registration on Hugging Face and agreeing to the terms of use. By downloading this dataset, you agree not to distribute, publish or reproduce a copy of the dataset. If another user within your organization wishes to use the hescape dataset, they must register as an individual user and agree to comply with the terms of use. Users may not attempt to re-identify the deidentified data used to develop the underlying dataset. - This dataset is provided “as-is” without warranties of any kind, express or implied. This dataset has not been reviewed, certified, or approved by any regulatory body, including but not limited to the FDA (U.S.), EMA (Europe), MHRA (UK), or other medical device authorities. Any application of this dataset in healthcare or biomedical settings must comply with relevant regulatory requirements and undergo independent validation. Users assume full responsibility for how they use this dataset and any resulting consequences. The authors, contributors, and distributors disclaim any liability for damages, direct or indirect, resulting from dataset use. Users are responsible for ensuring compliance with data protection regulations (e.g., GDPR, HIPAA) when using it in research that involves patient data. extra_gated_fields: Full Name (first and last): text Type of Affiliation: type: select options: - Industry - Academia - Other Current Affiliation (no abbreviations): text Current and Official Institutional Email: text Main use-case: type: select options: - Models Benchmarking - Biomarker Discovery - Diagnostics - Pathology Workflows Acceleration - Other Please add information on your intended research use: text I agree to use this dataset for non-commercial, academic purposes only: checkbox I agree not to distribute the dataset: checkbox dataset_info: - config_name: human-5k-panel features: - name: name dtype: class_label: names: '0': TENX157 '1': TENX158 '2': Xenium_Prime_Breast_Cancer_FFPE '3': Xenium_Prime_Cervical_Cancer_FFPE '4': Xenium_Prime_Human_Lung_Cancer_FFPE '5': Xenium_Prime_Human_Ovary_FF '6': Xenium_Prime_Ovarian_Cancer_FFPE_XRrun - name: image dtype: image - name: gexp dtype: array2_d: shape: - 1 - 5001 dtype: float32 - name: cell_coords dtype: array2_d: shape: - 1 - 2 dtype: float32 - name: source dtype: class_label: names: '0': hest - name: atlas dtype: string - name: age dtype: string - name: diagnosis dtype: string - name: cancer dtype: bool - name: oncotree_code dtype: string - name: tissue dtype: class_label: names: '0': breast '1': cervix '2': lung '3': ovary '4': prostate '5': skin - name: tumor_grade dtype: string - name: gender dtype: string - name: race dtype: string - name: treatment_type dtype: string - name: therapeutic_agents dtype: string - name: tumor_tissue_type dtype: string - name: assay dtype: string - name: preservation_method dtype: string - name: stain dtype: string - name: spaceranger dtype: string - name: species dtype: string - name: cytassist dtype: bool splits: - name: train num_bytes: 314500822907.79 num_examples: 178817 download_size: 311449865844 dataset_size: 314500822907.79 - config_name: human-breast-panel features: - name: name dtype: class_label: names: '0': NCBI783 '1': NCBI784 '2': NCBI785 '3': TENX94 '4': TENX95 '5': TENX96 '6': TENX97 '7': TENX98 '8': TENX99 - name: image dtype: image - name: gexp dtype: array2_d: shape: - 1 - 280 dtype: float32 - name: cell_coords dtype: array2_d: shape: - 1 - 2 dtype: float32 - name: source dtype: class_label: names: '0': hest - name: atlas dtype: string - name: age dtype: string - name: diagnosis dtype: string - name: cancer dtype: bool - name: oncotree_code dtype: string - name: tissue dtype: class_label: names: '0': breast - name: tumor_grade dtype: string - name: gender dtype: string - name: race dtype: string - name: treatment_type dtype: string - name: therapeutic_agents dtype: string - name: tumor_tissue_type dtype: string - name: assay dtype: string - name: preservation_method dtype: string - name: stain dtype: string - name: spaceranger dtype: string - name: species dtype: string - name: cytassist dtype: bool splits: - name: train num_bytes: 400920658633.13 num_examples: 234299 download_size: 