Skip to main content


Table 1 Description of disorder predictors analyzed in this work

From: MetaDisorder: a meta-server for the prediction of intrinsic disorder in proteins

Method Short description Availability Ref.
DisEMBL ANN trained to predict classic loops (DSSP), flexible loops with high B-factors, missing coordinates in X-ray structures, regions of low-complexity and prone to aggregation. local installation [21]
DISOPRED2 SVM trained to predict residues with missing coordinates. local installation [22]
DISpro Recursive neural networks (RNNs) trained to predict missing coordinates. local installation [23]
GlobPlot A simple method based on several hydrophobicity scales to predict regions of missing coordinates and loops with high B-factors. local installation [24]
iPDA Incorporates information about sequence conservation, predicted secondary structure, sequence complexity and hydrophobic clusters. web service [25]
IUPred Estimates pairwise interaction energies using a statistical potential. Two versions for predicting long and short disorder. web service [26]
Pdisorder Combination of neural network, linear discriminant function and acute smoothing procedure is used for recognition of disordered and ordered regions in proteins. web service [27]
Poodle-s SVM trained for short disorder detection (uses PSSMs generated by PSI-BLAST). web service [28]
Poodle-l Predicts long disorder using an SVM. web service [29]
PrDOS Predicts missing coordinates in 3D structure using SVM and PSSMs from PSI-BLAST. web service [30]
Spritz Predicts long and short disorder (missing coordinates) using two separate SVMs. Utilizes secondary structure. web service [31]
RONN Predicts missing coordinates using an ANN. local installation [33]