Difference between MEME discriminatory and differential enrichment mode

[peeps]

Hello,

I am having difficulty understanding the difference between MEME's discriminatory and differential enrichment mode options and under what scenarios one might be preferred over the other. Could someone explain the difference?

Thanks! 

[cegrant]

Discriminatory and differential enrichment modes address the same problem using two different algorithms.  In both cases you MEME is looking for motifs that appear more in the target set than in the control set. In the case of discriminatory search position specific priors are generated for the target set. The position specific priors are used to focus MEME on patterns in the target set that don't appear in the control set. Differential e

nrichment simply counts how often motifs appear in the target set than in the control set, and scores motifs that appear mostly in the target set, higher. You can try either method and use the one that seems to work better.

More information is available in the online help on the public MEME web application. Click on the red question mark icon next to "Select the motif discovery mode". 

This documentation may also be helpful:

MEME (command line)

psp-gen

Timothy L. Bailey, Mikael Bodén, Tom Whitington and Philip Machanick, "The value of position-specific priors in motif discovery using MEME", BMC Bioinformatics, 11(1):179, 2010