• DISpro is a web server to predict protein disordered regions using the 1D-Recursive Neural Networks and statistical methods. For the detailed descriptions of the method, please see the reference.

• Target name, optional, which is used to identify your query.
• Email address, where the prediction result is sent.
• Sequence: a raw text of sequence, white space are ignored.

• Disordered region predictions are presented in CASP format. Here is an example with descriptions:

         PFRMAT DR
         TARGET Query-Name 
         AUTHOR baldi-group-server 
         REMARK Predictor remarks
         METHOD Description of methods used
         METHOD Description of methods used
         METHOD Description of methods used
         MODEL  1
         H D 0.70           # <- Field 1: residue code,
         L D 0.80           # <- Field 2: order/disorder assignment code (D: disordered, O: ordered),
         E D 0.80           # <- Field 3: the number specifying the associated
         G D 0.60           #    confidence level: 0.5 - residue not predicted 
         S D 0.90           #                     >0.5 - disordered region 
         I O 0.50           #                     <0.5 - ordered region
         G O 0.40
         I O 0.40
         K O 0.30
         H O 0.20
         E O 0.20
         I O 0.40
         V O 0.45
         F D 0.60
         D D 0.90
         C D 0.80
         END

• On a data set of 723 protein chains, the precision and recall of disordered regions is 75.4% and 38.8% respectively. TPR (True Positive Rate) at 5% FPR (False Positive Rate) is 92.8%.

• J. Cheng, M. Sweredoski, P. Baldi, Accurate Prediction of Protein Disordered Regions by Mining Protein Structure Data, Data Mining and Knowledge Discovery, vol. 11, no. 3, pp. 213-222, 2005. [PDF]


• Download the dataset used in training DISpro here.
• Download DISpro 1.0 (Linux version).