29. Kruskal Wallis Test - raytonghk/genepiper GitHub Wiki

Kruskal-Wallis Test module performs kruskal-wallis test on individual taxa according to the assigned groups.

Load Data and Subsample

Analysis starts from loading data by Load Data panel. User can select project and data label to load the saved data. After loading data, subsampling can be done by Filter panel if needed.

Parameters

Parameters panel provides options for the test design.

  • Taxonomic Rank For Agglomeration: select taxonomic rank for agglomerate count. The naming of taxa will follow the taxonomical rank selected.
  • Abundance Type: choose between Rarefied Count (downsample the count to the lowest count amongst sample) and Relative Abundance (each OTU count divided by the sum of count in each sample).
  • Group Column: select column in sample data to group the sample. The follow checkboxes will updated and users can select which group they would like to include into the test.
  • Significant Level: set the p value cutoff for filtering the result.

Results

Result panel shows the table of taxa that significantly different amongst groups. User can select the taxa and shows the plot of the taxa. The four plot type was provided in Plot Type tab of Graphic Parameters panel.

  • Box Plot,

  • Notched Box Plot,

  • Violin Plot,

  • Dot Box Plot,

Graphic Parameters

Graphic Parameters panel provides options to control plot appearance.

Plot Type tab provides option to select the type of plot.

Mann Whitney tab provides option to check if plotting significant annotation.