Skip to main content
. 2017 Nov 21;2(6):e00092-17. doi: 10.1128/mSystems.00092-17

FIG 2 .

FIG 2 

Comparison of FDR methods on simulated data sets. (a) Average FDR for the DS, FBH, and BH procedures (red, orange, and blue lines, respectively) as a function of the number of samples per group in simulation I. The green line indicates the prespecified FDR control level (0.1) used for the tests. (b) Number of differentially abundant taxa identified by the three methods in panel a with 100 truly different taxa and 900 truly nondifferent taxa. (c) Average FDR for DS, FBH, and BH procedures as a function of the number of rare taxa in simulation II (50 samples per group). The green line indicates the prespecified FDR control level (0.1) used for the tests. (d) Number of differentially abundant taxa identified by the three methods in panel c.