Tool/Script To Get Consensus Sequences

Question

[upendrakumar.devisetty]

I have three fasta files and basically all i want is to get is the consensus or union of three fasta files. I mean, i want to pull out all the sequences that are common among three files and put them in a separate fasta file. The three fasta files were generated differently and have different headers but the sequences are identical among three files. Is there a tool or script that does this?

[Neilfws]

If there is no existing tool/script, this is a fairly basic bioinformatics programming task which you should learn how to do.

[upendrakumar.devisetty]

I do understand it is basic bioinformatics programming but i thought if there is something already written then i can just use it rather than trying to write by myself

[Damian Kao]

How large are the fasta files?

[upendrakumar.devisetty]

The files are not more than 100MB

[Michael]

search for "multiple sequence alignment" in a search engine of your choice or encyclopedia to get a starting point.

[Michael]

Your question title and use of the term "consensus sequence" is misleading, it is reserved for alignments; it seems that you are looking to find exactly identical sequences instead. Please take at least one quarter of the time it will take us to answer your question, to put a proper example, and to explain what you already have tried.

Answer 1

[upendrakumar.devisetty]

Sorry if i didn't explained my question well. Here is what i wanted...

I have three fasta files

fasta1

seq1 ATGATG seq2 GATAGATA seq3 TGGTGG

fasta2

m1 GATAGATA m2 TGGTGG m3 AGGAGG

fasta3

seq1 gene1 ATGATG seq3 gene3 TGGTGG seq4 gene4 AGTGTG

And what i wanted is basically a final fasta file containing the below

final_fasta

seq3 TGGTGG

As you can see this sequence is represented in all fasta files but the header i got is from fasa1

I am currently using a lengthy process to achieve this.

First blast fasta1 to fasta2 and take all the hits and use those hits to blasta fasta3.

Thought it should work it is lengthy and i though if somebody has done something like this before i can use their script/tools

Answer 2

[Damian Kao]

use CDHIT: http://weizhong-lab.ucsd.edu/cd-hit/