Déjà vu: Stimulating Open Drug Discovery for SARS-CoV-2

Abstract

In the past decade we have seen two major Ebola virus outbreaks in Africa, the Zika virus in Brazil and the current outbreak of coronavirus disease which has been named "severe acute respiratory syndrome coronavirus 2" (SARS-CoV-2). There is a strong sense of Déjà vu as the world is caught flat footed without effective treatments to administer to patients. Our team has been actively involved in several small molecule drug discovery efforts for the preceding virus outbreaks. In 2014 we used machine learning to identify 3 new molecules to test for the Ebola virus and these were subsequently shown to be active in vitro and in vivo. We have also been involved in open science approaches that leverage the community to help. In 2016 we launched the OpenZika project as an IBM World Community Grid Project that used distributed computing power of volunteers to dock large numbers of compounds into Zika and related flavivirus targets. This led us into several collaborations in which we validated computational predictions in vitro. With both of these initiatives there was some knowledge of the virus, many compounds had already been tested in the case of Ebola, whereas for Zika initially all we had was the virus RNA sequence. In the current SARS-CoV-2 outbreak, this was a completely new virus and the scientists in China and elsewhere have started from scratch. In the space of a few weeks since the outbreak is acknowledged to have started, there are now compounds suggested as active in vitro and molecules repurposed in clinical trials. While this has been impressive, we propose there may still be gaps in our approach to drug discovery for such outbreaks. There is an opportunity to repurpose additional approved drugs for this virus and we now suggest how these might be identified leveraging prior work on MERS-CoV, SARS-CoV and other viruses. We also describe some of the immense challenges and limitations of the open antiviral drug discovery approaches we have been involved in.

A Very Large-Scale Bioactivity Comparison of Deep Learning and Multiple Machine Learning Algorithms for Drug Discovery

“…Assay Central TM is proprietary software for curating high-quality datasets, generating Bayesian machine learning models with extended-connectivity fingerprints (ECFP6) generated from the CDK library 56 , and making prospective predictions of potential for bioactivity 1,[51][52][53][57][58][59][60][61][62][63][64] . We have previously described this software in detail 1,[51][52][53][57][58][59][60][61][62][63][64] as well as the interpretation of prediction scores 54,65 , and the metrics generated from internal five-fold cross-validation used to evaluate and compare predictive performances. These metrics include, but are not limited to, ROC score 51 , Cohen's kappa (CK) 66,67 , and Matthew's correlation coefficient (MCC) 68 as described by us previously.…”

Dispirotripiperazine-core compounds, their biological activity with a focus on broad antiviral property, and perspectives in drug design (mini-review)

“…All of these recent viral outbreaks point to the fact that even though they stimulated increased drug discovery efforts, none have yet to bear fruit in the form of an approved antiviral likely due to limited and inconsistent funding over the years. There is also a need to apply different computational technologies as early as possible in an attempt to speed up the process of antiviral drug discovery [ 6 , 7 ].…”

Current targets and drug candidates for prevention and treatment of SARS-CoV-2 (COVID-19) infection

“…Various drugs evaluated against SARS-CoV and MERS-CoV have been explored in terms of efficacy for repurposing during the SARS-CoV-2 outbreak (Horie et al, 2020;Kandimalla et al, 2020;Kalra et al, 2020;Pizzorno et al, 2020;Santos et al, 2020). Synthetic drugs (example chloroquine and hydroxychloroquine), and antivirals (remdesivir and favipiravir) and herbal phytoconstituents (baicalin, glycyrrhizin, hesperetin, quercetin, and scutellarin), are considered for COVID-19 due to their potential action on host ACE2 or RdRp (Chen et al, 2004;Ekins et al, 2020;Ho et al, 2007;Zhang and Yap, 2004). None of the other antiviral therapeutic interventions like Lopinavir/Ritonavir, and Remdesivir seemed to alter the natural clinical course of the disease based on the available literature (Romagnoli et al, 2020;Rosendaal, 2020), however multiple investigations are ongoing.…”