Table 1 Dataset summary. See 5: 5.4 for the details of these datasets. For each simulated dataset, the actual cell type label set is known. For each experimental dataset, the benchmark label set was obtained by the experimental and bioinformatics analysis process. This process includes fluorescence-activated cell sorting (FACS), known feature gene checking, cell screening, and clustering. Thus, although these benchmark label sets are not the actual label sets, they reflect our best knowledge of the cell types

From: Clustering Deviation Index (CDI): a robust and accurate internal measure for evaluating scRNA-seq data clustering