Meme is a program that facilitates the discovery of motifs in nucleic acid and protein sequences. You can run Meme off the web from Meme-Central (University of California at San Diego), from our own Watson server, or from the Hercules server. There are pros and cons to each choice.

In brief:

• Meme from the UCSD site (<-- click for more instructions)

Good: Easy interface to use (once you get there)
Good: Graphical output (HTML) is easy to understand
Bad: Awful turnaround. Don't expect output until the next day
 
• Meme on the Watson server (<-- click for more instructions)

Good: Output returned to you in seconds
Bad: Will not run FastA files. GCG requires you first to convert the file
Bad: Output is pure text and can be confusing
 
• Meme on the Hercules server (<-- click for more instructions)

Good: Output returned to you in seconds
Good: Default Graphical output (HTML) is easy to understand
Good: Uses a single FastA file as the source of sequences
Bad: User interface not as nice and interactive as with other options
Bad: You don't have a Hercules accounts, do you?

Running Meme from the UCSD site

• Go to Meme site (http://meme.sdsc.edu/meme/website/)
• Read the Introduction, Overview, etc, if you like, but eventually get to the Submission form link
• Fill in your e-mail address (output will be mailed to you)
• Either paste in or upload the FastA-formatted file
• Click the number of motifs you want to find per sequence (lower left on screen)
• Enter the number of motifs you want to find (lower left on screen)
• You probably don't want to fool with the other adjustable parameters
• Click START SEARCH
• Wait a long time (probably overnight)
• Download the file to your computer and read it with your favorite web browser

Running Meme from the Watson server

• Log in on Watson
• Upload the file
• Run:      fromfasta name-of-file
  This converts the sequences within the FastA-formatted file into GCG-formatted files. 
  Each sequence is put in a separate file with a pep extension (if a protein) or seq extension (if DNA)
• Run:      meme
• In response to:   Find motifs in what sequences? 
  ... you have a problem. If the uploaded sequences are the only ones in your directory, then you can easily respond *.pep or *.seq to induce meme to consider all of them. If you have other protein or DNA sequences in your directory, then you'll either have to type in (or copy/paste) the names of each sequence you want analyzed (ugh!) or play other games to be able to refer to them and only them.
• When asked, specify how many motifs you want Meme to find. Difficult to know a priori what's best.
• When asked for the name of the profile file, accept the default. You'll never read this file anyway.
• When asked for the name of the report file, give it a descriptive name you'll remember later. This is the important file.
• When Meme is finished (may take a minute or so) read or download the file.

Running Meme from the Hercules server

Log in on Hercules

Upload the file

Type in the line similar to the one below:
     meme name-of-file -dna -mod zoops -nmotifs 6 > name-of-output-file.html

Notes on above: 
   * Use -dna if using DNA sequences or -protein if using protein sequences
   * Use -mod oops if you want Meme to find precisely one motif-hit per sequence
            -mod zoops if you want Meme to find either zer or one motif-hit per sequence
            -mod tcm if any number will do
   * Type after -nmotifs the number of motifs you want Meme to find

You won't be able to read the output file within Hercules. Instead, download it to your own computer and read it with your favorite Web browser

Contact me
  Jeff Elhai
  Life Sciences Building, Room 335
  828-0794
  ElhaiJ@VCU.Edu