Limits...
Predictions of hot spot residues at protein-protein interfaces using support vector machines.

Lise S, Buchan D, Pontil M, Jones DT - PLoS ONE (2011)

Bottom Line: These are the amino acid types on which the original model did not perform well.We have therefore developed two additional SVM classifiers, specifically optimised for these cases.HSPred reaches an overall precision and recall respectively of 61% and 69%, which roughly corresponds to a 10% improvement.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science, University College London, London, United Kingdom.

ABSTRACT
Protein-protein interactions are critically dependent on just a few 'hot spot' residues at the interface. Hot spots make a dominant contribution to the free energy of binding and they can disrupt the interaction if mutated to alanine. Here, we present HSPred, a support vector machine(SVM)-based method to predict hot spot residues, given the structure of a complex. HSPred represents an improvement over a previously described approach (Lise et al, BMC Bioinformatics 2009, 10:365). It achieves higher accuracy by treating separately predictions involving either an arginine or a glutamic acid residue. These are the amino acid types on which the original model did not perform well. We have therefore developed two additional SVM classifiers, specifically optimised for these cases. HSPred reaches an overall precision and recall respectively of 61% and 69%, which roughly corresponds to a 10% improvement. An implementation of the described method is available as a web server at http://bioinf.cs.ucl.ac.uk/hspred. It is free to non-commercial users.

Show MeSH

Related in: MedlinePlus

Sample output for the HSPred server.Screenshot of the results page for the IL-4/IL-4R complex (PDB code: 1IAR). On top, predictions are visualised using a Jmol applet. On the left is IL-4 (chain A), on the right IL-4R (chain B). Predicted hot spots are in red, non hot spots in white. Residues not part of the interface are in blue. Below, predictions scores for each interface residues (excluding Pro and Gly amino acids) are reported (note that only the first few residues are displayed here). Scores greater than zero corresponds to predicted hot spots.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3046169&req=5

pone-0016774-g004: Sample output for the HSPred server.Screenshot of the results page for the IL-4/IL-4R complex (PDB code: 1IAR). On top, predictions are visualised using a Jmol applet. On the left is IL-4 (chain A), on the right IL-4R (chain B). Predicted hot spots are in red, non hot spots in white. Residues not part of the interface are in blue. Below, predictions scores for each interface residues (excluding Pro and Gly amino acids) are reported (note that only the first few residues are displayed here). Scores greater than zero corresponds to predicted hot spots.

Mentions: We have implemented HSPred as a fully automatic web server, available at http://bioinf.cs.ucl.ac.uk/hspred. As input it requires a PDB formatted file containing the structure of the protein-protein complex. The user needs to define the interface to analyse by specifying the chain identifiers for each protein on either side of the interface. The output consists of two components: (i) a Jmol applet to visualise and explore the predictions using the protein structures and (ii) a table listing HSPred scores for each interface amino acid. The output page for an illustrative example is reported in Figure 4. The complex tested is Interleukin 4 (IL-4) bound to its receptor chain (IL-4R) (PDB code: 1IAR). Alanine mutational data from experiments are available for this complex [20], [21]. Out of interface mutations, HSPred predicts true positives, true negatives, false positive and false negatives. These results further validate the predictive accuracy of HSPred.


Predictions of hot spot residues at protein-protein interfaces using support vector machines.

Lise S, Buchan D, Pontil M, Jones DT - PLoS ONE (2011)

Sample output for the HSPred server.Screenshot of the results page for the IL-4/IL-4R complex (PDB code: 1IAR). On top, predictions are visualised using a Jmol applet. On the left is IL-4 (chain A), on the right IL-4R (chain B). Predicted hot spots are in red, non hot spots in white. Residues not part of the interface are in blue. Below, predictions scores for each interface residues (excluding Pro and Gly amino acids) are reported (note that only the first few residues are displayed here). Scores greater than zero corresponds to predicted hot spots.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3046169&req=5

pone-0016774-g004: Sample output for the HSPred server.Screenshot of the results page for the IL-4/IL-4R complex (PDB code: 1IAR). On top, predictions are visualised using a Jmol applet. On the left is IL-4 (chain A), on the right IL-4R (chain B). Predicted hot spots are in red, non hot spots in white. Residues not part of the interface are in blue. Below, predictions scores for each interface residues (excluding Pro and Gly amino acids) are reported (note that only the first few residues are displayed here). Scores greater than zero corresponds to predicted hot spots.
Mentions: We have implemented HSPred as a fully automatic web server, available at http://bioinf.cs.ucl.ac.uk/hspred. As input it requires a PDB formatted file containing the structure of the protein-protein complex. The user needs to define the interface to analyse by specifying the chain identifiers for each protein on either side of the interface. The output consists of two components: (i) a Jmol applet to visualise and explore the predictions using the protein structures and (ii) a table listing HSPred scores for each interface amino acid. The output page for an illustrative example is reported in Figure 4. The complex tested is Interleukin 4 (IL-4) bound to its receptor chain (IL-4R) (PDB code: 1IAR). Alanine mutational data from experiments are available for this complex [20], [21]. Out of interface mutations, HSPred predicts true positives, true negatives, false positive and false negatives. These results further validate the predictive accuracy of HSPred.

Bottom Line: These are the amino acid types on which the original model did not perform well.We have therefore developed two additional SVM classifiers, specifically optimised for these cases.HSPred reaches an overall precision and recall respectively of 61% and 69%, which roughly corresponds to a 10% improvement.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science, University College London, London, United Kingdom.

ABSTRACT
Protein-protein interactions are critically dependent on just a few 'hot spot' residues at the interface. Hot spots make a dominant contribution to the free energy of binding and they can disrupt the interaction if mutated to alanine. Here, we present HSPred, a support vector machine(SVM)-based method to predict hot spot residues, given the structure of a complex. HSPred represents an improvement over a previously described approach (Lise et al, BMC Bioinformatics 2009, 10:365). It achieves higher accuracy by treating separately predictions involving either an arginine or a glutamic acid residue. These are the amino acid types on which the original model did not perform well. We have therefore developed two additional SVM classifiers, specifically optimised for these cases. HSPred reaches an overall precision and recall respectively of 61% and 69%, which roughly corresponds to a 10% improvement. An implementation of the described method is available as a web server at http://bioinf.cs.ucl.ac.uk/hspred. It is free to non-commercial users.

Show MeSH
Related in: MedlinePlus