Finding Signal Peptide Sequences

Hi all,

I have to get signal peptide sequence of the extracellular enzymes genes from few of the Bacillus spps.

Can anyone please let me know, how i should go about?



I m using NCBI for finding out that... but when i search i get whole genomes sequences. & its very difficult to find out signal peptide sequence from it.

Also, for some CDS entries the signal peptide seq. is mentioned separately whereas some dont separately mention it ...



Is there any other way by which it can be done more easily?

Please help...

Hi.

There are several tools available for signal peptide prediction.

http://www.cbs.dtu.de/services/SignalP-2.0/

http://www.cbs.dtu.dk/services/SignalP/

http://www.predisi.de

http://nar.oxfordjournals.org/cgi/content/full/32/suppl_2/W375

You may also search for your protein in the Signal peptide database of 175542 proteins.

Good luck!

Thanks a lot Varsha..

I still have a small query..

Say suppose we have to find all the existing signal peptide sequences in a particular genome e.g in B. subtilis 168 strain.. then how we should go about doing that?

We might have to run the whole genome sequences in those tools.. but how will i get whole genome genes in its amino acid composition... do i have to select indiviual gene on whole genome page & convert it to protein sequence & then search whether it has signal peptide or not?

Is there any other non-labourious method for doing that?

Please help..

Regards,

Hi Vinbio,

You can download all RefSeq proteins for B. subitlis str 168 by visting this page:

http://www.ncbi.nlm.nih.gov/sites/entrez?db=genomeprj&cmd=Retrieve&dopt=Overview&list_uids=76

and then clicking on the 'Project Data' icon on the top right corner of the page.  Afterwards, you could run them through some of the programs that were suggested by Varsha to see which ones are predicted to be secreted.  Check out the bioinformatics section of this paper for ideas:  http://genomebiology.com/2008/9/2/R35

However, at the end of the day, bioinformatic predictions are one thing, but eventually, you'll need some lab validation experiments...

Good luck.

Thanks a lot Sichan..

the project data gave 8282 items.. so  are these all of the proteins present in that strain??

so that we can process all of them for presence of signal peptides?

Awaiting your reply,

Regards....

Hi VinBio,

Yes, but you might want to use the 4105 RefSeq proteins because some of those have undergone manual curation and will be more reliable.  In the drop down box next to 'Display', you can select 'FASTa'.  Next, in the 'Show' drop down box, you can select '500'.  In my experience, this will return all entries, not just the 500.  Then you can click on 'Send to' and send the 4105 protein sequences to a fasta file on your local machine.  From there, you can start predicting which ones are secreted and which ones are not.

Good luck.

Hey thanks Sichan..

i'll try that surely..

have a great day...

regards