Limits...
ResBoost: characterizing and predicting catalytic residues in enzymes.

Alterovitz R, Arvey A, Sankararaman S, Dallett C, Freund Y, Sjölander K - BMC Bioinformatics (2009)

Bottom Line: The method effectively selects and combines rules of thumb into a simple, easily interpretable logical expression that can be used for prediction.ResBoost reduces the number of false positives by up to 56% compared to the use of evolutionary conservation scoring alone.We also illustrate the ability of ResBoost to identify recently validated catalytic residues not listed in the CSA.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Science, University of North Carolina at Chapel Hill, USA. ron@cs.unc.edu

ABSTRACT

Background: Identifying the catalytic residues in enzymes can aid in understanding the molecular basis of an enzyme's function and has significant implications for designing new drugs, identifying genetic disorders, and engineering proteins with novel functions. Since experimentally determining catalytic sites is expensive, better computational methods for identifying catalytic residues are needed.

Results: We propose ResBoost, a new computational method to learn characteristics of catalytic residues. The method effectively selects and combines rules of thumb into a simple, easily interpretable logical expression that can be used for prediction. We formally define the rules of thumb that are often used to narrow the list of candidate residues, including residue evolutionary conservation, 3D clustering, solvent accessibility, and hydrophilicity. ResBoost builds on two methods from machine learning, the AdaBoost algorithm and Alternating Decision Trees, and provides precise control over the inherent trade-off between sensitivity and specificity. We evaluated ResBoost using cross-validation on a dataset of 100 enzymes from the hand-curated Catalytic Site Atlas (CSA).

Conclusion: ResBoost achieved 85% sensitivity for a 9.8% false positive rate and 73% sensitivity for a 5.7% false positive rate. ResBoost reduces the number of false positives by up to 56% compared to the use of evolutionary conservation scoring alone. We also illustrate the ability of ResBoost to identify recently validated catalytic residues not listed in the CSA.

Show MeSH

Related in: MedlinePlus

Sensitivity vs. false-positive rate curve for ResBoost compared to Global Conservation, ET, and ConSurf based on normalized score thresholding. At 85% sensitivity, ResBoost cuts the false positive rate by 55% compared to global conservation, 48% compared to ConSurf, and 32% compared to ET.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2713229&req=5

Figure 3: Sensitivity vs. false-positive rate curve for ResBoost compared to Global Conservation, ET, and ConSurf based on normalized score thresholding. At 85% sensitivity, ResBoost cuts the false positive rate by 55% compared to global conservation, 48% compared to ConSurf, and 32% compared to ET.

Mentions: Figure 3 illustrates the Receiver-Operator Characteristic (ROC) curve indicating the trade-off between sensitivity and FPR for ResBoost for increasing values of k. As described in the Methods section, for methods that generate residue scores rather than classifications, we normalized all scores for each score type on a protein by protein basis and then used a score threshold to classify residues as catalytic or non-catalytic. Using the CSA as ground truth, ResBoost achieved a sensitivity of 73% at a false positive rate of 5.7% (k = 32), a sensitivity of 85% at a false positive rate of 9.8% (k = 128), and a sensitivity of 88% at a false positive rate of 14% (k = 256). Increasing k improves sensitivity but also introduces more false positives.


ResBoost: characterizing and predicting catalytic residues in enzymes.

Alterovitz R, Arvey A, Sankararaman S, Dallett C, Freund Y, Sjölander K - BMC Bioinformatics (2009)

Sensitivity vs. false-positive rate curve for ResBoost compared to Global Conservation, ET, and ConSurf based on normalized score thresholding. At 85% sensitivity, ResBoost cuts the false positive rate by 55% compared to global conservation, 48% compared to ConSurf, and 32% compared to ET.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2713229&req=5

Figure 3: Sensitivity vs. false-positive rate curve for ResBoost compared to Global Conservation, ET, and ConSurf based on normalized score thresholding. At 85% sensitivity, ResBoost cuts the false positive rate by 55% compared to global conservation, 48% compared to ConSurf, and 32% compared to ET.
Mentions: Figure 3 illustrates the Receiver-Operator Characteristic (ROC) curve indicating the trade-off between sensitivity and FPR for ResBoost for increasing values of k. As described in the Methods section, for methods that generate residue scores rather than classifications, we normalized all scores for each score type on a protein by protein basis and then used a score threshold to classify residues as catalytic or non-catalytic. Using the CSA as ground truth, ResBoost achieved a sensitivity of 73% at a false positive rate of 5.7% (k = 32), a sensitivity of 85% at a false positive rate of 9.8% (k = 128), and a sensitivity of 88% at a false positive rate of 14% (k = 256). Increasing k improves sensitivity but also introduces more false positives.

Bottom Line: The method effectively selects and combines rules of thumb into a simple, easily interpretable logical expression that can be used for prediction.ResBoost reduces the number of false positives by up to 56% compared to the use of evolutionary conservation scoring alone.We also illustrate the ability of ResBoost to identify recently validated catalytic residues not listed in the CSA.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Science, University of North Carolina at Chapel Hill, USA. ron@cs.unc.edu

ABSTRACT

Background: Identifying the catalytic residues in enzymes can aid in understanding the molecular basis of an enzyme's function and has significant implications for designing new drugs, identifying genetic disorders, and engineering proteins with novel functions. Since experimentally determining catalytic sites is expensive, better computational methods for identifying catalytic residues are needed.

Results: We propose ResBoost, a new computational method to learn characteristics of catalytic residues. The method effectively selects and combines rules of thumb into a simple, easily interpretable logical expression that can be used for prediction. We formally define the rules of thumb that are often used to narrow the list of candidate residues, including residue evolutionary conservation, 3D clustering, solvent accessibility, and hydrophilicity. ResBoost builds on two methods from machine learning, the AdaBoost algorithm and Alternating Decision Trees, and provides precise control over the inherent trade-off between sensitivity and specificity. We evaluated ResBoost using cross-validation on a dataset of 100 enzymes from the hand-curated Catalytic Site Atlas (CSA).

Conclusion: ResBoost achieved 85% sensitivity for a 9.8% false positive rate and 73% sensitivity for a 5.7% false positive rate. ResBoost reduces the number of false positives by up to 56% compared to the use of evolutionary conservation scoring alone. We also illustrate the ability of ResBoost to identify recently validated catalytic residues not listed in the CSA.

Show MeSH
Related in: MedlinePlus