skbio.stats.composition.ANCOMBCResult.pairwise_test#

ANCOMBCResult.pairwise_test(alpha='inherit', p_adjust='inherit')[source]#

Perform pairwise directional test between all group pairs.

Uses mixed directional FDR (mdFDR) correction via bootstrap.

Parameters:
alphafloat or “inherit”, optional

Significance level. Default is “inherit”, which will use the value supplied upstream.

p_adjuststr, optional

Family wise error (FWER) controlling method. Default is “inherit”, which will use the p-value correction method supplied upstream.

Returns:
pd.DataFrame

Pairwise test result with a (FeatureID, Comparison) multi-index. The index levels and columns are:

  • FeatureID: Feature identifier, i.e., dependent variable.

  • Comparison: Group contrast being tested.

  • Log(FC): Estimated difference in group effects on the natural-log abundance scale.

  • SE: Standard error of the estimated contrast.

  • W: W-statistic, calculated as the estimated contrast divided by its standard error.

  • pvalue: Uncorrected p-value of the W-statistic.

  • qvalue: p-value after mixed directional false discovery rate correction.

  • Signif: Whether the contrast is significantly different from zero.