Introduction

Principal component analysis (PCA) is commonly used to reduce the dimension of abundance profiling, and to view all samples in two-dimension graphs. The PCA visualization application contains the scatter plot of samples based on PCA coordinate, the default axes are PC1 and PC2, and users could swap to the other PCs provided in the input files. Users should provide at least one category of metadata for grouping samples with different colors in the visualization. With multiple metadata categories, users could use discrete values to shape samples and discrete/continuous values to color samples.

If you have the raw profiling table of samples without differential testing analysis, you can use PCA module.

If you want to visualize your processed results, you can upload the following inputs to interactively visualize online:

Input Files

Click on the Download Data button to check the demo input files.

1. Metadata/group information (TSV file)

• Example input

Here is an example of the header of the metadata file. The attribute will be used in the visualization for color, shape, and sider boxplot. The metadata can be continuous (e.g. Age) and discrete (e.g. gender) variables.

2. PCA coordinate (TSV file)

• Example input

This file contains samples with PCA coordinates.

Display Interactions

• Tooltips
  Hover on the box or the scatter, the info tooltips will show.

• Click to highlight scatter
  Click on a sample scatter to highlight it with auxiliary axises.

• Double click to hide scatter
  Double-click on a sample scatters to hide it. Hidden samples can be re-shown in the filter samples editor.

• Download
  One SVG file will be generated when the Download Chart button is clicked.

Editor Functions

• Files

  • Manage Files: checklist of files uploaded previously, delete or download files.
  • Upload: upload files. Note that the duplicated file name will be alerted and given a random postfix.
  • Choose: choose files uploaded previously.

• Data

  • Taxonomic Rank: select the taxonomic rank of the data to display.
  • Scatter shape by: selection of scatterplot sample shapes differentiation basis.
  • Scatter color by: select the basis of the scatter color to display.
  • Box category by: select the category of the boxplot.

• General Settings

  • Grid Length: input a value to adjust the length of the grid.
  • Box Height: input a value to adjust the height of the box.
  • Scatter Size: input an integer to adjust the size of the scatter.
  • Hollow Scatter: click to make scatters hollow.
  • Filter Samples: select the samples you want to display.

• Box Settings

  • Hollow box: check it to hollow boxes.
  • Outlier: check it to draw outliers.
  • Sample Scatters: check it to draw sample scatters over the boxes.
  • Draw Violin: check it to draw violins instead of boxes.

• Meta Panel

  The visualizer will automatically distinguish discrete and continuous metadata, and provide different editor functions.

  • $discrete-data: click reorder and recolor button to launch the corresponding interface.
  • $continuos-data: click reorder and recolor button to launch the corresponding interface; the value range can be adjusted in the recolor interface.
  • Apply changes: click it to apply the changes and redraw the chart.

Manual version=1.1, updated by Dr. JIANG Yiqi on 2022-04-19.

Manual version=1.0, written by Dr. JIANG Yiqi and Ms. WANG Yanfei on 2021-09-13.