This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Go Analysis, Bp, Cc, Mf Separate Or Combined

Hi All:

When we perform a GO enrichment analysis, is it more common practice to consider each root category (BP, CC, MF) separately or combined (i.e. consider all categories)? I saw some of the packages rank the terms on BP only (or CC, or MF), while others may have a list of all three roots. Are there any pros and cons for the two practices?

Thank you!

gene ontology multiple
• 27,137 views • link
updated by boczniak767 boczniak767 • written by alittleboy alittleboy

For purposes of simplicity for future visitors to this site, we should define abbreviations for the less initiated: GO = Gene ontology; BP = Biological process; CC = Cellular component; MF = Molecular function

• 0 views • link
written by Malachi Griffith Malachi Griffith

2 answers

This depends on the goals of your analysis. If you are only interested in enrichment of BP terms (which is commonly the case) then only include the BP category. The advantage of this is mainly that it reduces the number of tests and therefore might help with the multiple testing problem. If however you are also interested in enrichment of CC/MF terms then I would include all categories combined. Doing each separately might be considered "cheating" when it comes to correcting p-values for multiple testing.

• 0 views • link
written by Obi Griffith Obi Griffith

It also depends on test used. From developers I learned that mgsa method (used in Ontologizer program and mgsa package from bioconductor) should be performed separately for each category (CC/BP/MF).

• 202 views • link
written by boczniak767 boczniak767

Log in to answer this question.

More posts like this