Advances in exploration of machine learning methods for predicting functional class and interaction profiles of proteins and peptides irrespective of sequence homology

Juan Cui, Lianyi Han, Honghuang Lin, Zhiqun Tang, Zhiliang Ji, Zhiwei Cao, Yixue Li, Yuzong Chen

Research output: Contribution to journalReview article

6 Citations (Scopus)

Abstract

Various computational methods have been used for predicting protein function from clues contained in protein sequence. A particular challenge is the functional prediction of proteins that show low or no sequence similarity to proteins of known function. Recently, machine learning methods have been explored for predicting functional class of proteins from a variety of sequence-derived structural and physicochemical properties independent of sequence similarity, which showed promising potential for a broad spectrum of proteins including those that show low and no similarity to other proteins. These methods can thus be explored as potential tools to complement similarity-based, clustering-based and structure-based methods for predicting protein function. This article reviews the strategies, algorithms, current progresses, available software and web-servers, and underlying difficulties in using machine learning methods for predicting the functional class of proteins and peptides, and protein-protein interactions. The reported prediction performances in the application of these methods are also presented.

Original languageEnglish (US)
Pages (from-to)95-112
Number of pages18
JournalCurrent Bioinformatics
Volume2
Issue number2
DOIs
StatePublished - May 1 2007

Fingerprint

Sequence Homology
Peptides
Learning systems
Homology
Machine Learning
Proteins
Protein
Interaction
Profile
Class
Web Server
Performance Prediction
Protein-protein Interaction
Protein Sequence
Computational Methods
Complement
Computational methods
Clustering
Cluster Analysis
Software

Keywords

  • Machine learning method
  • Neural network
  • Peptide prediction
  • Protein function prediction
  • Protein-protein interaction
  • Support vector machine

ASJC Scopus subject areas

  • Biochemistry
  • Molecular Biology
  • Genetics
  • Computational Mathematics

Cite this

Advances in exploration of machine learning methods for predicting functional class and interaction profiles of proteins and peptides irrespective of sequence homology. / Cui, Juan; Han, Lianyi; Lin, Honghuang; Tang, Zhiqun; Ji, Zhiliang; Cao, Zhiwei; Li, Yixue; Chen, Yuzong.

In: Current Bioinformatics, Vol. 2, No. 2, 01.05.2007, p. 95-112.

Research output: Contribution to journalReview article

Cui, Juan ; Han, Lianyi ; Lin, Honghuang ; Tang, Zhiqun ; Ji, Zhiliang ; Cao, Zhiwei ; Li, Yixue ; Chen, Yuzong. / Advances in exploration of machine learning methods for predicting functional class and interaction profiles of proteins and peptides irrespective of sequence homology. In: Current Bioinformatics. 2007 ; Vol. 2, No. 2. pp. 95-112.
@article{f191488995074ecd8692d99a59d3d11f,
title = "Advances in exploration of machine learning methods for predicting functional class and interaction profiles of proteins and peptides irrespective of sequence homology",
abstract = "Various computational methods have been used for predicting protein function from clues contained in protein sequence. A particular challenge is the functional prediction of proteins that show low or no sequence similarity to proteins of known function. Recently, machine learning methods have been explored for predicting functional class of proteins from a variety of sequence-derived structural and physicochemical properties independent of sequence similarity, which showed promising potential for a broad spectrum of proteins including those that show low and no similarity to other proteins. These methods can thus be explored as potential tools to complement similarity-based, clustering-based and structure-based methods for predicting protein function. This article reviews the strategies, algorithms, current progresses, available software and web-servers, and underlying difficulties in using machine learning methods for predicting the functional class of proteins and peptides, and protein-protein interactions. The reported prediction performances in the application of these methods are also presented.",
keywords = "Machine learning method, Neural network, Peptide prediction, Protein function prediction, Protein-protein interaction, Support vector machine",
author = "Juan Cui and Lianyi Han and Honghuang Lin and Zhiqun Tang and Zhiliang Ji and Zhiwei Cao and Yixue Li and Yuzong Chen",
year = "2007",
month = "5",
day = "1",
doi = "10.2174/157489307780618222",
language = "English (US)",
volume = "2",
pages = "95--112",
journal = "Current Bioinformatics",
issn = "1574-8936",
publisher = "Bentham Science Publishers B.V.",
number = "2",

}

TY - JOUR

T1 - Advances in exploration of machine learning methods for predicting functional class and interaction profiles of proteins and peptides irrespective of sequence homology

AU - Cui, Juan

AU - Han, Lianyi

AU - Lin, Honghuang

AU - Tang, Zhiqun

AU - Ji, Zhiliang

AU - Cao, Zhiwei

AU - Li, Yixue

AU - Chen, Yuzong

PY - 2007/5/1

Y1 - 2007/5/1

N2 - Various computational methods have been used for predicting protein function from clues contained in protein sequence. A particular challenge is the functional prediction of proteins that show low or no sequence similarity to proteins of known function. Recently, machine learning methods have been explored for predicting functional class of proteins from a variety of sequence-derived structural and physicochemical properties independent of sequence similarity, which showed promising potential for a broad spectrum of proteins including those that show low and no similarity to other proteins. These methods can thus be explored as potential tools to complement similarity-based, clustering-based and structure-based methods for predicting protein function. This article reviews the strategies, algorithms, current progresses, available software and web-servers, and underlying difficulties in using machine learning methods for predicting the functional class of proteins and peptides, and protein-protein interactions. The reported prediction performances in the application of these methods are also presented.

AB - Various computational methods have been used for predicting protein function from clues contained in protein sequence. A particular challenge is the functional prediction of proteins that show low or no sequence similarity to proteins of known function. Recently, machine learning methods have been explored for predicting functional class of proteins from a variety of sequence-derived structural and physicochemical properties independent of sequence similarity, which showed promising potential for a broad spectrum of proteins including those that show low and no similarity to other proteins. These methods can thus be explored as potential tools to complement similarity-based, clustering-based and structure-based methods for predicting protein function. This article reviews the strategies, algorithms, current progresses, available software and web-servers, and underlying difficulties in using machine learning methods for predicting the functional class of proteins and peptides, and protein-protein interactions. The reported prediction performances in the application of these methods are also presented.

KW - Machine learning method

KW - Neural network

KW - Peptide prediction

KW - Protein function prediction

KW - Protein-protein interaction

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=34249291647&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34249291647&partnerID=8YFLogxK

U2 - 10.2174/157489307780618222

DO - 10.2174/157489307780618222

M3 - Review article

AN - SCOPUS:34249291647

VL - 2

SP - 95

EP - 112

JO - Current Bioinformatics

JF - Current Bioinformatics

SN - 1574-8936

IS - 2

ER -