Minor Size Differences Leading To Major Motifs Differences

Question

[Karen]

Hi! I've obtained some motifs from artificial DNA sequences but I'm not sure if I'm doing something wrong because when I change the motifs size just in one position the generated logos change dramatically as in this picture I posted in the link.

When I just change motifs size from 15 to 16 I get completely different logos.. I'm new to this type of analysis, I'm finding these motifs with the cosmo package http://bioconductor.wustl.edu/bioc/html/cosmo.html that works like MEME but it's an R package and seems to work fine but I'm lost here when I get this results

http://tinypic.com/view.php?pic=2wez51g&s=6

Thanks for any help

Karen

[Josh Herr]

Could you provide us with more information on what your research question is and what you are trying to accomplish? I'm unclear as to why you should be getting such differences, but I don't have any context with which to understand. Thanks.

[Karen]

Yes, I'm studing 50 sequences of letters (A, C and T) where I'm trying to find patters of different sizes. Each sequence length ranges from 100 letters to 500 and all the sequences lengths are different. I'm trying to find if there are subsequences of different lengths (5,10,15,20) that are repeated or slightly different and represent them using sequence logos.

Answer 1

[Farhat]

If you reverse complement the motif from 16 it looks quite similar to the one from 15.

[Karen]

Thanks Farhat, if I understand it right this means that if in size 16 I replace A/T it results in 15 (or also in 17) so all three motifs are quite similar.

But this only works if I'm dealing with DNA sequences but in my problem A, C and T have a different meaning, they are not nucleotides so A and T have different meanings and I cannot replace one for another..

Does a situation like mine happen when working with motifs? I mean if in real DNA sequences does it happen that when the lenght of the motif changes it also changes the identified motifs? I can't find any example of a similar situation, intuitively I should think that the identified motifs should not change that much but I don't know..

[Farhat]

If you are using MEME there is an option to turn off reverse complimented motifs. There may be an option to do that there as well. Another option may be asking it to treat it as an amino acid sequence in which case there is no complimenting.