The ongoing pandemic of SARS-CoV-2 virus has already affected most of the world in a dramatic way. It endangers the lives of many people, puts a tremendous strain on the medical system, while the second-order effects on the economy and daily life have been equally devastating. Governments and the private sector are undertaking massive concerted efforts in tackling the crisis. Unquestionably, a discovery of a vaccine and cure will be of paramount importance for long-term control of the spread of the disease. In support of the efforts, BioExcel experts are partnering with numerous international initiatives by bringing all of our advanced software applications and expertise.
BioExcel is partnering with MolSSI (www.molssi.org) on a community-driven data repository and curation service for molecular structures, models, therapeutics, and simulations related to computational research related to therapeutic opportunities for COVID-19: https://covid.bioexcel.eu/ and https://covid.molssi.org/. Join us in this endeavor by contributing any data (small molecules, models, simulation data, new targets, etc) which may be useful to the wider communities. Please visit https://covid.molssi.org/contributing for more information.
Nostrum Biodiscovery (NBD) is focusing on screening its proprietary virtual library, ChemistriX as well as other open libraries (Zinc), into the 3C main protease of SARS-CoV-2; blocking this enzyme, the virus cannot replicate efficiently. For this reason, they are using a hierarchical docking approach combining Glide (docking tool from Schrodinger) with NBD’s proprietary Monte Carlo code PELE, a computational protocol that has shown remarkable results in international blind competitions. In addition, this approach is highly parallel and can take advantage of the supercomputer facilities at Barcelona Supercomputing Center. The work is in collaboration with the Electronic and Atomic Protein Modelling group led Victor Guallar at the Barcelona Supercomputing Center, and contributes altruistically to the EXSCALATE4CoronaVirus consortia.
Within the same consortium, KTH Royal Institute of Technology is bringing GROMACS and consultancy expertise for the needs of tailoring molecular dynamics simulations for large-scale executions.
INM-9 (Institute of Neuroscience and Medicine) at Jülich Research Center, using local supercomputing resources, is identifying ligands targeting proteins contributing to the coronavirus-host interactions, key to viral pathogenesis. Within the project, the team (together with CINECA, Bologna, Italy) is currently performing HPC-based virtual ligand screening experiments (in parallel on different proteins) on the JSC machines. The resulting libraries of molecules will undergo in-vitro testing.
University of Jyväskyla (JYU) is applying large-scale molecular dynamics simulations of the SARS-CoV-2 coronavirus’ spike protein bound to the human ACE2 receptor as a first step to understand the interactions that control host cell recognition in lung tissue. The group is also investigating the spike protein complexed with aptamer candidates to systematically search for oligonucleotides that can selectively bind to the spike protein and prevent host cell recognition. The large-scale simulations are done via custom workflows, which are run efficiently on the HPC resources at CSC in Finland. Moreover, the workflows and protocol used in this project will be readily available when possible future outbreaks of other infections occur. The project involves a team of specialists in chemistry, molecular modeling, infectious diseases as well as in aptamer selection and modification. Collaborators include researchers from Moscow State University and the Russian Academy of Sciences, who have a strong track record in developing DNA aptamers for diagnostic and therapeutic purposes.
In a parallel project, JYU works together with Petri Pihko, an expert in organic synthesis, and Varpu Marjomäki, a virology specialist, on inhibition of the RdRp with nucleotide analogues. The Covid-19 RdRp structure was recently resolved to 2.9 Å resolution with cryo-EM (6m71.pdb) and kindly shared with the group by the authors. Since then, the group has already performed MD simulations of the RdRp complexed with double stranded RNA consisting of the template and primer. Next steps include free energy computations to predict the effects of chemical modification of nucleotide bases, both inside the nucleotide binding site before the polymerization reaction, as well as after insertion of the nucleotide into the RNA primer. In addition, work will be done on investigating possible cooperative effect of the inhibitor by modifying multiple bases in the primer strand, as these are speculated to induce conformational changes in the complex that ultimately can cause its arrest.
In close collaboration, the MPG team in Göttingen has started to simulate the complex of the SARS-CoV-2 coronavirus spike protein bound to the human ACE2 receptor. Using a combination of GROMACS and PMX, the interface will be optimized with the goal to design a derivative of ACE2 with enhanced affinity to the SARS-CoV-2 spike protein for either diagnostic or therapeutic purposes.
The WeNMR portal, developed and run by our partners at Utrecht University, has seen an increase of registrations over the last weeks with many users indicating they intend to use it for COVID-19 projects. The HADDOCK WeNMR team is already involved in several collaborations ranging from drug screening against the protease to modelling COVID-19 related protein-protein interactions. For this purpose, together with EGI/EOSC experts, the team is looking both into expanding the processing capacity of the HADDOCK portals and providing customized solutions to support researchers. These might take the form of dedicated virtual clusters with a HADDOCK frontend running on EOSC cloud resource, and customized virtual machines with ready to run local HADDOCK installation for experienced users wanting to use the software at the command line.
Institute for Biomedicine Barcelona (IRB) works on understanding how the virus has evolved by comparing its structure/genome to other coronaviridae including GATg13, SARS-CoV as well as the US-variant. Hence, to extensive unbiased and enhanced sampling MD simulations, as well as bioinformatic analysis, work aims at the unveiling the virus evolution and host selection mechanism by identifying key structural determinants and pathogenic mutations that occurred from one viral strain to the others; a point of utmost importance also for future pandemic surveillance. Importantly, the gathered knowledge will be used to identify virus inhibition opportunities by means of in-silico drug screening, starting from known and commercially available drugs first. The work is done in collaboration with Roberto Burioni’s laboratory for viral research in San Raffaele, Italy.
In addition, BioExcel is working on a collaboration with the Protein Data Bank, EMBL-EBI, EOSC-Life and MolSSI for providing a repository of curated modelling and simulation data related to the virus studies. The data can be used as a reliable starting point for further research by research groups worldwide.