scMARK an 'MNIST' like benchmark to evaluate and optimize models for unifying scRNA data

Here we present a novel benchmark dataset (scMARK), that consists of 100,000 cells and 10 studies and can be used to ask how well models integrate data from different scRNA studies. Data is provided as aData *h5ad files that can be read with Python's library Scanpy. File 10k_cells_per_study.tar.bz2 contains scMARK, with one *h5ad file for each of 10 studies, downsampled to 10,000 cells per study. File 2k_cell_per_study_10studies.tar.bz2 contains a single *h5ad file with all 10 studies, downsampled to 2,000 cells per study. The matrix of UMI counts contains the intersection of genes in all 10 studies.