Skip to main content

Table 1 Summary of detailed characteristics of the six available web servers for DNA-binding sites prediction.

From: MetaDBSite: a meta approach to improve protein DNA-binding sites prediction

  Machine learning methods Properties used in training Online website
DISIS Support Vector Machine (SVM)
Neural network
Evolutionary profile
Conservation
Predicted secondary structure
Predicted solvent accessibility
http://cubic.bioc.columbia.edu/services/disis
DNABindR Naïve Bayes classifier Relative solvent accessibility
Sequence entropy
Secondary structure
Electrostatic potential
Hydrophobicity
http://turing.cs.iastate.edu/PredDNA/predict.html
BindN SVM The side chain pKa value
Hydrophobicity index
Molecular mass
http://bioinfo.ggc.org/bindn/
BindN-rf Random forest The side chain pKa value
Hydrophobicity index
Molecular mass
Blast-based conservation
Biochemical feature
Position-specific scoring matrix (PSSM)
http://bioinfo.ggc.org/bindn-rf/
DP-Bind SVM
Kernel logistic regression
Penalized logistic regression
Sequence-based BLOSUM62
PSSM-based
http://lcg.rit.albany.edu/dp-bind/
DBS-PRED Neural network Protein sequence information
Solvent accessibility
Secondary structure
http://www.netasa.org/dbs-pred