Predicting small ligand binding sites in proteins using backbone structure

Andrew J. Bordner

Research output: Contribution to journalArticle

35 Citations (Scopus)

Abstract

Motivation: Specific non-covalent binding of metal ions and ligands, such as nucleotides and cofactors, is essential for the function of many proteins. Computational methods are useful for predicting the location of such binding sites when experimental information is lacking. Methods that use structural information, when available, are particularly promising since they can potentially identify non-contiguous binding motifs that cannot be found using only the amino acid sequence. Furthermore, a prediction method that can utilize low-resolution models is advantageous because high-resolution structures are available for only a relatively small fraction of proteins. Results: SitePredict is amachine learning-based method for predicting binding sites in protein structures for specific metal ions or small molecules. The method uses Random Forest classifiers trained on diverse residue-based site properties including spatial clustering of residue types and evolutionary conservation. SitePredict was tested by cross-validation on a set of known binding sites for six different metal ions and five different small molecules in a non-redundant set of protein-ligand complex structures. The prediction performance was good for all ligands considered, as reflected by AUC values of at least 0.8. Furthermore, a more realistic test on unbound structures showed only a slight decrease in the accuracy. The properties that contribute the most to the prediction accuracy of each ligand were also examined. Finally, examples of predicted binding sites in homology models and uncharacterized proteins are discussed.

Original languageEnglish (US)
Pages (from-to)2865-2871
Number of pages7
JournalBioinformatics
Volume24
Issue number24
DOIs
StatePublished - Dec 2008

Fingerprint

Binding sites
Backbone
Binding Sites
Ligands
Proteins
Protein
Metal ions
Metals
Ions
Protein Structure
Molecules
Spatial Clustering
Nucleotides
Cofactor
Computational methods
Random Forest
Prediction
Performance Prediction
Area Under Curve
Cluster Analysis

ASJC Scopus subject areas

  • Biochemistry
  • Molecular Biology
  • Computational Theory and Mathematics
  • Computer Science Applications
  • Computational Mathematics
  • Statistics and Probability

Cite this

Predicting small ligand binding sites in proteins using backbone structure. / Bordner, Andrew J.

In: Bioinformatics, Vol. 24, No. 24, 12.2008, p. 2865-2871.

Research output: Contribution to journalArticle

Bordner, Andrew J. / Predicting small ligand binding sites in proteins using backbone structure. In: Bioinformatics. 2008 ; Vol. 24, No. 24. pp. 2865-2871.
@article{3d3b5e62b0b14b47a8e7549411408e96,
title = "Predicting small ligand binding sites in proteins using backbone structure",
abstract = "Motivation: Specific non-covalent binding of metal ions and ligands, such as nucleotides and cofactors, is essential for the function of many proteins. Computational methods are useful for predicting the location of such binding sites when experimental information is lacking. Methods that use structural information, when available, are particularly promising since they can potentially identify non-contiguous binding motifs that cannot be found using only the amino acid sequence. Furthermore, a prediction method that can utilize low-resolution models is advantageous because high-resolution structures are available for only a relatively small fraction of proteins. Results: SitePredict is amachine learning-based method for predicting binding sites in protein structures for specific metal ions or small molecules. The method uses Random Forest classifiers trained on diverse residue-based site properties including spatial clustering of residue types and evolutionary conservation. SitePredict was tested by cross-validation on a set of known binding sites for six different metal ions and five different small molecules in a non-redundant set of protein-ligand complex structures. The prediction performance was good for all ligands considered, as reflected by AUC values of at least 0.8. Furthermore, a more realistic test on unbound structures showed only a slight decrease in the accuracy. The properties that contribute the most to the prediction accuracy of each ligand were also examined. Finally, examples of predicted binding sites in homology models and uncharacterized proteins are discussed.",
author = "Bordner, {Andrew J.}",
year = "2008",
month = "12",
doi = "10.1093/bioinformatics/btn543",
language = "English (US)",
volume = "24",
pages = "2865--2871",
journal = "Bioinformatics",
issn = "1367-4803",
publisher = "Oxford University Press",
number = "24",

}

TY - JOUR

T1 - Predicting small ligand binding sites in proteins using backbone structure

AU - Bordner, Andrew J.

PY - 2008/12

Y1 - 2008/12

N2 - Motivation: Specific non-covalent binding of metal ions and ligands, such as nucleotides and cofactors, is essential for the function of many proteins. Computational methods are useful for predicting the location of such binding sites when experimental information is lacking. Methods that use structural information, when available, are particularly promising since they can potentially identify non-contiguous binding motifs that cannot be found using only the amino acid sequence. Furthermore, a prediction method that can utilize low-resolution models is advantageous because high-resolution structures are available for only a relatively small fraction of proteins. Results: SitePredict is amachine learning-based method for predicting binding sites in protein structures for specific metal ions or small molecules. The method uses Random Forest classifiers trained on diverse residue-based site properties including spatial clustering of residue types and evolutionary conservation. SitePredict was tested by cross-validation on a set of known binding sites for six different metal ions and five different small molecules in a non-redundant set of protein-ligand complex structures. The prediction performance was good for all ligands considered, as reflected by AUC values of at least 0.8. Furthermore, a more realistic test on unbound structures showed only a slight decrease in the accuracy. The properties that contribute the most to the prediction accuracy of each ligand were also examined. Finally, examples of predicted binding sites in homology models and uncharacterized proteins are discussed.

AB - Motivation: Specific non-covalent binding of metal ions and ligands, such as nucleotides and cofactors, is essential for the function of many proteins. Computational methods are useful for predicting the location of such binding sites when experimental information is lacking. Methods that use structural information, when available, are particularly promising since they can potentially identify non-contiguous binding motifs that cannot be found using only the amino acid sequence. Furthermore, a prediction method that can utilize low-resolution models is advantageous because high-resolution structures are available for only a relatively small fraction of proteins. Results: SitePredict is amachine learning-based method for predicting binding sites in protein structures for specific metal ions or small molecules. The method uses Random Forest classifiers trained on diverse residue-based site properties including spatial clustering of residue types and evolutionary conservation. SitePredict was tested by cross-validation on a set of known binding sites for six different metal ions and five different small molecules in a non-redundant set of protein-ligand complex structures. The prediction performance was good for all ligands considered, as reflected by AUC values of at least 0.8. Furthermore, a more realistic test on unbound structures showed only a slight decrease in the accuracy. The properties that contribute the most to the prediction accuracy of each ligand were also examined. Finally, examples of predicted binding sites in homology models and uncharacterized proteins are discussed.

UR - http://www.scopus.com/inward/record.url?scp=57249116902&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=57249116902&partnerID=8YFLogxK

U2 - 10.1093/bioinformatics/btn543

DO - 10.1093/bioinformatics/btn543

M3 - Article

C2 - 18940825

AN - SCOPUS:57249116902

VL - 24

SP - 2865

EP - 2871

JO - Bioinformatics

JF - Bioinformatics

SN - 1367-4803

IS - 24

ER -