OpenZika Team Adds New Resources to Continue Massive Data Analysis

New research models, a new student, new publications, new collaborations, and continuing data analysis are all detailed in this comprehensive update from the OpenZika research team.

Current Lab Work

Dr. Carolina Andrade’s research team, LabMol, is working on analyzing the data generated by OpenZika project. Dr. Melina Mottin is leading this task.

We are analyzing the virtual screen from OpenZika against the ZIKV NS5 protein, a key viral enzyme that is required for the virus to mature and form infectious viral particles. The full-length NS5 protein contains two domains, representing two distinct targets: the NS5 methyltransferase and NS5 polymerase. NS5 methyltransferase domain methylates, or attaches methyl (CH3) radicals to, the RNA cap structure; and NS5 polymerase domain synthesizes the viral RNA, and is thus essential to ZIKV survival and establishment of the infection in host cells.

We are being assisted by other Brazilian researchers, who are helping to analyze the big data generated by the docking calculations from World Community Grid. Professor Roosevelt Alves da Silva (Federal University of Goias), Dr. Joao H. Martins Sena (Oswaldo Cruz Foundation, FIOCRUZ), and Dr. Pedro Torres (University of Cambridge) are helping Dr. Mottin and the LabMol team to handle the massive amount of data we have been generating. Also, our new Master’s student, Bruna Souza, is being trained by Dr. Mottin. She is already analyzing the data and preparing reports.

In addition, we have prepared capsid proteins (from Zika, dengue and West Nile viruses) to “feed the beast” (that is, to keep creating the docking jobs that you all crunch for us), and perform docking calculations. The flavivirus capsid proteins have no inhibitors co-crystalized, so we performed a binding pocket prediction of the ZIKV capsid structure. Two binding pockets were identified. Then, we submitted docking jobs regarding pocket 1 and 2 of capsid protein against the ZINC15 database (containing 30.2 million compounds).

Dr. Mottin Visits Dr. Ekins

Dr. Mottin spent almost a month as a visiting researcher at Dr. Sean Ekins’ lab in North Carolina. She worked on Bayesian and Random Forest models for Zika, dengue and Ebola using publicly available data. The main goal of this work was to develop and implement these models to help increase the efficiency of the final selection of the compounds to test against the viruses. (For example, we can build a machine learning model for Zika and related flaviviruses that could help predict which compounds are more likely to have whole-cell activity, which will complement our target-based docking in OpenZika.) Models are now being optimized with new datasets obtained from the literature and databases such as ChEMBL and PubChem, to build new and more robust machine learning models for predicting activity for Zika and dengue viruses.

New Candidate Inhibitors Recently Discovered

We recently finished analyzing a virtual screen from OpenZika against the ZIKV NS2B/NS3 protease, a key viral enzyme that is required for the virus to mature and form infectious viral particles. Approximately 6 million compounds were docked against the ZIKV protease, followed by interaction-based docking filters to find compounds that were predicted to dock well to the allosteric site. We then used machine learning models as filters (our suite of Bayesian models that predict mouse liver microsomal stability, lack of cytotoxicity, and solubility). We ultimately visually inspected the binding modes of the top 318 compounds, and we narrowed it down to 27 candidates. These candidates will be purchased and then assayed by our collaborators, to discern their potency at inhibiting the protease, as well as their efficacy in cell-based assays.

We also analyzed the docking results regarding the Chembridge database (~1 million compounds) against NS5 polymerase (RNA and NTP sites) and NS5 methyltransferase (active, SAM and GTP sites), as well as the NS3 helicase (ATP and RNA sites). We filtered the best ranked compounds through the developed ZIKV QSAR models and performed a medicinal chemistry inspection of the filtered compounds, selecting 33 compounds in total. The next steps will be to order these compounds and experimentally test them (Figure 1). 

Upcoming and Recent Publications

We recently published a review paper entitled “Computational drug discovery for the Zika virus.” This paper was published in a special issue of the Brazilian Journal of Pharmaceutical Sciences. In this paper, we summarize current computational drug discovery efforts and their application to the discovery of anti-ZIKV drugs. We also present successful examples of the use of computational approaches to ZIKV drug discovery, including our OpenZika project.

In November, we published a keynote review entitled “The A–Z of Zika drug discovery” in the journal Drug Discovery Today. This is a comprehensive review of the recent advances in ZIKV drug discovery efforts, highlighting drug repositioning and computationally guided compounds, including recently discovered viral and host cell inhibitors. Promising ZIKV molecular targets are also described and discussed, as well as targets belonging to the host cell, as new opportunities for ZIKV drug discovery. All this knowledge is not only crucial to advancing the fight against the Zika virus and other flaviviruses, but it will also help the scientific community prepare for the next emerging virus outbreak to which we will have to respond.

Dr. Ekins, Dr. Andrade, and Dr. Mottin, along with other researchers, recently had a review paper “High Throughput and Computational Repurposing for Neglected Diseases” accepted in the journal Pharmaceutical Research. This paper describes the many drug repurposing efforts that have been going on in different labs around the world to try to find treatments for many tropical diseases.

The OpenZika project results were presented at the 256th ACS National Meeting, on August 19-23, in Boston, MA, USA. Dr. Mottin gave an oral presentation and presented a poster entitled “OpenZika: Discovery of new antiviral candidates against Zika virus,” in the session Chemoinformatics Approaches to Enhance Drug Discovery Based on Natural Products.

