Identifying Functional Regions Of Proteins

Question

[Pals]

I am studying mammalian proteins and interested to analyze the surface conservation and functionally important residues. Presently, I am using Consurf but it generates redundant result as it depends on multiple sequence alignment (When I take all the sequences, important residues might be shadwed and if I take fewer seqeunces, majority of the residues are conserved). In addition, there is no any option to select only mammalian proteins in this program, because I want to ignoresequences from bacterial as well as amphibians and reptiles' . As my working set contains about 75 proteins I do not want to make alignments taking only relevant mamalian proteins and then fetch it to the server. I would like to know if there are better programs than consurf that would help my task easier and make my analysis more noise free. I would appreciate your suggestions and ideas.

[Pals]

Sorry, I meant manual selection of sequences. Because it is more time consuming in determining how many sequences to select and which ones to select. I just wanted to know if there are better programs than consurf or some ways to restrict the mammalian proteins in such programs.

[Jan Kosinski]

Why you think you need to do manual alignment? Can't you select proteins you want, align with MAFFT (http://mafft.cbrc.jp/alignment/server/) and submit to ConSurf?

Answer 1

[Larry_Parnell]

If you'd like to identify functionally important residues, I would use Pfam (longer motifs) and ELM or Prosite (shorter motifs). You search one protein query without the need for a multiple alignment. You can screen the results for motifs relevant to mammals.

Answer 2

[Michael Kuhn]

I'm not entirely sure that I understand your problem, but if you worry about mammals being overrepresented in your analysis, adding a phylogenetic tree seems to be a built-in option of ConSurf to overcome this bias. If you want to retain only mammalian sequences, you need to filter for that either manually or with the help of the NCBI taxonomy (assuming that you have taxon ids). If you need to do this only once, going through the 75 proteins and removing non-mammalian ones is probably the quickest.

[Pals]

In fact I will make analysis on 75 different proteins. But thanks for your suggestion of using phylogenetic tree.