Rows with only one hit (k = 1) display raw p-values, but their adjusted p-values (FDR, Holm) are set to NA. Multiple-testing corrections are computed exclusively across sets with two or more hits.
Reason: Reference libraries contain thousands of taxon sets, and a single taxon often appears across hundreds of them. Including these single-hit sets in the multiple-testing procedure severely inflates the total number of hypotheses tested, diluting statistical power for genuinely enriched sets. Furthermore, a single shared taxon rarely provides sufficient evidence of set enrichment on its own.
Note: Single-hit sets remain in the table for reference, but because they were excluded prior to multiple-testing adjustment, their raw p-values should not be used to claim statistical significance.