Classifying G protein-coupled receptors with multiple physicochemical properties

Jingyi Yang, Jitender Deogun

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Automated annotation of G protein-coupled receptors (GPCRs) has been an intriguing topic because of the important role of GPCRs in pharmaceutical research. The diverse nature of GPCRs results in the lack of overall sequence homolog among members, making the classification of GPCRs a challenging task. In this paper, we propose a new method to classify GPCRs based on only their primary sequences. We extract feature vectors from protein sequences based on various physicochemical properties and use the Support Vector Machine (SVM) for the classification. When features derived from multiple properties are used together, we obtain the accuracy of 97.61% on GPCR Level I subfamily classification and 99.94% on GPCR superfamily recognition in double cross-validation tests. The results compare favorably with those reported in previous publications.

Original languageEnglish (US)
Title of host publicationBioMedical Engineering and Informatics
Subtitle of host publicationNew Development and the Future - Proceedings of the 1st International Conference on BioMedical Engineering and Informatics, BMEI 2008
Pages93-97
Number of pages5
DOIs
StatePublished - 2008
EventBioMedical Engineering and Informatics: New Development and the Future - 1st International Conference on BioMedical Engineering and Informatics, BMEI 2008 - Sanya, Hainan, China
Duration: May 27 2008May 30 2008

Publication series

NameBioMedical Engineering and Informatics: New Development and the Future - Proceedings of the 1st International Conference on BioMedical Engineering and Informatics, BMEI 2008
Volume1

Conference

ConferenceBioMedical Engineering and Informatics: New Development and the Future - 1st International Conference on BioMedical Engineering and Informatics, BMEI 2008
Country/TerritoryChina
CitySanya, Hainan
Period5/27/085/30/08

ASJC Scopus subject areas

  • Information Systems
  • Signal Processing
  • Biomedical Engineering

Fingerprint

Dive into the research topics of 'Classifying G protein-coupled receptors with multiple physicochemical properties'. Together they form a unique fingerprint.

Cite this