WURST |
 |
This is the wurst server. It takes a protein sequence, performs sequence to structure alignments
to a library and then mails results back to you.
If you want to do structure comparisons, please go straight to our salami server.
What happens..
- Enter a title for the job or calculation
- Fill in your email address
- Type (paste) your sequence below
- Click submit
- Wait for results to be mailed to you
Your sequence will be aligned to about 21000 PDB template structures.
The alignments will be ranked and the results mailed back to you.
By default, the 50 top ranked proteins will be listed and the top 10 alignments printed out.
Models will be built for 5 proteins.
You can change those defaults in the boxes below.
Requirements and limitations
- your email address
- an identifying title
- your sequence must be more than 20 and less than 1000 amino acids
- the calculation will be run with default gap and insertion penalties
Explanation of fields
- N models to build says how many sets of coordinates should be calculated and mailed back.
- N attached files per email - Of the sets of coordinates built, send back this many per message. If you ask for 20 models and set this parameter to 4, then you will get 5 mail messages, each containing 4 sets of attached coordinates. Note If you set this to zero, then
the models will be calculated and sent back in the body of the mail message, not as
attachments.
- N proteins in hit list tells us how many proteins to list in the brief summary, with scores, but without alignments.
- N detailed alignments Detailed alignments will be printed out for this many
proteins.
- Sequence is the one letter code for the protein sequence.
- It must be between 20 and 1000 characters long.
- If the first line begins with a '>' character, it is treated as a comment (like FASTA format).
- White spaces are ignored.
- Only valid amino acids are accepted (no "x" or "z" characters)
More on fold recognition
In most ways, this is a classic protein fold recognition / protein threading server.
We have a score function (force field) based on an optimised classification
of small protein fragments. On top of that, we add a sequence to sequence alignment
score, but using an amino acid substitution matrix, optimised for alignment
quality. There is a slightly more expansive description of the server ideas
here.
Confidentiality
We may save your email address so we can send a message if things break (or better if they get fixed). Nothing else is stored, but do not rely on it. You should assume we will take your sequence, patent it and become rich.
A reference
Please cite this
paper:
Torda, A.E., Procter, J.B., and Huber, T.
Nucleic Acids Res., 32, W532-W535, 2004
"Wurst: A protein threading server with a structural scoring function, sequence
profiles and optimized substitution matrices."
Some of the philosophy is described in a
very old paper describing the predecessor of wurst.
For robots and other servers
Our server gives the following responses..
- 200 initially
- 202 when a sequence is accepted and the form is filled out properly and a calculation seems to have begun
- 412 if something is wrong (sequence too short / long ...)
- 500 for an internal server error
People
This server comes from the Torda group in
Hamburg, Germany and the Huber
group in Brisbane, Australia.Further Information |
