Shaow/mousehippocampus_slideseqv2

    # Mouse Hippocampus Slide-seqV2 · Puck_200115_08

    Curated, ready-to-load spatial transcriptomics dataset.

    ## Source

    - Paper: [Stickels et al., Nat. Biotechnol. 2021](https://www.nature.com/articles/s41587-020-0739-1)
    - Canonical download: figshare 21829905 (file id 38736651) — derived from Single Cell Portal SCP815

    ## Scale

    | Property | Value |
    |---|---|
    | Technology | Slide-seqV2 (10 µm beads) |
    | Species | Mus musculus |
    | Tissue | Mouse hippocampus |
    | Sections / slices | 1 |
    | Total cells / spots | 35,349 |

    ## Files

    - `AdataMH1.h5ad`

    Each `.h5ad` follows the AnnData spec:
    - `.X` — gene expression matrix (cells × genes), sparse where natural
    - `.obs` — per-cell annotations (see "Metadata" below)
    - `.obsm['spatial']` — `(n_cells, 2)` float32 spatial coordinates
    - (where present) `.layers['count']` — raw integer counts
    - (where present) `.obsm['spatial3d']` — `(n_cells, 3)` float32 (x, y, z=section)

    ## Metadata (`obs` columns)

    `leiden`, `n_genes_by_counts`, `total_counts`

    ## Notes

    Puck_200115_08 mouse-hippocampus Slide-seqV2 puck (35,349 beads × 20,527 genes), QC-filtered and normalized. `obsm['spatial']` is set from `obs[xcoord, ycoord]`. Raw counts in `.X` (sparse float32, max ≈ 1900).

    ## Usage

    ```python

import scanpy as sc from huggingface_hub import snapshot_download d = snapshot_download(repo_id='Shaow/mousehippocampus_slideseqv2', repo_type='dataset') adata = sc.read_h5ad(f'{d}/AdataMH1.h5ad')

        ## Citation

        If you use this dataset, please cite the source paper above.

        ## License

        MIT for the curation/preparation. Underlying data inherits the license of
        the upstream publication (typically CC-BY-4.0); please see the source paper.