GOMO specificity and ranking parameter

[Devinder Kaur]

Hi Charles,

I have few queries from the GOMO search results

- Does higher specificity should be considered for selecting the function of motif?

- Does GOMO look for motif in normal as well as reverse complementary ?

- What sense does the ranking number make ? 

[CharlesEGrant]

- Does higher specificity should be considered for selecting the function of motif?

If you click on the red question mark next to the specificity column in the GOMo output, you'll see the following definition:

A measure of the relative position in the GO hierarchy showing how specific a term is. A value of 100% implies that the GO term is the most specific form whereas a value of 0% implies that the GO term is one of the three roots of the Gene Ontology with only terms below it. A tilda (~) infront of the percentage implies that the value shown has been rounded and is not exact.

So yes, you should definitely consider the specificity. The lower the specificity, the less useful the associated GO term is likely to be.

- Does GOMO look for motif in normal as well as reverse complementary ?

Yes, internally GOMo uses AMA  to compute motif affinity. AMA defaults to using both forward and reverse orientations of the motif. AMA has command line options to turn this off, but those options are accessible for GOMo.

- What sense does the ranking number make ?

GOMo  creates a list of genes ordered by increasing affinity of the motif for the region upstream of the gene. The ranking number is the position of the gene in that list. That is, genes with lower ranking  have less affinity for the motif.