Time-resolved single-cell RNA-seq datasets used in CellDyc

This collection contains six time-resolved single-cell RNA-seq datasets provided as processed AnnData (.h5ad) files. Five datasets are derived from previously published studies, and one dataset is a synthetic time-series dataset generated using scDesign3 and the CellDyc method. All datasets are formatted for use with CellDyc (“Experimental Time Points Guided Transcriptomic Velocity Inference”) and as example data for the CellDyc package. In addition to the datasets, this collection also provides a compressed archive celldyc_model, which contains precomputed models used for reproducing the results in the CellDyc manuscript. Datasets included: - gbm_mono2tam.h5ad (in vivo timestamping of monocyte infiltration and differentiation into Arg1⁺/Acp5⁺ TAMs in glioblastoma models via Zman-seq; Kirschenbaum et al. 2023) - GSE126954_ab_unique.h5ad (a curated subset of the C. elegans AB lineage (333 cells) spanning three experimental time points; Packer et al. 2019) - reprogramming_lineage.h5ad (a CellTagging lineage-traced dataset of 3,049 cells across six time points; Biddy et al. 2018) - gastrulation_erythropoiesis.h5ad (an scRNA-seq dataset capturing the complete erythroid lineage specification during mouse gastrulation across eight experimental time points; Pijuan-Sala et al. 2019) - zebrafish.h5ad (a curated subset of 2,341 cells from a zebrafish embryogenesis atlas; Farrell et al. 2018) - scdesign3_5000c500g.h5ad (a synthetic time-series dataset of 5,000 cells and 500 genes generated via scDesign3 and CellDyc; Song et al. 2024 and CellDyc preprint) If you wish to use any of these datasets in your own research, please cite the corresponding publications listed in the References section.