This page briefly describes the use of COUDES webserver. You can go to:
Principle
COUDES is a web-server that makes beta-turns prediction as well as their type. The basic principle is to predict first secondary structures with an accurate method such as PSIPRED (Jones, 1999), and to predict beta-turns within (almost) coil regions. The prediction within coil is done by using propensities weighted by scores coming from position specific scoring matrices (PSSMs) generated by PSI-BLAST (Altschul & al., 1997). For each predicted beta-turn, its type is then predicted using solely propensities.
For further details please see Fuchs & Alix (2005).
Quick Help
Predicting beta-turns through COUDES webserver is pretty easy. Go to the query page; it should look like the following:
The only thing you have to do is to paste your sequence in the large textarea (in plain or fasta format), and click on the button "Predict Turns!". All option values are set in order to get the best results in terms of well balanced under and over-prediction (see Fuchs & Alix (2004)).
For further details please see Fuchs & Alix (2005).
Detailed Help
Here is a brief explanation of each element in the Web-Page:
- Textarea: this is intended to paste your amino-acid sequence in plain or fasta format. All non natural amino-acids letter code are not taken into account, but might generate a warning. Numbers, spaces, tabulations are completely ignored.
- Choice of a propensities data set: this list allows the user to choose which propensities data set he wants to use. The first four letters indicate the beta-turn extraction method (PROM for PROMOTIF (Hutchinson and Thornton, 1996), EXTR for extract_turn), whereas the number indicates the number of protein chains used to calculate these propensities.
- PSI-BLAST database: this list allows the user to choose a database of sequences for generating the position specific scoring matrix (PSSM) with the program PSI-BLAST. It is possible to use NR (non redundant database), SWISSPROT or PDBAA (all sequences present in the PDB). This is recommanded to use the larger one, i.e. NR.
- Secondary structure prediction method: this list allows the user to choose a secondary structure prediction method used with COUDES (by default, COUDES predicts beta-turns only within coil regions). Sorry, for the time being, only PSIPRED is available.
- COUDES options:
- Score threshold: for each tetrapeptide, a score is evaluated using propensities and PSSMs. This threshold represents the minimum score from which a beta-turn is predicted.
- PSI-BLAST threshold: for each position of the sequence, each amino-acid of the multiple aligments possessing at least this PSI-BLAST threshold is taken into account for evaluating the score.
- Number of flanking residues: to predict turns, COUDES uses a sliding window all along the sequence. For each tetrapeptide, a score is evaluated. The number of flanking residues represents the number of residues taken into account around the tetrapeptide (on the right and on the left), e.g. if it is equal to 4, one considers: i-4 i-3 i-2 i-1 i i+1 i+2 i+3 i+4 i+5 i+6 i+7 (i i+1 i+2 i+3 is the central tetrapeptide on which the score is evaluated).
- Output format option: this allows the user to get a vertical, horizontal or both output(s).
For further details please see Fuchs & Alix (2005).
Note
May 2015: COUDES has been unavailable for a few months, sorry for the inconvenience. The service is now back and fully implemented within the Mobyle platform. The nr database of protein sequences that has been used to test COUDES in the original publication (DOI: 10.1002/prot.20461) is no longer avalaible. Thus you might get slightly different results compared to the original version published in 2005 (and so does for the psipred prediction). We now use uniref90 (version 2006-04-27), more recent versions containing many more sequences are not worth the extra computational cost.
References
- Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res., 25, 3389-3402.
- Fuchs PFJ, Alix AJP (2005). High accuracy prediction of beta-turns and their types using propensities and multiple alignments. Proteins, 59, 828-839.
- Hutchinson EG, Thornton JM. (1996). PROMOTIF--a program to identify and analyze structural motifs in proteins. Protein Sci., 5, 212-220.
- Jones DT. (1999). Protein secondary structure prediction based on position-specific scoring matrices. J. Mol. Biol., 292, 195-202.
Go to the query page
Go to the additional data page