11. Taxonomical Heatmap - raytonghk/genepiper GitHub Wiki

Taxonomical Heatmap module provides method to plot heatmap of samples at a specified taxonomic rank.

Load Data and Subsample

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

Parameters

Parameters panel provides the options to customise the heatmap:

Taxonomic Rank For Agglomeration: select the taxonomic rank to agglomerate counts. The taxa will be shown on the y-axis.

Abundance Type: choose between

  • Raw Count the original read counts,
  • Rarefied Count rarefying sample read counts to the lowest counts amongst samples, and
  • Relative Abundance counts are divided by the sum of counts in each sample.

Sample Order By: select how the samples should be arranged along the x-axis. Three options are provided:

  • Name (by sample name in alphabetical order);
  • Group (by group which is assigned by a variable in the sample data table and the sample label will be coloured according to the grouping), and
  • Hierarchical Clustering (by clustering result calculated with a specified method and distance, see hclust).

Taxa Order By: select how the taxa should be arranged along the y-axis. Two options provided:

  • Name by taxa name in alphabetical order, and
  • Hierarchical Clustering by clustering result calculated with a specified method and distance.

Display: select the method to filter the results for the best display. Normally, two options are provided:

  • Top Abundance displays only the top n (specified by user) taxa with the highest total read counts,
  • Least Abundance displays only the n taxa with the lowest total read counts.

If samples are assigned to groups, two more options are available for the filtering:

  • Anova Significant displays only the top n taxa according to ANOVA comparison amongst groups, arranged by the significance and p values, and
  • Kruskal Significant displays only the top n taxa according to Kruskal-Wallis comparison amongst groups, arranged by the significance and p values.

Rescale Taxa?: an option to rescale the count of each taxa amongst samples:

  • None no rescaling,
  • Normalization rescales read counts into range between 0 and 1, and
  • Standardization rescales read counts to mean = 0, sd = 1.

Result

Taxonomical Heatmap module plots heatmap of samples at a specified taxonomic rank with options to filter and arrange the results. The plot can be downloaded by clicking the Download button.

Graphic Parameters

Graphic Parameters panel provides options in the Title, Axis and Legend tabs to customise the plot.

In Color tab

Color Scale: select a color scale in the heatmap, options are:

  • Blue (default)
  • Heat Color
  • Green Red
  • Viridis
  • Green
  • Red