Skip to main content
. 2020 Jun 3;21:225. doi: 10.1186/s12859-020-03552-z

Fig. 5.

Fig. 5

Comparison of recall and precision with data generated from the zero-inflated log-normal model across different δ. a and b Recall and precision of DESeq2, ANCOM, metagenomeSeq, Wrench, and that of t-test, which was applied after counts were normalized by ALDEx2 or eBay