MEME-ChIP: Matching FIMO and SPAMO Outputs with MEME and DREME Outputs

[i...@alum.mit.edu]

I have run MEME-ChIP on some ChIP-seq datasets.  I noticed that the indexes in combined.meme do not correspond to the names of the fimo_out and spamo_out files.  There are also many fewer outputs from fimo and spamo than there are from MEME + DREME + CENTRIMO + TOMTOM.  From looking at the html file, it seems like, when motifs get clustered together, fimo and spamo are run for only one of them.  Is there some output from MEME-ChIP that I can use to connect the fimo and spamo output file names to the indexes in combined.meme?

Thanks so much!

[CharlesEGrant]

I noticed that the indexes in combined.meme do not correspond to the names of the fimo_out and spamo_out files.

Which version of MEME-ChIP are you running? Starting with MEME 4.11.4 MEME and DREME began including an 'alternate-id' derived from the consensus sequence for the motif, and it's index number. These alternate ids are included in the combined.meme file. For example,

MOTIF 1 WGGGTGTGGYYS-MEME-1

MOTIF 2 MCRCCCW-DREME-1

This indicates that MOTIF 1 in the combined.meme file was the first motif in the MEME output, and MOTIF 2 in the combined.meme file was the first motif in the DREME output file.

 From looking at the html file, it seems like, when motifs get clustered together, fimo and spamo are run for only one of them. 

This is correct. FIMO and SPAMO are only run using the most statistically significant motif in a cluster. If you would like FIMO and SPAMO to run on the full set of motifs in combined.meme, you'll have to run FIMO and SPAMO separately, using combined.meme as the motif input file.

[i...@alum.mit.edu]

Thanks for clarifying.  I switched to the latest version of MEME to take advantage of the alternate ids.