GPCRmd: easy sharing, visualization, and analysis of GPCR MD data

A community-driven research resource that aims to bring MD data within reach of all scientists interested in GPCRs.

Like Comment
Read the Article

It is often said that a picture is worth a thousand words. Sometimes, however, one static picture cannot explain the entire complexity of nature. For instance, based on a picture of a caterpillar and a butterfly, it would be very difficult to guess how one can transform into the other. In this scenario, a series of pictures -or a movie- would tell the whole story. The same happens with proteins such as G protein-coupled receptors (GPCRs), whose signaling outcome is known to be highly determined by their dynamic structural fluctuations. You can extract an extensive amount of useful information on GPCR functionality based on static structures but, due to the flexible nature of these receptors, studying their structural motions provide the missing keys necessary to achieve a complete understanding. Such insights can be obtained from molecular dynamics (MD) simulations (Figure 1): dynamic and interactive “movies” that yield information on the structural motions of GPCRs at atomic resolution. This can be useful for multiple different scientific disciplines, including protein engineering (e.g. detection of flexible receptor regions that require stabilization), drug design (e.g. detection of key drug-receptor interactions), and personalized medicine (e.g. impact of genetic variants on drug response). However, because of their complexity and size, generating, analyzing, or even visualizing MD simulations is not an easy task. It requires specialized knowledge of structural biology and the usage of specific and complex software, apart from some programming skills. In the end, because of this, MD simulation data tend to be restricted only to computational and structural biologists.

Figure 1. MD simulations provide information on the structural motions of a molecular system at atomic resolution.

Knowing the potential of MD simulations, we wanted to find a way to make this valuable data accessible to the whole community of scientists studying GPCRs -the motivation that originated the idea of GPCRmd (www.gpcrmd.org). We aimed to create an open-access platform to freely share MD data and include in it simulations of as many GPCRs as possible. However, we knew that this idea was ambitious and that we could not make it alone: we needed the support of a strong community for the development and later sustainability of such a platform. Thanks to the motivation and efforts of many other researchers who share our vision, we constituted the GPCRmd community, including research groups of 23 institutions from 10 different countries in Europe and the US. With them, we joined expertise to generate a common simulation protocol and use it to simulate around 70% of the 3D-GPCRome (Figure 2), which is the set of GPCR subtypes with solved structure. 

Figure 2. The first GPCRmd dataset of simulated structures covers 100% of GPCR classes, 80% of GPCR families, and 71% of receptors subtypes with solved structure at the time of writing, accounting for approximately 35% of all GPCR structures deposited in the PDB (black PDB identifiers). Colored circles represent active (green), intermediate (yellow), or inactive (red) receptor states.

We then needed to effectively share all this data and make it accessible for both MD experts and non-experts. For that, we built the online platform of GPCRmd (Figure 3). Apart from providing easy and free access to all the simulation data that we generated, our platform is intended to simplify the analysis of such data. Thus, we equipped it with fast, intuitive, and easy-to-use tools for the visualization and analysis of the simulations. We also included large-scale analysis tools to cluster and compare the simulations of different receptors based on their interaction patterns. This feature is especially valuable, as inspecting such an extensive dataset can help to uncover both universal or specific mechanisms that govern the structural dynamics of these receptors. As an example, in our publication, we demonstrate how this tool can be used to pinpoint differences in the water-mediated networks of the β2-adrenoceptor and OX2-receptor, potentially involved in receptor activation and G protein coupling. Also, in case anyone prefers to perform their own analyses, it is possible to freely download the data in GPCRmd and study it with external software or analysis scripts.

Figure 3. Summary of the main features available in GPCRmd. A GPCR-specific workbench enables interactive visualization (GPCRmd viewer) and analysis (GPCRmd toolkit) for individual simulations. The receptor meta-analysis provides tools for the comparative analysis and clustering of multiple MD simulations based on their interaction pattern.

As an open community, we are continuously working on extending our platform. On one hand, we aim to cover the whole 3D-GPCRome, which is constantly growing on its own. For that, we organize bi-annual updates from the GPCRmd community. On the other hand, we also encourage other researchers to submit and share their data. With this, we want to invite everyone to join our community and contribute either with data or suggestions for improvements or new tools for GPCRmd.

To sum up, GPCRmd is a community-driven platform designed to facilitate MD data availability, analysis, and exchange between different disciplines. Hence, it has the potential to boost the multidisciplinary research of GCRs with the final aim to achieve a deeper understanding of the molecular basis of their functionality. We thus welcome all scientists interested in GPCRs to visit it at www.gpcrmd.org.

Go to the profile of Mariona Torrens Fontanals

Mariona Torrens Fontanals

Predoctoral researcher, Pompeu Fabra University

1 Contributions
0 Followers
0 Following

No comments yet.