Protein family classification with partial least squares

Stephen O. Opiyo, Etsuko Moriyama

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

The quality of protein function predictions relies on appropriate training of protein classification methods. Performance of these methods can be affected when only a limited number of protein samples are available, which is often the case in divergent protein families. Whereas profile hidden Markov models and PSI-BLAST presented significant performance decrease in such cases, alignment-free partial least-squares classifiers performed consistently better even when used to identify short fragmented sequences.

Original languageEnglish (US)
Pages (from-to)846-853
Number of pages8
JournalJournal of Proteome Research
Volume6
Issue number2
DOIs
StatePublished - Feb 1 2007

Fingerprint

Least-Squares Analysis
Proteins
Hidden Markov models
Classifiers

Keywords

  • Amino acid composition
  • G-protein coupled receptors
  • Partial least square
  • Physico-chemical properties
  • Profile hidden Markov model

ASJC Scopus subject areas

  • Genetics
  • Biotechnology
  • Biochemistry

Cite this

Protein family classification with partial least squares. / Opiyo, Stephen O.; Moriyama, Etsuko.

In: Journal of Proteome Research, Vol. 6, No. 2, 01.02.2007, p. 846-853.

Research output: Contribution to journalArticle

@article{81bb0e8b3bff4b5c9b7b5cd82fbd9792,
title = "Protein family classification with partial least squares",
abstract = "The quality of protein function predictions relies on appropriate training of protein classification methods. Performance of these methods can be affected when only a limited number of protein samples are available, which is often the case in divergent protein families. Whereas profile hidden Markov models and PSI-BLAST presented significant performance decrease in such cases, alignment-free partial least-squares classifiers performed consistently better even when used to identify short fragmented sequences.",
keywords = "Amino acid composition, G-protein coupled receptors, Partial least square, Physico-chemical properties, Profile hidden Markov model",
author = "Opiyo, {Stephen O.} and Etsuko Moriyama",
year = "2007",
month = "2",
day = "1",
doi = "10.1021/pr060534k",
language = "English (US)",
volume = "6",
pages = "846--853",
journal = "Journal of Proteome Research",
issn = "1535-3893",
publisher = "American Chemical Society",
number = "2",

}

TY - JOUR

T1 - Protein family classification with partial least squares

AU - Opiyo, Stephen O.

AU - Moriyama, Etsuko

PY - 2007/2/1

Y1 - 2007/2/1

N2 - The quality of protein function predictions relies on appropriate training of protein classification methods. Performance of these methods can be affected when only a limited number of protein samples are available, which is often the case in divergent protein families. Whereas profile hidden Markov models and PSI-BLAST presented significant performance decrease in such cases, alignment-free partial least-squares classifiers performed consistently better even when used to identify short fragmented sequences.

AB - The quality of protein function predictions relies on appropriate training of protein classification methods. Performance of these methods can be affected when only a limited number of protein samples are available, which is often the case in divergent protein families. Whereas profile hidden Markov models and PSI-BLAST presented significant performance decrease in such cases, alignment-free partial least-squares classifiers performed consistently better even when used to identify short fragmented sequences.

KW - Amino acid composition

KW - G-protein coupled receptors

KW - Partial least square

KW - Physico-chemical properties

KW - Profile hidden Markov model

UR - http://www.scopus.com/inward/record.url?scp=33847341188&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33847341188&partnerID=8YFLogxK

U2 - 10.1021/pr060534k

DO - 10.1021/pr060534k

M3 - Article

VL - 6

SP - 846

EP - 853

JO - Journal of Proteome Research

JF - Journal of Proteome Research

SN - 1535-3893

IS - 2

ER -