 | Author: Bob MacCallum, Stockholm Bioinformatics Center, Stockholm University, Sweden. |
DRIPPRED
Proteins don't always look like the pictures shown on the covers of journals. Most proteins have one or more residues that cannot be pinned down by X-ray crystallography or NMR. These regions are often functionally important and shouldn't just be thought of as "junk" or "linker" polypeptides, particularly in eukaryotic proteins. The prediction of these disordered regions can be useful for structural biologists who may have difficulties expressing or crystallising proteins. For molecular biologists, predicted disordered regions may help in the dissection of functional domains by mutagenesis.
| Please follow these links to use our web based predictor for disordered regions in proteins, called DRIP-PRED. |
The predictor was ranked somewhere between 2nd and 4th in the CASP6 evaluation (group number 347), and self-assessment shows approximately equal performance to David Jones' DISOPRED (but is non-overlapping in terms of per-residue predictions). DRIP-PRED is available free of charge for protein-by-protein research purposes (no automated requests allowed).
NEWS: service moved to a smaller compute server on 26 Aug 2008. No automated requests please!
Old news: protein sequence databases (for PSI-BLAST profile generation) were updated on 17 Feb 2006
If you find DRIP-PRED useful, please make a link to it on your site. A journal article may appear during 2006, meanwhile please see this online paper for more details about the approach.
The author of the DRIP-PRED disorder predictor is Bob MacCallum.