Significance Analysis

The Significance Analysis tab is divided into two main sections: the side panel and the main panel.

Side Panel đź“š

In the side panel, you have the following options:

1. Choose Groups to Compare

Select the groups from your data for which you want to perform significance analysis.

  • For DESeq preprocessing, select from predefined factors.
  • For other preprocessing methods, choose from available sample annotation columns.

2. Choose Comparisons

Select the specific pairings of groups for which you want to perform significance analysis.

  • Automatically generates possible pairings based on selected groups.
  • Notation is “Treatment:Control” and indicates the direction of the comparison.

3. Choose Test Method

Select the statistical test method for significance analysis.

4. Choose Significance Level

Set the desired significance level (alpha) for hypothesis testing.

  • Slider input allowing selection between 0.005 and 0.1 with a default value of 0.05.

5. Choose Test Correction

Select the method for correcting p-values to account for multiple testing.

1. None No correction is applied to p-values. Each test is considered independently.
2. Bonferroni Adjusts the significance level by dividing it by the number of tests. Controls the family-wise error rate.
3. Benjamini-Hochberg Controls the false discovery rate (FDR) by adjusting p-values. Often more powerful than Bonferroni.
4. Benjamini Yekutieli An extension of Benjamini-Hochberg for controlling the FDR under dependence. Suitable when tests are correlated.
5. Holm A step-down method that controls the family-wise error rate. Adjusts p-values sequentially.
6. Hommel A method that controls the family-wise error rate and is more powerful than Bonferroni. Accounts for arbitrary dependency structures.
7. Hochberg A step-up method that controls the family-wise error rate. Adjusts p-values sequentially.
8. FDR (False Discovery Rate) Controls the expected proportion of falsely rejected null hypotheses. More liberal than family-wise error rate methods.
For a bit more information on multiple testing correction, [read here](https://en.wikipedia.org/wiki/Multiple_comparisons_problem).

Main Panel đź’ˇ

The main panel displays the results of the significance analysis. The main panel has several tabs:

  • Result Visualization: This tab shows similarities and differences between the groups you compared.
    • Visualization Choices: Allows you to customize the visualization of the results.
      • Select your desired comparisons to visualize: Choose which comparisons to include in the visualization.
      • Visualization method: Select the type of plot to use. If the number of comparisons is smaller than 4, you can choose between UpSetR plot and Venn diagram. If the number of comparisons is larger than 4, only the UpSetR plot is available.
      • Type of significance to look at: Choose between Significant and Significant unadjusted.
      • Intersections to highlight: Specify the number of intersections to highlight in the plot. (Only available for UpSetR plot)
    • Venn Diagram: Shows the intersections of significant results across different comparisons as overlapping circles. The number within each intersection is equal to the number of significant results that the circles sharing the intersection have in common.
    • UpSetR Plot: The UpSetR plot section visualizes the intersections of significant results across different comparisons.
      • Intersection Size: The bars represent the number of significant results for each intersection.
      • Set Size: The horizontal bars show the number of significant results in each comparison group.
      • Intersections: The dots and lines indicate which groups are part of each intersection.
    • Download Options: You can save the plots in different file formats such as PNG, TIFF, and PDF. You also have the option to download the underlying R code and data. For more information, check out Interface Details.
  • Comparison tabs: For each selected comparison, a separate tab is created in which the results can be viewed. They consist of a Volcano plot and a table.
    • Table: A sortable table, displaying pvalue, padj, log2FoldChange, and other information for each entity tested. The table can be downloaded as a .csv or . xlsx file. The table can be sorted by clicking on the column headers.
    • Volcano Plot: The volcano plot section is divided into two plots:
      • Corrected p-Values: Shows the log fold change versus the corrected p-values.
      • Uncorrected p-Values: Shows the log fold change versus the uncorrected p-values.
      • Both plots simultaneously or separately can be downloaded directly in common formats (e.g., PNG, TIFF, PDF) or sent to the report. You can also download the underlying R code and data. For more information, check out Interface Details.

Other Notes đź“Ś

  • Question Marks: The displayed question marks provide quick and immediate help. However, since you are reading this documentation, you found the extensive version. Hope it helped!
  • Interpretation: The plots and table can give insights into the significant changes between the compared groups. Adjusting the significance level and thresholds can help in identifying the most relevant results. Checking the intersections can point towards commonalities between the groups.

Further Navigation

Do you want to…