A genetic approach for building different alphabets for peptide and protein classification
<p>Abstract</p> <p>Background</p> <p>In this paper, it is proposed an optimization approach for producing reduced alphabets for peptide classification, using a Genetic Algorithm. The classification task is performed by a multi-classifier system where each classifier (Li...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
BMC
2008-01-01
|
Series: | BMC Bioinformatics |
Online Access: | http://www.biomedcentral.com/1471-2105/9/45 |
id |
doaj-8e56371397a54d85ac27a9b25125f376 |
---|---|
record_format |
Article |
spelling |
doaj-8e56371397a54d85ac27a9b25125f3762020-11-24T21:18:05ZengBMCBMC Bioinformatics1471-21052008-01-01914510.1186/1471-2105-9-45A genetic approach for building different alphabets for peptide and protein classificationLumini AlessandraNanni Loris<p>Abstract</p> <p>Background</p> <p>In this paper, it is proposed an optimization approach for producing reduced alphabets for peptide classification, using a Genetic Algorithm. The classification task is performed by a multi-classifier system where each classifier (Linear or Radial Basis function Support Vector Machines) is trained using features extracted by different reduced alphabets. Each alphabet is constructed by a Genetic Algorithm whose objective function is the maximization of the area under the ROC-curve obtained in several classification problems.</p> <p>Results</p> <p>The new approach has been tested in three peptide classification problems: HIV-protease, recognition of T-cell epitopes and prediction of peptides that bind human leukocyte antigens. The tests demonstrate that the idea of training a pool classifiers by reduced alphabets, created using a Genetic Algorithm, allows an improvement over other state-of-the-art feature extraction methods.</p> <p>Conclusion</p> <p>The validity of the novel strategy for creating reduced alphabets is demonstrated by the performance improvement obtained by the proposed approach with respect to other reduced alphabets-based methods in the tested problems.</p> http://www.biomedcentral.com/1471-2105/9/45 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Lumini Alessandra Nanni Loris |
spellingShingle |
Lumini Alessandra Nanni Loris A genetic approach for building different alphabets for peptide and protein classification BMC Bioinformatics |
author_facet |
Lumini Alessandra Nanni Loris |
author_sort |
Lumini Alessandra |
title |
A genetic approach for building different alphabets for peptide and protein classification |
title_short |
A genetic approach for building different alphabets for peptide and protein classification |
title_full |
A genetic approach for building different alphabets for peptide and protein classification |
title_fullStr |
A genetic approach for building different alphabets for peptide and protein classification |
title_full_unstemmed |
A genetic approach for building different alphabets for peptide and protein classification |
title_sort |
genetic approach for building different alphabets for peptide and protein classification |
publisher |
BMC |
series |
BMC Bioinformatics |
issn |
1471-2105 |
publishDate |
2008-01-01 |
description |
<p>Abstract</p> <p>Background</p> <p>In this paper, it is proposed an optimization approach for producing reduced alphabets for peptide classification, using a Genetic Algorithm. The classification task is performed by a multi-classifier system where each classifier (Linear or Radial Basis function Support Vector Machines) is trained using features extracted by different reduced alphabets. Each alphabet is constructed by a Genetic Algorithm whose objective function is the maximization of the area under the ROC-curve obtained in several classification problems.</p> <p>Results</p> <p>The new approach has been tested in three peptide classification problems: HIV-protease, recognition of T-cell epitopes and prediction of peptides that bind human leukocyte antigens. The tests demonstrate that the idea of training a pool classifiers by reduced alphabets, created using a Genetic Algorithm, allows an improvement over other state-of-the-art feature extraction methods.</p> <p>Conclusion</p> <p>The validity of the novel strategy for creating reduced alphabets is demonstrated by the performance improvement obtained by the proposed approach with respect to other reduced alphabets-based methods in the tested problems.</p> |
url |
http://www.biomedcentral.com/1471-2105/9/45 |
work_keys_str_mv |
AT luminialessandra ageneticapproachforbuildingdifferentalphabetsforpeptideandproteinclassification AT nanniloris ageneticapproachforbuildingdifferentalphabetsforpeptideandproteinclassification AT luminialessandra geneticapproachforbuildingdifferentalphabetsforpeptideandproteinclassification AT nanniloris geneticapproachforbuildingdifferentalphabetsforpeptideandproteinclassification |
_version_ |
1726010379327242240 |