Dr. Mottin also gave an oral presentation “Applying Molecular Dynamics to Drug Discovery for Zika Virus and Schistosoma mansoni”, in the South American Initiative for Cooperation on Molecular Simulations (SAIMS) meeting, held at Institut Pasteur, Montevideo, Uruguay, November 4-7. The meeting was a great opportunity to exchange experiences and collaborate with South American researchers who work with Zika.

Past Publications and Outreach

Dr. Sean Ekins presented a poster at Cell Symposia: Emerging and Re-emerging Viruses, on October 1-3, 2017, in Arlington, VA, USA, entitled “OpenZika: Opening up the discovery of new antiviral candidates against Zika virus.”

Our PLoS Neglected Tropical Diseases paper, "OpenZika: An IBM World Community Grid Project to Accelerate Zika Virus Drug Discovery," was published on October 20 2016, and it has already been viewed more than 5,200 times. Anyone can access and read this paper for free. Another research paper “Illustrating and homology modeling the proteins of the Zika virus” was published in F1000Research and viewed > 4,200 times. 

We also published a research paper entitled “Molecular Dynamics simulations of Zika Virus NS3 helicase: Insights into RNA binding site activity” in October 2017, as part of a special issue on Flaviviruses for the journal Biochemical and Biophysical Research Communications. This study of the NS3 helicase system helped us learn more about this promising target for blocking Zika replication. The results will help guide how we analyze the virtual screens that we performed against NS3 helicase, and the Molecular Dynamics simulations generated new conformations of this system that we have been using as targets in new virtual screens that we performed as part of OpenZika.

New Collaborations

These articles and presentations are helping to bring additional attention to the project and encouraging the formation of new collaborations. 

We started a very important collaboration with the Center for Innovation in Biodiversity and Drug Discovery (CIBFar), coordinated by Professor Glaucius Oliva, hosted at University of Sao Paulo (USP), Brazil, São Carlos Institute of Physics (IFSC). The main goal of this collaboration is to test our compounds directly on enzymatic assays with Zika virus proteins. Our selected compounds are being tested to see if they can bind to the NS3 helicase, using the differential scanning fluorescence (DSF) technique and/or if they can inhibit the ATPase activity of this protein.

Another collaboration has been started with a Brazilian group working on semi-synthetic and natural products, led by Professor Luis Octavio Regasini, from the Department of Chemistry and Environmental Sciences, São Paulo State University (UNESP), and Professor Ana Carolina Gomes Jardim, from the Institute of Biomedical Sciences, Federal University of Uberlandia (UFU), a virologist expert in running phenotypic assays (cell-based assays) with viruses. They are screening a library of natural and synthetic compounds which include flavonoids, alkaloids and diarylamines to assess their antiviral potential against ZIKV infection in vitro. For the discovered experimental hits, we are performing docking calculations using AutoDock Vina with ZIKV proteins, using the same protocols and models that we developed in the OpenZika project. The results of this collaboration are being written in a research paper, which will be submitted soon to a scientific journal for publication.

Dr. Sean Ekins is continuing a collaboration with Dr. Scott Laster and Dr. Frank Scholle at NC State University to study flaviviruses. They have access to cell-based plaque assays for these viruses. We have also been evaluating several natural products they identified to predict potential targets in Zika and evaluate them experimentally in the labs of our collaborators described above.

New Student Team Member

As mentioned above, Dr. Andrade has hired a new graduate student, Bruna Sousa, to work on the OpenZika project. She started working in Dr. Andrade’s lab in March 2018, and she is very enthusiastic about learning and collaborating on the project. She is a volunteer for the OpenZika project with an Android device, and has invited students and professors of the graduate program, during classes and through her personal social media, to participate in the project. She is joining in the OpenZika calls with World Community Grid.

She has presented her new results in a presentation entitled “Computational and experimental strategies for the identification of protease and helicase proteins of the dengue and Zika viruses”. She was selected to participate at the IX School of Molecular Modeling in Biological Systems (IX EMMSB), August 20- 24, at the National Laboratory of Scientific Computing (LNCC) in Petropolis - Rio de Janeiro, Brazil.

She also presented a poster entitled “Molecular Docking studies of a novel diarylamine compound active against Zika virus replication.” In October 15-17, she participated in the XV Congress of Research, Teaching and Extension of Federal University of Goiás (UFG), where she presented a poster entitled “Molecular docking of an antranilic acid derivative against Zika proteins."

Status of the Calculations

In total, we have submitted almost 6.8 billion docking jobs, which involved 427 different target sites. Our initial screens used an older library of 6 million commercially available compounds, and our current experiments utilize the new ZINC15 library of 30.2 million compounds. We have already received approximately 6.3 billion of these results on our server. (As a reminder, there is some lag time between when the calculations are performed on our volunteer’s machines and when we get the results, since all of the results per “package” of approximately 10,000 – 50,000 different docking jobs need to be returned to World Community Grid, re-organized, and then compressed before they are sent to our server.)

So far, the > 80,000 volunteers who have donated their spare computing power to OpenZika have given us > 63,023 CPU years’ worth of docking calculations, at a current average of 66.3 CPU years per day!  Thank you all very much for your help!

Except for a few stragglers, we have received all of the results for our experiments that involve docking 6 million compounds versus NS1, NS3 helicase (both the RNA binding site and the ATP site), NS5 (both the RNA polymerase and the methyltransferase domains), NS2B/NS3 protease and capsid (binding pockets 1 and 2). 

Thank you to all the volunteers who are donating their unused computing time to this project!  We appreciate your help!

Related Articles