Dear all
I have the protein sequences encoded by a gene from a few organisms. I make an example below:
species1 AHLLAHLAHHLALJALHA
# ^----^
species2 AHLHAHLAHHHJJJAHHJ
species3 ALLLLLLAHLLHLJALHA
# ^----^
If performing CLUSTALX, species1 and species2 are group together, because of high overall similarity, however, I would like to find the motifs that could group species1 and species3, for example, the terminal LJALHA motif.
I appreciate any of your suggestions. THANKS!
There are very short sequences that are common between different combinations (Highlighted below). So, you need to first make a set of rules that defines a motif. Then based on that anybody can suggest you a tool or provide a short script to do that.
species1 AHLLAHLAHHLALJALHA
# ^---^ ^----^
species2 AHLHAHLAHHHJJJAHHJ
# ^---^
species3 ALLLLLLAHLLHLJALHA
# ^----^
There are motif discovery programs (e.g MEME) and python libraries (e.g Motility) that can calculate Position weighted matrices or motifs, but you need to provide more information on your problem.
•
written
by
GouthamAtla
Hi, Geeky, thanks. Actually my analysis focuses on one protein. I found that if doing CLUSTALX, species1 and 2 are grouped together. However, this is not true and is different from taxonomic classifications. I need to find motifs (for example, larger than 7 amino acids) that can group species1 and 3, and separate species2. This is my general aim.
I will try MEME. If you have any ideas, please feel free to comment. THANKS.
You can supply ClustalX with the real tree so it should highlight the similarities between them.
I think that defining a motif for 2 sequences is basically finding a long stretch of AA that are similar (or very close), you can do it by simply align them (using needle of EMBOSS for instance) and locating stretches of 7 AA that are similar.
Log in to answer this question.
