BioTech: Searching Protein Sequence Databases

Here is an example of how to search the Swiss-Prot database from start to finish. We suggest that, if possible, you open another window in your web browser so that you can follow along with the instructions while you are doing the search.

First we need a DNA sequence. Let's use this sequence:

	ttctaactgc aacctttcga agcctttgct ctggcacaac aggtagtagg cgacactgtt
	cgtgttgtca acatgaccaa caagtgtctc ctccaaattg ctctcctgtt gtgcttctcc
	actacagctc tttcc

Now we need to translate the nucleotide sequence into a protein sequence. There is a handy tool on the WWW for automatic translation of nucleotide sequences in to amino acid sequences. Copy the DNA sequence above and we will use this tool to translate that sequence.
You are offered a choice of translation tables. For now, we'll just use "Standard." Later, when you have a sequence from an organism that you've been working with, feel free to choose another. Now paste the DNA sequence into the input box. And then click on "Generate Protein."
The output will now show your translated and then transcribed protein as your output. Select and copy your translated protein sequence.
Now go back to the main Tools page on the Swiss-Prot server and click on Blitz. Blitz is a search tool that allows you to search the databases for protein sequence.

(Note: you may also use BLAST for this. However, information about Blast searching is already available on our nucleic acid sequence page.)

(Also note: Blitz searching requires access to an email account. If you do not have access to an email account, skip down to our instructions for performing a BioSCAN similarity search with your protein sequence.)
Once you are in the Blitz entry screen, you will be asked to submit your email address and a title for your search (the title is for your own record keeping). You should end up with 2 email messages, the first being a confirmation of your job's submission and the second being the output of your job. Your output will contain dozens of proteins that resemble the sequence that you've given. At the bottom of your output will be the name and accession number of the protein that most closely resembles the sequence that you've submitted.

Congratulations! Your most closely resembled output should read:

RESULT    1     Score 104;  Match 0.0%;  Predicted No. 5.55e-17;

ID   INB_HUMAN      STANDARD;      PRT;   187 AA.

DE   INTERFERON BETA PRECURSOR (FIBROBLAST).

(Note: there will be many results, but the first result is the most closely-related result.)

BioSCAN
Alternate instructions for similarity searches using BioSCAN: Go to the BioSCAN input page and paste your copied protein sequence in the input area. All of the other settings should default to normal (which is all we need for now), so now choose "Submit Request."
Congratulations! Your results should read:
```
Db|Acc|Name            Description                               

sw|P01574|INB_HUMAN    INTERFERON BETA PRECURSOR (FIBROBLAST)... 
```
This means that your sequence most closely matched this precursor sequence in humans. All of the entries in BioSCAN are generated in hypertext, so you can easily find more information about your results.

Searching PIR

The most difficult part of searching the PIR database is finding the search engine on the WWW. Although many sites have been put together to search multiple databases at once, very few concentrate specifically on PIR. As a matter of fact, there is not even a link to the WWW search site from the PIR home page.

Because PIR is really a series of databases covering different topics and differing amounts of information, the PIR WWW search engine allows the user to select which database to search. However, because the engine is formatted to be an specific information retrieval site, you cannot search on a protein sequence. Rather, you may search the database to retrieve protein sequences of specific characteristics that you determine.

PIR is a good starting point to find information about sequence availability and detection. It is a very thoroughly cross-linked report tool that will help you get to more detailed information that you may be seeking. However, if you need to identify a protein sequence that you do not know the history of, you should try the search engines that are affiliated with Swiss-Prot.

To learn about how to search PIR, visit PIR Search Help. This is an excellent guide to searching PIR.

Return to the Protein Sequence Databases page.

Last updated: July 23, 1998
URL: http://biotech.icmb.utexas.edu/pages/science/protseq_search.html
Questions or Comments: feedback@biotech.icmb.utexas.edu
BioTech Resources Web Project
Indiana Institute for Molecular and Cellular Biology
©, 1995-1998 The Trustees of Indiana University