Marker genes are ranked by the Wilcoxon rank-sum test within each cell type, against all other cells. Because an adjusted p-value is minute for almost every gene once a dataset runs to tens of thousands of cells, the table is better filtered on log2 fold change and percent detected. Click any gene to plot its expression; click any cell type to isolate it in the atlas.
Where an identifier could be established, the gene is shown as the accession the rest of this site uses for Hydra vulgaris, with the source dataset's own ID in parentheses (hover for why; the rule and the coverage are in the manual). Those IDs are the keys the underlying data files use, so they are what the links here carry. 280 of the 630 identifiers listed on this page carry a not mapped mark: they are the source study's own gene models, and no accession could be justified for them. They are left as they are rather than given a best guess.
