Prepared sections and references for a spatial transcriptomics imputation baseline
收藏资源简介:
Prepared sections and references for the spatial transcriptomics imputation baseline Input files of the benchmark in https://github.com/LazarSmiljkovic/st-imputation-baseline, as prepared for the CoDi cell-type annotation tool, with CoDi's per-spot labels. The benchmark reads them from the folder given by the ZONEIMPUTE_DATA environment variable; the other inputs are downloaded from their original deposits by scripts/download_data.sh in the repository (see DATA.md there). Spatial sections (spots x genes, raw counts) ST_Mouse_kidney.h5ad Slide-seqV2 mouse kidney puck, GEO GSE190094 (Marshall et al. 2022); obsm["X_spatial"] ST_Adult_Mouse_Brain.h5ad Stereo-seq adult mouse brain section B5; obsm["spatial"], obs["celltype_pred"] ST_Mouse_embryo_whole_brain.h5ad MOSTA E16.5 E1S3 whole brain cell bins (STOmicsDB STDS0000058), depth-equalised values; the benchmark uses its cells, genes and coordinates and takes the raw counts from the original MOSTA file ST_Mouse_Brain_Visium.h5ad Visium mouse brain section used in development runs only Single-cell references SC_Mouse_Kidney.h5ad GSE157079 kidney reference, log-normalised, obs["clusters"] SC_mouse_brain_L5.h5ad Zeisel et al. 2018 mouse nervous system atlas, level 5, raw counts, obs["cell_class"], obs["cell_subclass"] SC_Mouse_Brain_Visium.h5ad reference of the development Visium section Labels mouse_kidney_CoDi.csv, adult_mouse_brain_B5_CoDi.csv, whole_brain_mouse_embryo_CoDi.csv, visium_mouse_brain_CoDi.csv CoDi labels per spot: cell_id, CoDi_dist, confidence_dist, CoDi_contrastive, confidence_contrastive, CoDi tools_predictions.zip labels of the same sections by Tangram, cell2location, Seurat and CytoSPACE md5sums.txt lists the checksums. The original data are public under the terms of their deposits; please cite the original studies when using them.



