HMM-based Protein Structure Prediction, SAM-T08  TESTING Page

Can't determine if this is the Test or Live server. Assuming Live. Please tell the administrator.

HMM-based Protein Structure Prediction, SAM-T08

TEST SERVER: Warning: this is a TEST page that is under construction. Don't try to use it!

Test Server TO-DO list (shouldn't show up on the real server):

Put PARENT field into TS files for alignment-based models? At least add explanation to METHODS line about the first model versus models 2-5.
averaging of fold-recognition results needs to be reweighted, and better selection of alphabets made.
target09 needs to be written and template library made (replacing t2k or t04)
Add "first look" access to molecular viewing.
Arrange for animated bar to be replaced by "crash" or "finished" icon when job terminates.
Arrange for reverse-order printing of .html file, by using server-side includes and appending to different subsection files.

In summer 2014, the server failed because the scratch disks on the old cluster ran out of space for hold the non-redundant protein database. New hardware with more (and faster) scratch space should be installed in Fall 2014.

This server was tested as SAM_T08_server in the CASP8 experiment. Preliminary results indicate that it is a good primary server, but that metaservers combining results from several primary servers are likely to produce somewhat better results.

This server finds similar protein sequences in NR and aligns them, providing sequence logos that show relative conservation of different positions. Local structure predictions are done with neural nets for several different local structure alphabets, and hidden Markov models are created. Fold recognition and alignment to proteins in the Protein Data Bank are done, and a full three-dimensional model is constructed. Look carefully at the E-values for the best hits, as full 3D models are constructed even when the E-values are high (and the resulting models are not then not trustworthy).

Email Address (required):

Subject line:

Email the secondary structure predictions (CASP SS format). Subject line:

Email the residue-residue contact predictions (CASP RR format). Subject line:

Email the top 5 models (CASP TS format). Subject line:

Either cut and paste your FASTA-formatted sequence directly in the box below:
>T0437 SfR90 , Shigella flexneri, 99 residues (see PDB 2k3iA) MKDVVDKCSTKGCAIDIGTVIDNDNCTSKFSRFFATREEAESFMTKLKELAAAASSADEG ASVAYKIKDLEGQVELDAAFTFSCQAEMIIFELSLRSLA

...or else upload it from a file on your local system, using the button below:

Sequences should be no longer than amino acids, to avoid overloading the server.

The amino acids should all be in UPPER CASE, but the server currently fixes the case for you if you forget.

The first character of the first line must be ">" (the "greater than" symbol).

The sequence ID must immediately follow the ">" (e.g. ">test ABC" is OK, but "> test ABC" with a leading space is not).

SAM-T08 will only process one sequence at a time.

You can find an explanation of common errors on our SAM-T08 frequently-asked questions page (FAQ). If you do not find a solution to your problem there, please inform us at sam-info@cse.ucsc.edu.

If you checked any of the "email me" boxes in Step 1, you will receive an email when your query is complete. You can also just wait for the resulting page to load (this could take a while, depending on how many other requests we are currently processing).

This server has evolved over the years under the direction of Kevin Karplus. Major contributors to the web design include Alex Williams, Rachel Karchin, Spencer Tu, and Christian Barrett.

This document was last modified on .

`Email Address (required):`
`Subject line:`
`Email the secondary structure predictions (CASP SS format).`	`Subject line:`
`Email the residue-residue contact predictions (CASP RR format).`	`Subject line:`
`Email the top 5 models (CASP TS format).`	`Subject line:`