User-friendly shiny application for the interactive utilization of DESeq2. In addition to providing the standard DESeq2 methods, DESeq2-Vis provides further normalization options, as well as the visualization of gene profiles.
The following section provides a quick rundown of the standard worklow and functions of DESeq2-Vis.
Start R
Enter:
library(shiny)
runGitHub("Shiny_DESeq2", "Integrative-Transcriptomics")
When running DESeq2-Vis for the first time, all required packages will be installed automatically.
The following files are required for running DESeq2-Vis on your experimental data:
- A .tsv-file containing the raw counts for each sample in one column and locus tags as rownames. Tables acquired from featureCounts can also be directly uploaded to DESeq2-Vis.
- An experimental design table (.tsv-file) containing sample names as row names and experimental conditions in columns. Each column containing an experimental condition to be analyzed must be indicated by containing the keyword condition. DESeq2-Vis will automatically scan this file and replace the column name of each sample in the counts-table by a merged string of the corresponding experimental conditions. A suffix indicating the replicate number is added to each sample name. Rows corresponding to samples that are not contained in the counts-table will be removed from the design table.
- A GFF-file containing annotations and gene descriptions. Based on the locus tags, the corresponding gene name is added to each locus tag (unannotated genes will only contain the locus tag).
- Press the Upload! button to start the upload- and scanning process.
- Select an Experimental Variable of your choice as a basis for the normalization and differential expression analysis.
- Select a normalization method.
- Adjust the significance level.
- Press the Run DESeq! button in order to apply normalization.
The Normalization-tab provides an overview of the normalized data and general data analysis tools for quality control and gene profile analysis.
- Normalized Counts: Contains normalized counts based on the specified method as well as a TPM-normalized table.
- Boxplots: Provides simple boxplots of the normalized counts for each sample for quality control purposes (WIP).
- PCA: Dot shape and color can be adjusted based on up to two experimental conditions. Use sliders to adjust plot- and font-size. Plot the PCA by pressing the Refresh Plot!-button.
- Heatmaps: Contains two different, UPGMA-clustered heatmaps:
- Pairwise distance between samples (euclidean distance of log2-normalized counts).
- Heatmap of genes with the highest variance. Use the slider to adjust the amount of displayed genes and plot size.
- Profile Plots: Use to display the the gene expression of one or multiple genes per condition or individual sample (use the checkbox Average replicates to switch). Gene profiles can be displayed as gene-wise expression profile (tab Sample Profiles) or Mean Expression Profile over all selected genes.
The calculation of differential expression is carried out under Differential Expression-tab.
The differential expression between two experimental groups is calculated as follows:
- Select a first and second experimental group. The log2-foldchange is calculated as
$log$ (1st group) -$log$ (2nd group). - Upon pressing Add to table!, the amount of signficantly up- and down-regulated genes as well as the total amount of significantly differentially expressed genes is display in tabular format.
- Press Show Genes to view the differential expression between the two conditions, gene descriptions as well as an interactive volcano plot.