Hybrid learning algorithm in neural network system for enzyme classification

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Nucleic acid and protein sequences store a wealth of information which ultimately determines their functions and characteristics. Protein sequences classification deals with the assignment of sequences to known categories based on homology detection properties. In this paper, we developed a hybrid learning algorithm in neural network system called Neural Network Enzyme Classification (NNEC) to classify an enzyme found in Protein Data Bank (PDB) to a given family of enzymes. NNEC was developed based on Multilayer Perceptron with hybrid learning algorithm combining the genetic algorithm (GA) and Backpropagation (BP), where one of them acts as an operator in the other. Here, BP is used as a mutation-like-operator of the general GA search template. The proposed hybrid model was tested with different topologies of network architecture, especially in determining the number of hidden nodes. The precision results are quite promising in classifying the enzyme accordingly.

Original languageEnglish
Pages (from-to)209-220
Number of pages12
JournalInternational Journal of Advances in Soft Computing and its Applications
Volume2
Issue number2
Publication statusPublished - Jul 2010

Fingerprint

Learning algorithms
Enzymes
Neural networks
Proteins
Backpropagation
Genetic algorithms
Nucleic acids
Multilayer neural networks
Network architecture
Mathematical operators
Topology

Keywords

  • Enzyme
  • Hybrid GA-BP
  • Neural networks
  • Protein classification

ASJC Scopus subject areas

  • Computer Science Applications

Cite this

@article{03fe6647dfe64b2894dcd5fc0401b4de,
title = "Hybrid learning algorithm in neural network system for enzyme classification",
abstract = "Nucleic acid and protein sequences store a wealth of information which ultimately determines their functions and characteristics. Protein sequences classification deals with the assignment of sequences to known categories based on homology detection properties. In this paper, we developed a hybrid learning algorithm in neural network system called Neural Network Enzyme Classification (NNEC) to classify an enzyme found in Protein Data Bank (PDB) to a given family of enzymes. NNEC was developed based on Multilayer Perceptron with hybrid learning algorithm combining the genetic algorithm (GA) and Backpropagation (BP), where one of them acts as an operator in the other. Here, BP is used as a mutation-like-operator of the general GA search template. The proposed hybrid model was tested with different topologies of network architecture, especially in determining the number of hidden nodes. The precision results are quite promising in classifying the enzyme accordingly.",
keywords = "Enzyme, Hybrid GA-BP, Neural networks, Protein classification",
author = "Osman, {Mohd Haniff} and Liong, {Choong Yeun} and Ishak Hashim",
year = "2010",
month = "7",
language = "English",
volume = "2",
pages = "209--220",
journal = "International Journal of Advances in Soft Computing and its Applications",
issn = "2074-8523",
publisher = "International Center for Scientific Research and Studies (ICSRS)",
number = "2",

}

TY - JOUR

T1 - Hybrid learning algorithm in neural network system for enzyme classification

AU - Osman, Mohd Haniff

AU - Liong, Choong Yeun

AU - Hashim, Ishak

PY - 2010/7

Y1 - 2010/7

N2 - Nucleic acid and protein sequences store a wealth of information which ultimately determines their functions and characteristics. Protein sequences classification deals with the assignment of sequences to known categories based on homology detection properties. In this paper, we developed a hybrid learning algorithm in neural network system called Neural Network Enzyme Classification (NNEC) to classify an enzyme found in Protein Data Bank (PDB) to a given family of enzymes. NNEC was developed based on Multilayer Perceptron with hybrid learning algorithm combining the genetic algorithm (GA) and Backpropagation (BP), where one of them acts as an operator in the other. Here, BP is used as a mutation-like-operator of the general GA search template. The proposed hybrid model was tested with different topologies of network architecture, especially in determining the number of hidden nodes. The precision results are quite promising in classifying the enzyme accordingly.

AB - Nucleic acid and protein sequences store a wealth of information which ultimately determines their functions and characteristics. Protein sequences classification deals with the assignment of sequences to known categories based on homology detection properties. In this paper, we developed a hybrid learning algorithm in neural network system called Neural Network Enzyme Classification (NNEC) to classify an enzyme found in Protein Data Bank (PDB) to a given family of enzymes. NNEC was developed based on Multilayer Perceptron with hybrid learning algorithm combining the genetic algorithm (GA) and Backpropagation (BP), where one of them acts as an operator in the other. Here, BP is used as a mutation-like-operator of the general GA search template. The proposed hybrid model was tested with different topologies of network architecture, especially in determining the number of hidden nodes. The precision results are quite promising in classifying the enzyme accordingly.

KW - Enzyme

KW - Hybrid GA-BP

KW - Neural networks

KW - Protein classification

UR - http://www.scopus.com/inward/record.url?scp=84860269162&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84860269162&partnerID=8YFLogxK

M3 - Article

VL - 2

SP - 209

EP - 220

JO - International Journal of Advances in Soft Computing and its Applications

JF - International Journal of Advances in Soft Computing and its Applications

SN - 2074-8523

IS - 2

ER -