Single-cell Data

In this module, a brief summary of single-cell RNA-seq datasets are listed as follows.

Quality control. Each dataset is filtered on its own terms rather than by one fixed UMI window: where the source publication states thresholds they are reproduced exactly, and where it does not, outliers are called per library by median absolute deviation. The Cell Atlas records the figures actually applied to every dataset, including cells before and after filtering and doublet rates. Cells reads before filtering → after filtering, and Libraries is how many libraries were integrated into that dataset, which is what makes its retention figure readable. A row marked not re-analysed is one whose count matrix could not be obtained from the public deposit; the reason is given in the cell, because the deposits differ — some hold only per-sample matrices, some hold normalised values rather than counts, and one is scATAC rather than scRNA-seq.