Seems that the initial GO analysis is not correcting for sampling bias (tissue/technology) by using the study specific background?

Dear James,

Our protocol was designed to aid in the visualization of the user's GO enrichment results of choice. The intent was to provide a simple script to visualize enriched GO terms, and not for performing GO enrichment. For example, we used the GO enrichment option in TAIR with specific settings and described how this was done. Therefore, regardless of how you perform the GO enrichment and obtained the initial input GO results, our script allows you to visualize the output. Indeed methods that account for biases in GO enrichment analyses should be considered by the users and their desired methods of choice. Thank you for your thoughtful comment.

Thanks Dawn. The use I’ve noted didn’t understand how to generate a valid initial GO analysis- perhaps because the working example doesn’t emphasise the critical need to use the experiment specific background to generate the adjusted p-values for the regulated list. Perhaps worth making an explicit statement for this requirement within your script to avoid “misuse”? Thanks for replying.