Sequence Search

Search your sequence against the SUPERFAMILY hidden Markov models.

Paste sequences in FASTA format into the text box below, or upload a file, then click on 'Submit'. Up to 20 sequences can be submitted in the FASTA format. After submitting you will be shown the sequences which have been accepted by the server. Your submission will queue on the server, and then run for at least 5 minutes (time depending on your sequences). If you select e-mail notification you will be e-mailed when the run is complete. If you select browser notification, you must wait until the browser indicates that the search is 'done', and then click on the link to the output. The output page will remain active for 15 days. The output will give you a list of hits that your sequences make to models belonging to superfamilies, their alignments to the model, and assigned genome sequences.

After your sequences have been accepted by the server they are passed through a filter to find any domain assignments for similar sequences already stored in the database. If no hits are found in our pre-computed database, the sequences are scored against the model library. If this does not produce any significant hits then you are presented with the option of using PRC to score the sequences. PRC is an extremely sensitive profile-profile comparison method.

Large numbers of sequences can be accommodated by contacting superfamily@mrc-lmb.cam.ac.uk.