Visual Exploratory Assessment of Class C GPCR Extracellular Domains Discrimination Capabilities
G protein-coupled receptors (GPCRs) are integral membrane-bound proteins. They are divided in five main classes with Class C members receiving recently special attention because of their involvement in many neurologic diseases. These receptors are composed of the seven-helix transmembrane domain (typical of all GPCRs) and a large extracellular domain where the endogenous ligand binds. In the present absence of crystal structures for complete Class C receptors, their primary sequences can provide limited but useful information that can be used for subtype discrimination in first instance. In this paper, we show that the extracellular part of these sequences provides as effective a discrimination as the complete sequence. With that purpose, we describe a process of exploratory sequence visualization using different data transformations and manifold learning techniques for dimensionality reduction. Class discriminability is assessed using an entropy-based measure.
KeywordsGPCRs N-terminus Data visualization GTM Kernel-GTM
Unable to display preview. Download preview PDF.
- 8.Vellido, A., et al.: Seeing is believing: the importance of visualization in real-world machine learning applications. In: Proceedings of ESANN 2011, pp. 219–226 (2011)Google Scholar
- 12.Cárdenas, M.I., Vellido, A., Giraldo, J.: Visual interpretation of class C GPCR subtype overlapping from the nonlinear mapping of transformed primary sequences. In: Proceedings of IEEE BHI 2014, pp. 764–767 (2014)Google Scholar
- 14.Olier, I., Vellido, A., Giraldo, J.: Kernel Generative topographic mapping. In: Proceedings of ESANN 2010, pp. 481–486 (2010)Google Scholar