Prediction of inhibitor binding free energies by quantum neural networks. Nucleoside analogues binding to trypanosomal nucleoside hydrolase

Benjamin B. Braunheim, Robert W. Miles, Vern L. Schramm, Steven D. Schwartz

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

A computational method has been developed to predict inhibitor binding energy for untested inhibitor molecules. A neural network is trained from the electrostatic potential surfaces of known inhibitors and their binding energies. The algorithm is then able to predict, with high accuracy, the binding energy of unknown inhibitors. IU-nucleoside hydrolase from Crithidia fasciculata and the inhibitor molecules described previously [Miles, R. W. Tyler, P. C. Evans, G. Fumeaux R. H., ParK(i)n, D. W., and Schramm, V. L. (1999) Biochemistry 38, xxxx-xxxx] are used as the test system. Discrete points on the molecular electrostatic potential surface of inhibitor molecules are input to neural networks to identify the quantum mechanical features that contribute to binding. Feed-forward neural networks with back- propagation of error are trained to recognize the quantum mechanical electrostatic potential and geometry at the entire van der Waals surface of a group of training molecules and to predict the strength of interactions between the enzyme and novel inhibitors. The binding energies of unknown inhibitors were predicted, followed by experimental determination of K(i) values. Predictions of K(i) values using this theory are compared to other methods and are more robust in estimating inhibitory strength. The average deviation in estimating K(i) values for 18 unknown inhibitor molecules, with 21 training molecules, is a factor of 5 x K(i) over a range of 660 000 in K(i) values for all molecules. The a posteriori accuracy of the predictions suggests the method will be effective as a guide for experimental inhibitor design.

Original languageEnglish (US)
Pages (from-to)16076-16083
Number of pages8
JournalBiochemistry
Volume38
Issue number49
DOIs
StatePublished - Dec 7 1999

Fingerprint

N-Glycosyl Hydrolases
Static Electricity
Nucleosides
Free energy
Neural networks
Molecules
Binding energy
Crithidia fasciculata
Electrostatics
Enzyme Inhibitors
Biochemistry
Corrosion inhibitors
Research Design
Feedforward neural networks
Computational methods
Backpropagation
Geometry

ASJC Scopus subject areas

  • Biochemistry

Cite this

Prediction of inhibitor binding free energies by quantum neural networks. Nucleoside analogues binding to trypanosomal nucleoside hydrolase. / Braunheim, Benjamin B.; Miles, Robert W.; Schramm, Vern L.; Schwartz, Steven D.

In: Biochemistry, Vol. 38, No. 49, 07.12.1999, p. 16076-16083.

Research output: Contribution to journalArticle

Braunheim, Benjamin B. ; Miles, Robert W. ; Schramm, Vern L. ; Schwartz, Steven D. / Prediction of inhibitor binding free energies by quantum neural networks. Nucleoside analogues binding to trypanosomal nucleoside hydrolase. In: Biochemistry. 1999 ; Vol. 38, No. 49. pp. 16076-16083.
@article{48e058e1018048879526dd59e1df7bc6,
title = "Prediction of inhibitor binding free energies by quantum neural networks. Nucleoside analogues binding to trypanosomal nucleoside hydrolase",
abstract = "A computational method has been developed to predict inhibitor binding energy for untested inhibitor molecules. A neural network is trained from the electrostatic potential surfaces of known inhibitors and their binding energies. The algorithm is then able to predict, with high accuracy, the binding energy of unknown inhibitors. IU-nucleoside hydrolase from Crithidia fasciculata and the inhibitor molecules described previously [Miles, R. W. Tyler, P. C. Evans, G. Fumeaux R. H., ParK(i)n, D. W., and Schramm, V. L. (1999) Biochemistry 38, xxxx-xxxx] are used as the test system. Discrete points on the molecular electrostatic potential surface of inhibitor molecules are input to neural networks to identify the quantum mechanical features that contribute to binding. Feed-forward neural networks with back- propagation of error are trained to recognize the quantum mechanical electrostatic potential and geometry at the entire van der Waals surface of a group of training molecules and to predict the strength of interactions between the enzyme and novel inhibitors. The binding energies of unknown inhibitors were predicted, followed by experimental determination of K(i) values. Predictions of K(i) values using this theory are compared to other methods and are more robust in estimating inhibitory strength. The average deviation in estimating K(i) values for 18 unknown inhibitor molecules, with 21 training molecules, is a factor of 5 x K(i) over a range of 660 000 in K(i) values for all molecules. The a posteriori accuracy of the predictions suggests the method will be effective as a guide for experimental inhibitor design.",
author = "Braunheim, {Benjamin B.} and Miles, {Robert W.} and Schramm, {Vern L.} and Schwartz, {Steven D.}",
year = "1999",
month = "12",
day = "7",
doi = "10.1021/bi990830t",
language = "English (US)",
volume = "38",
pages = "16076--16083",
journal = "Biochemistry",
issn = "0006-2960",
publisher = "American Chemical Society",
number = "49",

