Training_sc

This dataset contains various scRNA-seq datasets that are used in the omnideconv benchmarking study. Each dataset is stored as a SummarizedExperiment object in RDS format (compressed file format for R).The objects contain the following data:[assay] 'counts': raw gene expression counts[assay] 'tpm': CPM-normalized gene expression counts[metadata] 'batch': batch information of samples in dataset[metadata] 'cell_type': the cell-type annotation that was used in the benchmark. Often, this included manual refinement and filtering of the original dataset. Some datasets contain annotations with different resolution levels (fine, coarse).Included datasets are:Hao: blood, human [https://doi.org/10.1038/s41591-021-01329-2]HaoSub: subset of Hao, 10% of each cell-typeWu: breast cancer, human [https://doi.org/10.1038/s41588-021-00911-1]Allen: brain post-mortem, human [https://doi.org/10.1038/s41586-021-03465-8]TabulaSapiens: blood, human [https://doi.org/10.1126/science.abl4896]Lambrechts: lung cancer, human [https://doi.org/10.1038/s41591-018-0096-5]SchulteSchrepping: blood, human [https://doi.org/10.1016/j.isci.2021.103115]Arunachalam: blood, human [https://doi.org/10.1016/j.isci.2021.103115]Heimlich: blood, human [https://doi.org/10.1182/bloodadvances.2023011445]DominguezConde: blood, human [https://doi.org/10.1126/science.abl5197]Maynard: lung cancer, human [https://doi.org/10.1016/j.cell.2020.07.017]Lee: blood, human [https://doi.org/10.1126/sciimmunol.abd1554]TabulaMuris: spleen, mouse [https://doi.org/10.1038/s41586-018-0590-4]