GPCRdb in 2018: adding GPCR structure models and ligands.

Pándy-Szekeres G, Munk C, Tsonkov TM, Mordalski S, Harpsøe K, Hauser AS, Bojarski AJ, Gloriam DE. · 2018

Read GPCRdb in 2018: adding GPCR structure models and ligands. by Pándy-Szekeres G, Munk C, Tsonkov TM, Mordalski S, Harpsøe K, Hauser AS, Bojarski AJ, Gloriam DE. (2018) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc5753179.txt.

Introduction

Pándy-Szekeres and co-authors report the 2018 update of GPCRdb, a database for G protein-coupled receptors. These receptors mediate much human signaling and many drug effects. The update adds structure models across the full receptor set. It also adds about 150,000 ligands with activity data.

Each model uses a main template plus local ones for weak spots. Models come in inactive, middle, and active states for most classes. Ligand browsers sort by receptor, activity, traits, and shop availability. Users can download structures and data sheets for their own work. Three major studies are linked so readers can explore their findings in the database.

Catalog record

CREATORPándy-Szekeres G, Munk C, Tsonkov TM, Mordalski S, Harpsøe K, Hauser AS, Bojarski AJ, Gloriam DE.
YEAR2018
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC5753179

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

GPCRdb in 2018: adding GPCR structure models and ligands By Gáspár Pándy-Szekeres, Christian Munk, Tsonko M Tsonkov, Stefan Mordalski, Kasper Harpsøe, Alexander S Hauser, Andrzej J Bojarski, David E Gloriam Published in: Nucleic Acids Research DOI: 10.1093/nar/gkx1109 ABSTRACT Abstract G protein-coupled receptors are the most abundant mediators of both human signalling processes and therapeutic effects. Herein, we report GPCRome-wide homology models of unprecedented quality, and roughly 150 000 GPCR ligands with data on biological activities and commercial availability. Based on the strategy of ‘Less model – more Xtal’, each model exploits both a main template and alternative local templates. This achieved higher similarity to new structures than any of the existing resources, and refined crystal structures with missing or distorted regions. Models are provided for inactive, intermediate and active states—except for classes C and F that so far only have inactive templates. The ligand database has separate browsers for: (i) target selection by receptor, family or class, (ii) ligand filtering based on cross-experiment activities (min, max and mean) or chemical properties, (iii) ligand source data and (iv) commercial availability. SMILES structures and activity spreadsheets can be downloaded for further processing. Furthermore, three recent landmark publications on GPCR drugs, G protein selectivity and genetic variants have been accompanied with resources that now let readers view and analyse the findings themselves in GPCRdb. Altogether, this update will enable scientific investigation for the wider GPCR community. GPCRdb is available at http://www.gpcrdb.org. INTRODUCTION G protein-coupled receptors represent both the largest membrane and drug target protein families, accounting for 800 (4%) human genes and 34% of FDA-approved drug targets (1). The human GPCRs have been divided into six classes (or families): A (Rhodopsin), B1 (Secretin), B2 (Adhesion), C (Glutamate), F (Frizzled) and Taste 2 based on evolutionary homology (2,3). All GPCRs share a common heptahelical transmembrane domain carrying the signal from the extracellular ligand to the intracellular signalling protein, typically one of four families of G proteins or β-arrestin, which can also internalise the receptor. Recently, structural biology and pharmacology breakthroughs have opened up new GPCR research fields exploring structural mechanisms and complexes, and novel principles to achieve a more selective action through ‘biased agonism’ preferentially activating a specific signal protein or ‘allosteric modulation’ of the natural ligand response from an alternative binding site. The GPCR database, GPCRdb (4–6) serves the wide GPCR community, currently ∼1800 monthly users, with reference data, web server analysis tools and dynamic visualisation of data and statistics. Current data ranges all 398 human non-olfactory GPCRs and 16 G proteins, over 14 000 species orthologues, 30 328 binding site mutations, all 218 experimental structures, and 10 059 extracted ligand interactions. GPCR crystal structures have now been determined for the major human classes A–C, F (D and E fungal and amoeba receptors, respectively), but not yet for the Adhesion (B1) and Taste 2 classes, which have atypical pharmacology in that they are not activated by an endogenous agonist. In total, structures are available for 46 unique receptors, however close to 90% of the non-olfactory receptors still lack