Building A Consensus Sequence From A Set Of Sequences

Question

[Rishika Sengupta]

I have a list of around 50 pdb files/fasta sequences (they do not belong to any family). I badly need to build up 10-15 consensus sequences from them representing sets of PDB files. I have used the web servers ClustalW and Consensus could not really understand the results. Please help.

Rishika CARLBio group

Answer 1

[Pals]

I have once found the program PAGAN to be useful in this sort of cases. You could go through the Manuscript first. Also Codoncode aligner can address your problem.

[Rishika Sengupta]

Thanks but cant install PAGAN

[Pals]

Probably because it is targeted to debian based linux environment. It would be really difficult if not impossible if you are using other linux distributions.

[2184687-1231-83-]

You need libboost libraries installed: sudo apt-get install libboost-dev libboost-program-options1.42-dev libboost-regex1.42-dev

[Ari]

If you have problems installing PAGAN, please contact the author and ask for help. You can find the contact details on the program web site.

PAGAN requires two Boost packages (available for all platforms, including OSX and Windows) but shouldn't then compile fine.

[Ari]

If you have problems installing PAGAN, please contact the author and ask for help. You can find the contact details on the program web site. PAGAN requires two Boost packages (available for all platforms, including OSX and Windows) but should then compile fine.

Answer 2

[Neilfws]

If what you want to do is cluster the sequences into groups and choose a representative of each group, look no further than CD-HIT. It has a web server too, if you don't want to run it locally.

[Rishika Sengupta]

I have just now used the CD-HIT web server. But I want to input my sequences in fasta format, cluster the sequences into groups and choose a representative of each group, as you said correctly. Can I do that with the help of CD-HIT? Please reply. Thank you so much for your help.

[Neilfws]

Well yes, you can - as I said in the answer.

[Neilfws]

Well yes, you can - as I said in the answer. I've used the standalone program to do exactly that; I have not used the server, but assume it can do the same.

[Rishika Sengupta]

Thanks a tonne....could do that after installing the G++ compiler. :):)

Answer 3

[brentp]

After generating your clusters, look here for how to generate a consensus sequence.

Every time someone asks about consensus sequences, I recommend motility. It has a C++ and a python interface.

[Pals]

Will this program consider protein sequences also?

[Pals]

But will this program consider protein sequence also?

[Rishika Sengupta]

The CD-HIT worked out....

[Rishika Sengupta]

Thanks for the response.

[brentp]

@Kisun, good point. You'd have to convert to nucleotide sequence to use motility.

Answer 4

[Assa Yeroslaviz]

How about doing it with R?

the Biostring package is very helpfull and easy to understand. have a look at the vignette of Biostring