how to generate meme index file without installing the MEME Suite Web Application

[Mathilde Perez]

Hi,

I am willing to use some of your motif databases for an external tool called BagFoot (https://www.ncbi.nlm.nih.gov/pubmed/28538187).

I already managed to get the tool to work and now I would like to use the latest motif database available.

To do so I need the meme file you provide (i.e. jolma2013.meme), I already have installed the motif databases on our linux server and the latest  MEME suite: meme-5.0.2 .

But in order to run BagFoot, I would also need  the related index files (i.e.:  jolma2013_index.csv), in the following format:

ID,Name,Species,Description1,Description2,SRC

M00002,E47,Homo sapiens  ,E47,T00207 E47; Species: human,Transfac

M00005,AP-4,Homo sapiens  ,activator protein 4,T00036 AP-4; Species: human,Transfac

M00006,MEF-2,Homo sapiens  ,myogenic enhancer factor 2,T00505 aMEF-2; Species: mouse,Transfac

M00007,Elk-1,Homo sapiens  ,Elk-1,T00250 Elk-1; Species: human,Transfac

M00008,Sp1,Homo sapiens  ,stimulating protein 1,T00759 Sp1; Species: human,Transfac

M00017,ATF,Homo sapiens  ,activating transcription factor,T00051 ATF; Species: human,Transfac

M00024,E2F,Homo sapiens  ,E2F,T00221 E2F; Species: human,Transfac

M00025,Elk-1,Homo sapiens  ,Elk-1,T00250 Elk-1; Species: human,Transfac

....

How can I generate this file?

In your documentation, there is this 'update-sequence-db' tool that should generate it, but I dont have this tool install in the MEME suite (I dont want/need the MEME Suite Web Application),

update-sequence-db --csv /motif_databases

Could you help me? 

Thanks a lot in advance.

Mathilde

[cegrant]

The update-sequence-db program is used to generate the DNA and protein sequence databases for a MEME Suite web site, not the motif databases, so it wouldn't be useful to you.

We provide the motif databases as a courtesy to the users of the MEME Suite public web site. We don't curate them ourselves, we just take publicly available data sets and convert them into the MEME motif format. We don't generate any index files for the motifs.  The jolma2013_index.csv file you mentioned is apparently something peculiar to BagFoot. So you might ask them about it. We don't have any connection to the BagFoot developers. 

If you look in the top directory of the MEME Suite motif databases you'll will find a file named 'motif_db.csv'. This describes the source motif databases.You could write a script that would read in that file, read in the  jolma2013.meme, and use the combination of the two to generate  the jolma2013_index.csv file. The format of the MEME motif file is described here. Helping  you write such a script is beyond the scope of help we can provide. You should contact your local bioinformatics resources for assistance.

[Mathilde Perez]

Thanks for the quick reply!

In the same time I figured out that indeed this file is need for Bagfoot but not coming from MEME suite.

I will make my own script to produce it.

Thanks again