Artificial neural network prediction of clozapine response with combined pharmacogenetic and clinical data

Chao Cheng Lin, Ying Chieh Wang, Jen Yeu Chen, Ying Jay Liou, Ya Mei Bai, I. Ching Lai, Tzu Ting Chen, Hung Wen Chiu, Yu Chuan Li

Research output: Contribution to journalArticle

29 Citations (Scopus)

Abstract

Although one third to one half of refractory schizophrenic patients responds to clozapine, however, there are few evidences currently that could predict clozapine response before the use of the medication. The present study aimed to train and validate artificial neural networks (ANN), using clinical and pharmacogenetic data, to predict clozapine response in schizophrenic patients. Five pharmacogenetic variables and five clinical variables were collated from 93 schizophrenic patients taking clozapine, including 26 responders. ANN analysis was carried out by training the network with data from 75% of cases and subsequently testing with data from 25% of unseen cases to determine the optimal ANN architecture. Then the leave-one-out method was used to examine the generalization of the models. The optimal ANN architecture was found to be a standard feed-forward, fully-connected, back-propagation multilayer perceptron. The overall accuracy rate of ANN was 83.3%, which is higher than that of logistic regression (LR) (70.8%). By using the area under the receiver operating characteristics curve as a measure of performance, the ANN outperformed the LR (0.821 ± 0.054 versus 0.579 ± 0.068; p <0.001). The ANN with only genetic variables outperformed the ANN with only clinical variables (0.805 ± 0.056 versus 0.647 ± 0.066; p = 0.046). The gene polymorphisms should play an important role in the prediction. Further validation of ANN analysis is likely to provide decision support for predicting individual response.

Original languageEnglish
Pages (from-to)91-99
Number of pages9
JournalComputer Methods and Programs in Biomedicine
Volume91
Issue number2
DOIs
Publication statusPublished - Aug 2008

Fingerprint

Clozapine
Pharmacogenetics
Neural networks
Logistic Models
Neural Networks (Computer)
ROC Curve
Electric network analysis
Network architecture
Logistics
Multilayer neural networks
Genes
Polymorphism
Backpropagation
Refractory materials
Testing

Keywords

  • Clozapine
  • Genetic polymorphism
  • Neural network models
  • Schizophrenia

ASJC Scopus subject areas

  • Software

Cite this

Artificial neural network prediction of clozapine response with combined pharmacogenetic and clinical data. / Lin, Chao Cheng; Wang, Ying Chieh; Chen, Jen Yeu; Liou, Ying Jay; Bai, Ya Mei; Lai, I. Ching; Chen, Tzu Ting; Chiu, Hung Wen; Li, Yu Chuan.

In: Computer Methods and Programs in Biomedicine, Vol. 91, No. 2, 08.2008, p. 91-99.

Research output: Contribution to journalArticle

Lin, Chao Cheng ; Wang, Ying Chieh ; Chen, Jen Yeu ; Liou, Ying Jay ; Bai, Ya Mei ; Lai, I. Ching ; Chen, Tzu Ting ; Chiu, Hung Wen ; Li, Yu Chuan. / Artificial neural network prediction of clozapine response with combined pharmacogenetic and clinical data. In: Computer Methods and Programs in Biomedicine. 2008 ; Vol. 91, No. 2. pp. 91-99.
@article{bd097873b2144bc38d5a760c830270ce,
title = "Artificial neural network prediction of clozapine response with combined pharmacogenetic and clinical data",
abstract = "Although one third to one half of refractory schizophrenic patients responds to clozapine, however, there are few evidences currently that could predict clozapine response before the use of the medication. The present study aimed to train and validate artificial neural networks (ANN), using clinical and pharmacogenetic data, to predict clozapine response in schizophrenic patients. Five pharmacogenetic variables and five clinical variables were collated from 93 schizophrenic patients taking clozapine, including 26 responders. ANN analysis was carried out by training the network with data from 75{\%} of cases and subsequently testing with data from 25{\%} of unseen cases to determine the optimal ANN architecture. Then the leave-one-out method was used to examine the generalization of the models. The optimal ANN architecture was found to be a standard feed-forward, fully-connected, back-propagation multilayer perceptron. The overall accuracy rate of ANN was 83.3{\%}, which is higher than that of logistic regression (LR) (70.8{\%}). By using the area under the receiver operating characteristics curve as a measure of performance, the ANN outperformed the LR (0.821 ± 0.054 versus 0.579 ± 0.068; p <0.001). The ANN with only genetic variables outperformed the ANN with only clinical variables (0.805 ± 0.056 versus 0.647 ± 0.066; p = 0.046). The gene polymorphisms should play an important role in the prediction. Further validation of ANN analysis is likely to provide decision support for predicting individual response.",
keywords = "Clozapine, Genetic polymorphism, Neural network models, Schizophrenia",
author = "Lin, {Chao Cheng} and Wang, {Ying Chieh} and Chen, {Jen Yeu} and Liou, {Ying Jay} and Bai, {Ya Mei} and Lai, {I. Ching} and Chen, {Tzu Ting} and Chiu, {Hung Wen} and Li, {Yu Chuan}",
year = "2008",
month = "8",
doi = "10.1016/j.cmpb.2008.02.004",
language = "English",
volume = "91",
pages = "91--99",
journal = "Computer Methods and Programs in Biomedicine",
issn = "0169-2607",
publisher = "Elsevier Ireland Ltd",
number = "2",

