FindMarkers adjusted p value. Why is n = all genes in the dataset?

[amparora]

Hi,

The differential expression analysis done by the FindMarkers function also performs Bonferroni correction using all genes in the dataset. However, one could test, for example, only those genes expressed in at least 10% of the cell in either of the groups to compare. This can reduce quite a lot, the number of genes tested. In this case, using the total number of genes would overcorrect the p-values and significant DEGs will be lost.
Why is the total number of genes used for correction instead of the number of genes tested?

Thanks in advance for the feedback!