Fig 1.
Overview of the FAVABEAN and FALAPhyl workflow for 16S rRNA amplicon analysis.
Fig 2.
Beta diversity (PhILR distances) Principal Coordinate Analysis (PCoA) plots.
A) ASVs of V1-V3, B) ASVs of V4-V5, and C) Sidle-reconstructed taxonomies. Graphs were adjusted for clarity by modifying axis labels inside the graphs and legend placement.
Table 1.
Statistics of the pairwise PhILR distances of three samples (buccal, saliva, or tongue) to the supra- and sub-gingival samples. Note that PhILR is not boundary constrained between −1 and 1, therefore distance comparisons should only be done within the same feature resolution level.
Fig 3.
Breakdown of Jaccard dissimilarity.
Fig 3A: Dissimilarity between supragingival and subgingival plaque in microbial membership within the same participant, illustrated through the two components that form the Jaccard dissimilarity; turnover, and nestedness. Fig 3B: Observed features alpha diversity between supragingival and subgingival plaque samples within the same individual.
Table 2.
Non-self sources of subgingival plaque using V1-3 region.
Fig 4.
Principal Coordinates Analysis (PCoA) based on PhILR distances.
(A) PCoA plots show dispersion using the standard deviation ellipses (representing one standard deviation of the spread of the samples away from the centroid, which may become cumbersome to illustrate any differences). (B) An alternate visualization of the same data using the probability density functions along principal axes, providing clearer distributional insight. The pipeline also generates Sample-labeled PCoA plots automatically.
Fig 5.
Scores from the paired testing in Intervention arm, baseline and post-intervention paired samples.
The test shows the dataset interrogated against multiple spike-in trials, with different effect sizes. Black triangle illustrates the average of the effect sizes. As none of the averages reach the zero threshold, none of the methods are capable of reliably identifying the statistically significant features.