}

TY - JOUR

T1 - Artificial neural network prediction of clozapine response with combined pharmacogenetic and clinical data

AU - Lin, Chao Cheng

AU - Wang, Ying Chieh

AU - Chen, Jen Yeu

AU - Liou, Ying Jay

AU - Bai, Ya Mei

AU - Lai, I. Ching

AU - Chen, Tzu Ting

AU - Chiu, Hung Wen

AU - Li, Yu Chuan

PY - 2008/8

Y1 - 2008/8

N2 - Although one third to one half of refractory schizophrenic patients responds to clozapine, however, there are few evidences currently that could predict clozapine response before the use of the medication. The present study aimed to train and validate artificial neural networks (ANN), using clinical and pharmacogenetic data, to predict clozapine response in schizophrenic patients. Five pharmacogenetic variables and five clinical variables were collated from 93 schizophrenic patients taking clozapine, including 26 responders. ANN analysis was carried out by training the network with data from 75% of cases and subsequently testing with data from 25% of unseen cases to determine the optimal ANN architecture. Then the leave-one-out method was used to examine the generalization of the models. The optimal ANN architecture was found to be a standard feed-forward, fully-connected, back-propagation multilayer perceptron. The overall accuracy rate of ANN was 83.3%, which is higher than that of logistic regression (LR) (70.8%). By using the area under the receiver operating characteristics curve as a measure of performance, the ANN outperformed the LR (0.821 ± 0.054 versus 0.579 ± 0.068; p <0.001). The ANN with only genetic variables outperformed the ANN with only clinical variables (0.805 ± 0.056 versus 0.647 ± 0.066; p = 0.046). The gene polymorphisms should play an important role in the prediction. Further validation of ANN analysis is likely to provide decision support for predicting individual response.

AB - Although one third to one half of refractory schizophrenic patients responds to clozapine, however, there are few evidences currently that could predict clozapine response before the use of the medication. The present study aimed to train and validate artificial neural networks (ANN), using clinical and pharmacogenetic data, to predict clozapine response in schizophrenic patients. Five pharmacogenetic variables and five clinical variables were collated from 93 schizophrenic patients taking clozapine, including 26 responders. ANN analysis was carried out by training the network with data from 75% of cases and subsequently testing with data from 25% of unseen cases to determine the optimal ANN architecture. Then the leave-one-out method was used to examine the generalization of the models. The optimal ANN architecture was found to be a standard feed-forward, fully-connected, back-propagation multilayer perceptron. The overall accuracy rate of ANN was 83.3%, which is higher than that of logistic regression (LR) (70.8%). By using the area under the receiver operating characteristics curve as a measure of performance, the ANN outperformed the LR (0.821 ± 0.054 versus 0.579 ± 0.068; p <0.001). The ANN with only genetic variables outperformed the ANN with only clinical variables (0.805 ± 0.056 versus 0.647 ± 0.066; p = 0.046). The gene polymorphisms should play an important role in the prediction. Further validation of ANN analysis is likely to provide decision support for predicting individual response.

KW - Clozapine

KW - Genetic polymorphism

KW - Neural network models

KW - Schizophrenia

UR - http://www.scopus.com/inward/record.url?scp=44949169359&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=44949169359&partnerID=8YFLogxK

U2 - 10.1016/j.cmpb.2008.02.004

DO - 10.1016/j.cmpb.2008.02.004

M3 - Article

C2 - 18508152

AN - SCOPUS:44949169359

VL - 91

SP - 91

EP - 99

JO - Computer Methods and Programs in Biomedicine

JF - Computer Methods and Programs in Biomedicine

SN - 0169-2607

IS - 2

ER -