400723909494 dataset_size: 400920658633.13 - config_name: human-colon-panel features: - name: name dtype: class_label: names: '0': TENX111 '1': TENX114 '2': TENX147 '3': TENX148 '4': TENX149 - name: image dtype: image - name: gexp dtype: array2_d: shape: - 1 - 322 dtype: float32 - name: cell_coords dtype: array2_d: shape: - 1 - 2 dtype: float32 - name: source dtype: class_label: names: '0': hest - name: atlas dtype: string - name: age dtype: string - name: diagnosis dtype: string - name: cancer dtype: bool - name: oncotree_code dtype: string - name: tissue dtype: class_label: names: '0': bowel - name: tumor_grade dtype: string - name: gender dtype: string - name: race dtype: string - name: treatment_type dtype: string - name: therapeutic_agents dtype: string - name: tumor_tissue_type dtype: string - name: assay dtype: string - name: preservation_method dtype: string - name: stain dtype: string - name: spaceranger dtype: string - name: species dtype: string - name: cytassist dtype: bool splits: - name: train num_bytes: 93910317910.089 num_examples: 61067 download_size: 93850601554 dataset_size: 93910317910.089 - config_name: human-immuno-oncology-panel features: - name: name dtype: class_label: names: '0': TENX138 '1': TENX139 '2': TENX140 '3': TENX141 '4': TENX142 - name: image dtype: image - name: gexp dtype: array2_d: shape: - 1 - 380 dtype: float32 - name: cell_coords dtype: array2_d: shape: - 1 - 2 dtype: float32 - name: source dtype: class_label: names: '0': hest - name: atlas dtype: string - name: age dtype: string - name: diagnosis dtype: string - name: cancer dtype: bool - name: oncotree_code dtype: string - name: tissue dtype: class_label: names: '0': bowel '1': brain '2': lung '3': ovary '4': pancreas - name: tumor_grade dtype: string - name: gender dtype: string - name: race dtype: string - name: treatment_type dtype: string - name: therapeutic_agents dtype: string - name: tumor_tissue_type dtype: string - name: assay dtype: string - name: preservation_method dtype: string - name: stain dtype: string - name: spaceranger dtype: string - name: species dtype: string - name: cytassist dtype: bool splits: - name: train num_bytes: 116194418252.55 num_examples: 67050 download_size: 116118073866 dataset_size: 116194418252.55 - config_name: human-lung-healthy-panel features: - name: name dtype: class_label: names: '0': NCBI856 '1': NCBI857 '2': NCBI858 '3': NCBI859 '4': NCBI860 '5': NCBI861 '6': NCBI864 '7': NCBI865 '8': NCBI866 '9': NCBI867 '10': NCBI870 '11': NCBI873 '12': NCBI875 '13': NCBI876 '14': NCBI879 '15': NCBI880 '16': NCBI881 '17': NCBI882 '18': NCBI883 '19': NCBI884 - name: image dtype: image - name: gexp dtype: array2_d: shape: - 1 - 343 dtype: float32 - name: cell_coords dtype: array2_d: shape: - 1 - 2 dtype: float32 - name: source dtype: class_label: names: '0': hest - name: atlas dtype: string - name: age dtype: string - name: diagnosis dtype: string - name: cancer dtype: bool - name: oncotree_code dtype: string - name: tissue dtype: class_label: names: '0': lung - name: tumor_grade dtype: string - name: gender dtype: string - name: race dtype: string - name: treatment_type dtype: string - name: therapeutic_agents dtype: string - name: tumor_tissue_type dtype: string - name: assay dtype: string - name: preservation_method dtype: string - name: stain dtype: string - name: spaceranger dtype: string - name: species dtype: string - name: cytassist dtype: bool splits: - name: train num_bytes: 97238707878.741 num_examples: 56689 download_size: 97181626515 dataset_size: 97238707878.741 - config_name: human-multi-tissue-panel features: - name: name dtype: class_label: names: '0': TENX105 '1': TENX106 '2': TENX116 '3': TENX118 '4': TENX119 '5': TENX120 '6': TENX121 '7': TENX122 '8': TENX123 '9': TENX124 '10': TENX125 '11': TENX126 '12': TENX132 '13': TENX133 '14': TENX134 - name: image dtype: image - name: gexp dtype: array2_d: shape: - 1 - 377 dtype: float32 - name: cell_coords dtype: array2_d: shape: - 1 - 2 dtype: float32 - name: source dtype: class_label: names: '0': hest - name: atlas dtype: string - name: age dtype: string - name: diagnosis dtype: string - name: cancer dtype: