TopBackground
The computational studies are dealt with different aspects of molecular properties of GPCRs, such as sequence-based identification and classification, sequence analyses, topology prediction and three-dimensional model generation.
Many distinct methodologies based on full-length sequence and motif-based search approaches, machine learning, and several alignment-free techniques have all been used successfully to identify and then classify GPCRs. GPCRTOP (Sokouti, Rezvan, Yuchdav, & Dastmalchi, 2014), GPCRHMM (Wistrand, Kall, & Sonnhammer, 2006), 7TMHMM (Moller, Vilo, & Croning, 2001), Pred-GPCR (Papasaikas, Bagos, Litou, Promponas, & Hamodrakas, 2004), and GPCRsClass (Bhasin & Raghava, 2005) are illustrative examples of many available methodologies being used for detection and classification of GPCRs. Topology prediction of GPCRs can be considered in the context of more general problem of IMP topology prediction. Consequently, methods such as TOP-PRED (von Heijne, 1992), SOSUI (Hirokawa, Boon-Chieng, & Mitaku, 1998), HMMTOP (Tusnady & Simon, 2001) and TMHMM (Krogh, Larsson, von Heijne, & Sonnhammer, 2001) are used frequently for the purpose of topology prediction of GPCRs. In its simplest form, a GPCR topology prediction can be regarded as search for conserved seven hydrophobic stretches of residues in the sequence. Combination of the above mentioned methods was also used to improve discrimination of a particular family of GPCRs. According to latest classifications, GPCRs are grouped into six categories based on their shared sequence homologies, namely class A (Rhodoposin-like), class B (Secretin receptor family), class C (Metabotropic glutamate and pheromone), class D (Fungal mating pheromone receptors), class E (Cyclic AMP receptors) and class F (Frizzled and Smoothened) (X. Xiao, Wang, & Chou, 2011) (Figure 1). The importance of this superfamily of proteins, urged scientific community to setup specialized database servers to make the study of these proteins more accessible. GPCRDB (Vroling et al., 2010), IUPHAR (Sharman et al., 2012) and GPCR section of UniProt (“The Universal Protein Resource (UniProt) in 2010,” 2009) are the well known publicly available databases. Structural information on GPCRs can be predicted at different levels ranging from detection of such proteins from their sequence information, prediction of locations and numbers of transmembrane segments within their sequences up to constructing their 3D model structures.
Figure 1. The GPCR Universe. (A) Statistics about annotated GPCR sequences in Uniprot and available experimental structures. (B) GPCRDB database statistics.
Sequence Analysis of GPCRs, Identification and Classification
The prediction of sequence location and transmembrane orientation (collectively called topology) are the essential first steps in the analysis of GPCRs primary structures. The properties of the two-dimensional lipid bilayer impose a set of constraints on the folding of GPCRs, which can make predicting the location of TM segments relatively straightforward using a range of currently available methodologies (Rath et al., 2013) .