Finding Common Motifs In Sequences

Question

[Fabio]

I have a few hundred yeast sequences (20-80bp long) and I want to find common motifs (conserved bases at certain indices) in them. I am using a Mac

Answer 1

[Stew]

I would recommend MEME as others, however Weeder is also very easy to install and run on a Mac. It is word based and runs a lot faster than Meme, with a far fewer options.

If you would like something to run specifically on your Mac you could try iMotifs, which incorporates the NestedMICA algorithm for discovering over-represented motifs.

For a small set of sequences, such as yours, you could also use Meme, Weeder or others online without installing them locally. Finally, the Regulatory Sequence Analysis Tools (RSAT) (http://rsat.ulb.ac.be/rsat/ or http://rsat.ccb.sickkids.ca/) are a great place to start for pattern matching and discovery.

In case it is of any use I have put up a short presentation about DNA motif finding from a course I gave earlier this year. You can find it here.

Answer 2

[Tom Koerber]

You can also use MEME: http://meme.sdsc.edu/.

Answer 3

[Marcos De Carvalho]

Some time ago I used SOMBRERO with a good degree of success on finding motifs in a very diverse set of sequences. They have a Mac version for download as well as parallel versions for Irix and Linux.

Answer 4

[Michael]

The first step when looking for conservation of single bases or motives is often a multiple sequence alignment that will align the sequences in a way such that conserved regions are best visible. This can be a first step before using explicit motif finders like MEME. A good way of visualizing multiple alignments is the sequence-logo that will give a graphical representation of base conservation.

Here is the wikipedia list of multiple-sequence alignment tools.

I recommend to start with the EBI web-server of ClustalW though, if that is not enough you can also try MAFFT or T-Coffee.

Weblogo can generate sequence-logo graphics from the output and also from fasta input directly.

Advantage of these tools is that you don't need to install them, so good for a first attempt irrespective of using a Mac.

Answer 5

[Giovanni M Dall'Olio]

Meme has been the first program to be published for doing that.

As an alternative you can find one of the EMBOSS tools; if you are scared by a terminal and want to do it from a web-based interface, you can use the EMBOSS tools from galaxy

Answer 6

[Darked89]

You may check out these pages:

• Bioinfo tutorial (Promoter prediction)
• Wikiomics:Sequence_motifs

These are ca 2 years old (links may not work etc.) but as a starting point should be OK. Also in unlikely case you did not found it yet: in yeast there has been an extensive motif search study done by Kellis with insane number of citations:

• Nature. 2003 May 15;423(6937):241-54.
• Kellis M, Patterson N, Endrizzi M, Birren B, Lander ES. Sequencing and comparison of yeast species to identify genes and regulatory elements

Answer 7

[Zhenhai Zhang]

Try this out? http://fraenkel.mit.edu/webmotifs/form.html

Answer 8

[Suk211]

ACGGGCCCGACGATGCGTCGTA
ACGTACGTCGAACCGTCGTCGT
ACGTGCGTCGAAACGTCAGTCG
ACGGGTTCGATCGTCGTCGTCG

may be in Python I will break down the first sequence of required motif length into a sliding window and will search for those list of motifs in the rest of sequences using regular expression in python using re.search() method.

[Istvan Albert]

post the python code as well (put it into pre tags then it will be shown nicely formatted, see help on the right)

Answer 9

[sanchezcavani]

I would recommend to use MEME.