We strongly advise switching to the newer SAM-T02 server, which has been greatly outperforming SAM-T99 in both our internal tests and in the LiveBench on-going evaluation of protein-structure servers. We keep the SAM-T99 server around for comparison purposes, but will no longer be actively maintaining it. We plan to eliminate in the next year, as it takes up far more resources and performs worse than the newer server.
email address (required):
Subject for e-mail (optional)
Either cut and paste your sequence in the box below or upload it from a file on your local system. Sequences must be in a readseq-compatible format---we recommend FASTA format. You may also use a multiple alignment as a seed, which must be in a2m format.
Cut and paste your sequence here, ...
or specify the sequence-containing file on your system here.
Further options appear below.
This page offers the ability to perform remote protein homolog searches and protein structure prediction using HMM technology. It is part of a larger suite of UCSC's HMM applications that are available on the web. More details on the methods used can be found in an HMM tutorial, though that tutorial talks about the predecessor SAM-T98 method, which differs in several parameter choices. A simpler web page that just searches our HMM library with the query sequence and doesn't build a multiple alignment, is the SAM-T99 model-library web page, which may provide more limited results much faster.
The PDB sequence database is updated weekly, but the template libraries are only updated sporadically.
You can check the SAM-T99 server queue status here . (Local users at UCSC check here .)
Please read the explanation of common errors on our SAM-T99 Frequently-asked Questions page. If you do not find a solution to your problem there, please inform us at email@example.com.
return T99 alignment in format(s): A2M Pretty Align HTML Sequence Logos
Search PDB (sequences updated weekly)
return secondary structure prediction in format(s): CASP RDB FASTA
(Last Update: )