a crystal structure (http://gpcrdb.org/structure/statistics). Furthermore, only eleven receptors have a structure in the active state. Hence, many GPCR studies utilise homology models for virtual screening, lead optimisation, elucidation of ligand binding sites, off-target rationalisation, modeling of complexes, and molecular dynamics (7). The latest GPCR Dock assessment (8) showed that the accuracy of such models has greatly increased, and the best models utilize more than one template. Our 5-HT1B model achieved the best RMSD compared to the crystal structure and was based on five backbone templates and a position-specific rotamer library derived from all GPCR structures. This strategy of ‘Less model—more Xtal’ covers more of the model with experimental data by extending the main template with alternative local templates for missing or distorted segments. Herein, we present the first resource with pre-generated multi-template models of all human non-olfactory GPCRs in their inactive, intermediate and active states (except for classes C and F for which active and intermediate templates are still missing). GPCR ligands have previously been collated in a dedicated resource, GLIDA database (9), but maintenance was discontinued in October 2010 when the database contained 30,399 ligands. Today, the Guide to PHARMACOLOGY database offers curated reference ligands for a large number of protein families, including GPCRs (10), however this is just a small subset of the publically available ligands. The most comprehensive resource for biological activities is the ChEMBL database, which has compiled data from publications, patents and commercial sources (11). However, ChEMBL data is not presented and filtered to be directly accessible by many users. A further resource is the Ki database of the psychoactive drug screening programme containing broad target profiling, however most of which are single-point values (12). Herein, we present a new GPCR ligand resource that applies the same quality requirements as Guide to PHARMACOLOGY, but encompasses ligands from ChEMBL, GLIDA and the PDSP Ki database. It gives users novel possibilities, such as commercial availability, target selectivity and selection of ligands suitable for crystallography experiments. METHODS Technical features of the homology modeling pipeline The new GPCRdb homology modeling pipeline (Figure 1) updates the templates and models upon each database update. It was developed in Python in the Django web framework, utilizes several tools from Biopython (13), saves data in a PostgreSQL database, builds the final models with the help of MODELLER 9.18 (14) and uses NGL (15) for model visualisation. It leverages on integration of several unique manually annotated GPCRdb resources, including structure-based sequence alignments with generic residue numbers (16), GPCR structures (all in PDB) with defined preferred chain, helix segment borders, and distorted regions; and a position-specific sidechain rotamer library (17). [Figure: The GPCRdb homology model pipeline builds on the principle of ‘Less model – more Xtal’ meaning that more of the target is covered by structural templates, as an incomplete or partly distorted main template is complemented with alternative local templates. All templates and models are automatically updated upon each database update. Models are produced for all the human non-olfactory GPCRs in the inactive, and where templates exist (classes A and B1), also intermediate and active states.] Pairwise sequence alignment Target-template pairwise sequence alignments are fetched from GPCRdb. Main template selection For each receptor with a structure at least one (the most complete) representative template (currently totalling 169) has been manually annotated to define aligned segments (TM1–7, H8, ICL1–2 and ECL1–2), bulges, constrictions and any missing or distorted regions. A model main template is selected based on: (i) class (B2 to B1 and Taste2 to A), (ii) state (active/intermediate/inactive), (iii) receptor sequence similarity, (iv) structure coverage (no. generic residue positions (16) with coordinates) and (v) resolution. The template states are defined based on the extend of outward and inward movement of TM6 and TM5, respectively, on the intracellular, G protein binding side (http://docs.gpcrdb.org/structures.html#structure-state). Helix end adjustment Transmembrane helix ends are in some places (typically in ICL3) altered due to protein fusions, H8 is sometimes missing or truncated, and both can be distorted due to crystal packing effects. Non-native parts of helices are removed from the initial model and missing segments are modeled based on a local alternative template, which is superposed onto the last helical turn of the main template.

Read the complete text