How to set up a pair-wise comparison for differential expression analysis

Comparisons can be defined during the experiment design process and after the experiment has finished processing.

To create a new comparison during experiment set up, click . To create a new comparison after experiment processing, select .

Comparisons can be created at the replicate group level (using the attributes) or individual sample level (using the sample names) by dragging and dropping different attributes or samples into the condition and control boxes on the right side of the page. Combinations of two or more different attributes can be used to create comparisons with specific subsets of samples. In the example image below, a comparison is set up across two treatment groups (Radiation vs Control) at a specific time point (1 hour). Alternatively, the comparison could be set up specifically at only the 48 hour time point, or including both time points simultaneously.

The number on the bottom of each group box shows the total number of samples that qualify. You can view and deselect specific samples by clicking on the EDIT button in the bottom right corner and then clicking on the triangle to see a check list. For best statistical practices, it is recommended to include a minimum of 3 samples on each side of the comparison. 

Additionally, covariate correction can be added to the statistical model using the "SELECT COVARIATE" button. Click here to learn more about covariate correction for RNA-seq analysis.

Click the  button to save the created comparison.