TABLE 2.

Troubleshooting table.

Step	Problem	Possible reason	Solution
23	Error report: ‘The sequences submitted for the input dataset appear to be in a format that MEME-ChIP does not recognize. Please check to be sure that your data is formatted properly.’	Make sure that you did not submit a Word document or a BED file or other invalid file format	Re-format your sequence file to ensure that it is in FASTA sequence format
	Confirmation email is not received	You might have entered your email address incorrectly or your email program may be filtering the MEME-ChIP email as spam	Check your folders designated as junk e-mail or spam. Try to configure your mail program not to filter out emails from http://meme.nbcr.net or http://meme.ebi.edu.au
24	The job stopped with a time-out error	Very large MEME-ChIP jobs may run out of time on the web server	Resubmit the job on an alternative server. Currently the hardware on the server at http://meme.ebi.edu.au is twice as fast as that at http://meme.nbcr.net Alternatively, run the job on your own computer after installing MEME Suite (Box 4)
24	In the ‘PROGRAMS’ section of the results page, getsize and MEME finished with warnings like: ‘Duplicate sequence name. XX’. In this case, you may notice that the sequence counts in the ‘INPUT FILES’ section is unusually small	The sequences in the FASTA file have duplicate FASTA IDs	Please make sure every sequence has a unique FASTA ID. (The FASTA ID is everything between the ‘>’ and the first white space on the FASTA header line.) The second (or later) duplicate sequence will be ignored by MEME-ChIP during motif discovery
	In the ‘PROGRAMS’ section of the results page, CentriMo finished with a warning like ‘Skipping sequence XX as its length (XX) does not match the expected length (XX)’	The sequences in the FASTA file do not have the same length	Please make sure every sequence has the same length. Otherwise, CentriMo will skip them
	When analyzing CLIP-seq data, the de novo motifs identified are weakly palindromic, and only half of each palindromic motif matches with a known RBP motif in the database	The program searched for motifs in both strands of the RNA sequences	Make sure you check the ‘scan given strand only’ box in Step 8
	MEME or DREME identified a motif similar to the known canonical motif of the binding factor, but Tomtom annotates it as some other similar or known motif	The canonical motif might not be included in the selected motif database	Please make sure that the appropriate motif database is selected in Step 4 Alternatively, make a customized motif data- base that includes the known canonical motif matrix and select that database in Step 4
	MEME-ChIP finished normally but the known primary known binding motif is not present in the results	The ChIP-seq or the CLIP-seq experiment failed; the peak calling program or the cross-linking site identification method did not perform as expected	There are many reasons that a ChIP-seq or CLIP-seq experiment could go wrong. To help identify the problem, check a few positive control (known TF- or RBP-binding) regions. In addition, check to ensure that the correct genome assembly was used to retrieve the sequences