bool - name: oncotree_code dtype: string - name: tissue dtype: class_label: names: '0': bone '1': heart '2': kidney '3': liver '4': lung '5': lymphoid '6': pancreas '7': skin - name: tumor_grade dtype: string - name: gender dtype: string - name: race dtype: string - name: treatment_type dtype: string - name: therapeutic_agents dtype: string - name: tumor_tissue_type dtype: string - name: assay dtype: string - name: preservation_method dtype: string - name: stain dtype: string - name: spaceranger dtype: string - name: species dtype: string - name: cytassist dtype: bool splits: - name: train num_bytes: 223137438189.84 num_examples: 132040 download_size: 222975005987 dataset_size: 223137438189.84 configs: - config_name: human-5k-panel data_files: - split: train path: human-5k-panel/train-* - config_name: human-breast-panel data_files: - split: train path: human-breast-panel/train-* - config_name: human-colon-panel data_files: - split: train path: human-colon-panel/train-* - config_name: human-immuno-oncology-panel data_files: - split: train path: human-immuno-oncology-panel/train-* - config_name: human-lung-healthy-panel data_files: - split: train path: human-lung-healthy-panel/train-* - config_name: human-multi-tissue-panel data_files: - split: train path: human-multi-tissue-panel/train-* --- # HESCAPE • PyArrow Format HESCAPE (**H&E + Spatial Contrastive Pretraining Benchmark**) is a large-scale benchmark for multimodal learning in spatial transcriptomics. This repository hosts the **PyArrow-formatted Hugging Face datasets** for HESCAPE, organized by panel as dataset **configs**. --- ## Available Configs (Panels) This dataset repo exposes the following configs: - `human-5k-panel` - `human-breast-panel` - `human-colon-panel` - `human-immuno-oncology-panel` - `human-lung-healthy-panel` - `human-multi-tissue-panel` Each config corresponds to an independent HESCAPE dataset panel. --- ## Schema Each dataset entry contains the following columns: | Column | Type | Description | |---------------|-----------|-------------| | `name` | class_label| Unique identifier for the sample | | `image` | image | Image patch | | `gexp` | array | Transcriptomic expression based on gene panel| | `cell_coords` | array | Coords of the image-gexp pair in tissue | | `source` | string | Source of data | | `atlas` | string | Label for atlas | | `age` | string | Age | | `diagnosis` | string | Diagnosis : "Cancer" or "None" | | `cancer` | bool | Whether cancer or not | | `oncotree_code` | string | Oncotree code | | `tissue ` | class_label | Tissue label | | `tumor_grade` | string | Grade of tumor | | `gender` | string | Gender | | `race` | string | Race | | `treatment_type` | string | Treatement type | | `therapeutic_agents` | string | Therapeutic agent | | `tumor_tissue_type` | string | Tumor tissue type | | `assay` | string | Assay used | | `preservation_method` | string | Preservation method used | | `stain` | string | Stain of histology | | `spaceranger` | string | Spaceranger version | | `species` | string | Species | | `cytassist` | string | Boolean | --- ## Usage Load a specific panel (config): ```python from datasets import load_dataset # Example: load the human breast panel ds = load_dataset( "Peng-AI/hescape-pyarrow", name="human-breast-panel", split="train", streaming=True ) print(ds) ``` ## List all configs ```python from datasets import get_dataset_config_names get_dataset_config_names("Peng-AI/hescape-pyarrow") ``` ### How to cite: ``` @misc{gindra2025largescalebenchmarkcrossmodallearning, title={A Large-Scale Benchmark of Cross-Modal Learning for Histology and Gene Expression in Spatial Transcriptomics}, author={Rushin H. Gindra and Giovanni Palla and Mathias Nguyen and Sophia J. Wagner and Manuel Tran and Fabian J Theis and Dieter Saur and Lorin Crawford and Tingying Peng}, year={2025}, eprint={2508.01490}, archivePrefix={arXiv}, primaryClass={q-bio.GN}, url={https://arxiv.org/abs/2508.01490}, } ``` ### Contact: - Rushin Gindra Helmholtz Munich, Munich (`rushin.gindra@helmholtz-munich.de`) - The dataset is distributed under the Attribution-NonCommercial-ShareAlike 4.0 International license (CC BY-NC-SA 4.0 Deed)