}

TY - JOUR

T1 - Prediction of inhibitor binding free energies by quantum neural networks. Nucleoside analogues binding to trypanosomal nucleoside hydrolase

AU - Braunheim, Benjamin B.

AU - Miles, Robert W.

AU - Schramm, Vern L.

AU - Schwartz, Steven D.

PY - 1999/12/7

Y1 - 1999/12/7

N2 - A computational method has been developed to predict inhibitor binding energy for untested inhibitor molecules. A neural network is trained from the electrostatic potential surfaces of known inhibitors and their binding energies. The algorithm is then able to predict, with high accuracy, the binding energy of unknown inhibitors. IU-nucleoside hydrolase from Crithidia fasciculata and the inhibitor molecules described previously [Miles, R. W. Tyler, P. C. Evans, G. Fumeaux R. H., ParK(i)n, D. W., and Schramm, V. L. (1999) Biochemistry 38, xxxx-xxxx] are used as the test system. Discrete points on the molecular electrostatic potential surface of inhibitor molecules are input to neural networks to identify the quantum mechanical features that contribute to binding. Feed-forward neural networks with back- propagation of error are trained to recognize the quantum mechanical electrostatic potential and geometry at the entire van der Waals surface of a group of training molecules and to predict the strength of interactions between the enzyme and novel inhibitors. The binding energies of unknown inhibitors were predicted, followed by experimental determination of K(i) values. Predictions of K(i) values using this theory are compared to other methods and are more robust in estimating inhibitory strength. The average deviation in estimating K(i) values for 18 unknown inhibitor molecules, with 21 training molecules, is a factor of 5 x K(i) over a range of 660 000 in K(i) values for all molecules. The a posteriori accuracy of the predictions suggests the method will be effective as a guide for experimental inhibitor design.

AB - A computational method has been developed to predict inhibitor binding energy for untested inhibitor molecules. A neural network is trained from the electrostatic potential surfaces of known inhibitors and their binding energies. The algorithm is then able to predict, with high accuracy, the binding energy of unknown inhibitors. IU-nucleoside hydrolase from Crithidia fasciculata and the inhibitor molecules described previously [Miles, R. W. Tyler, P. C. Evans, G. Fumeaux R. H., ParK(i)n, D. W., and Schramm, V. L. (1999) Biochemistry 38, xxxx-xxxx] are used as the test system. Discrete points on the molecular electrostatic potential surface of inhibitor molecules are input to neural networks to identify the quantum mechanical features that contribute to binding. Feed-forward neural networks with back- propagation of error are trained to recognize the quantum mechanical electrostatic potential and geometry at the entire van der Waals surface of a group of training molecules and to predict the strength of interactions between the enzyme and novel inhibitors. The binding energies of unknown inhibitors were predicted, followed by experimental determination of K(i) values. Predictions of K(i) values using this theory are compared to other methods and are more robust in estimating inhibitory strength. The average deviation in estimating K(i) values for 18 unknown inhibitor molecules, with 21 training molecules, is a factor of 5 x K(i) over a range of 660 000 in K(i) values for all molecules. The a posteriori accuracy of the predictions suggests the method will be effective as a guide for experimental inhibitor design.

UR - http://www.scopus.com/inward/record.url?scp=0033534170&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033534170&partnerID=8YFLogxK

U2 - 10.1021/bi990830t

DO - 10.1021/bi990830t

M3 - Article

VL - 38

SP - 16076

EP - 16083

JO - Biochemistry

JF - Biochemistry

SN - 0006-2960

IS - 49

ER -