G-Protein-Coupled Receptors (GPCRs) are cell membrane proteins of relevance to biology and pharmacology. Their supervised classification in subtypes is hampered by label noise, which stems from a combination of expert knowledge limitations and lack of clear correspondence between labels and different representations of the protein primary sequences. In this brief study, we describe a systematic approach to the analysis of GPCR misclassifications using Support Vector Machines and use it to assist the discovery of database labeling quality problems and investigate the extent to which GPCR sequence physicochemical transformations reflect GPCR subtype labeling. The proposed approach could enable a filtering approach to the label noise problem.
CitationKönig, C., Vellido, A., Alquezar, R., Giraldo, J. Misclassification of class C G-protein-coupled receptors as a label noise problem. A: European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. "ESANN 2014: 22st European Symposium on Artificial Neural Networks, Computational Intelligence And Machine Learning: Bruges April 23-24-25, 2014: proceedings". Bruges: 2014, p. 695-700.
All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder. If you wish to make any use of the work not provided for in the law, please contact: email@example.com