2
$\begingroup$

I am performing a BLAST search and I need to filter the output to a certain family of proteins.

Specifically, I need to get matches within the CYP152 family of the Cytochrome P450 proteins. I can parse the results to just P450 proteins fairly easily as the "hit def" field in the output will contain "Cytochrome P450" somewhere in there if it is in fact a P450 protein sequence, but the same is not true for CYP152.

Is there some way to do this? I am using ExPASy BLAST:

http://web.expasy.org/blast/

Thanks in advance

$\endgroup$
  • $\begingroup$ Do you have a list of protein names in the family? $\endgroup$ – terdon Jun 1 '16 at 17:55
  • $\begingroup$ Can't you just use taxonomy filter?, but I'm sure @terdon can help you out :) $\endgroup$ – KingBoomie Jun 1 '16 at 20:28
  • $\begingroup$ @terdon I do not, that would make my life much easier. And Rick Beeloo the taxonomy filter is not that specific, at least as far as I have found. $\endgroup$ – Jason Drain Jun 1 '16 at 20:40
  • 1
    $\begingroup$ You're basically doing it backwards. You should start by getting the sequences of the proteins in the family and then blast your query against them. Are you looking for hits against any species? Only some? Do you have access to a Linux or Unix (OSX will do) machine? $\endgroup$ – terdon Jun 1 '16 at 22:00
  • $\begingroup$ I run Ubuntu on my laptop. And okay, are there any good sources of information that you can recommend to me? I have been searching around and can find lots of info on the P450s but not much on the CYP152 subfamily. What I have is a sequence from one protein in the CYP152 subfamily (called OleT JE) and I am trying to compare it with all CYP152 proteins. $\endgroup$ – Jason Drain Jun 3 '16 at 16:05
1
$\begingroup$

The gene identifiers that you get in BLAST output depends on the identifiers used during the BLAST database construction. So you may not find the exact identifier you are looking for.

For restricting the BLAST results to a smaller set of proteins/nucleic acid you can choose either of the two:

  1. Make a database with the fasta sequences of the proteins/nucleic acids that you want. You have to install BLAST and run it from the commandline. Use makeblastdb to make the database. Now you align your queries against this database. Note that now your e-values would not be the same as what you would obtain from a search against the bigger database.
  2. As you said in the question, you can parse the results later. You would have to know what identifiers were used in the blast database that you searched against. ExPASy-BLAST would most probably use the UniProt IDs (there is an ID convertor in the UniProt website to map UniProt IDs to other IDs). If you use NCBI-BLAST (nr database) then you would need to know the GenBank IDs.
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.