Deep Learning-based Network Pharmacology for Exploring the Mechanism of Licorice for the Treatment of COVID-19 (preprint)

Licorice, a traditional Chinese medicine, has been widely used for the treatment of COVID-19, but all active compounds and the corresponding targets are still not clear. Therefore, this study proposed a deep learning-based network pharmacology approach to identify more potential active compounds and targets of licorice and to collect information regarding different representative compounds. A graph convolutional neural network was used to construct a molecular map and a convolutional neural network was used to develop a Morgan fingerprint. Twenty core compounds and 6 core targets were predicted, among which 4 compounds (quercetin, naringenin, liquiritigenin, and licoisoflavanone), 2 targets (SYK and JAK2) and the relevant pathways (P53, cAMP, and NF-kB) were associated with SARS-CoV-2-infection, which were confirmed by previous studies. In addition, 2 new active compounds (glabrone and vestitol) and 2 new targets (PTEN and MAP3K8) were further validated by molecular docking, and the results showed that these active compounds bound to SARS-CoV-2 related targets, including the main protease (Mpro, also called 3CLpro), the spike protein (S protein), and the angiotensin-converting enzyme 2 (ACE2). Overall, we conclude that the findings of this study has the value of further exploration in the following experiment and clinical application.

Efficacy of remdesivir in COVID-19 patients with a simulated two-arm controlled study (preprint)

While the recent study on the compassionate use of remdesivir for COVID-19 patients has shown a 68% clinical improvement7 it is a one-arm study that renders the evaluation of the efficacy in reducing death and the length of stay of hospitalization intractable due to a lacking of the control group. We came up with a two-arm controlled study design to simulate the treated and the untreated (control group) group by applying two respective transition models to the empirical data on dynamics of the disease severity (Figure 2 of the original article7) that are classified into low- (no and low oxygen supplement), medium- (non-invasive ventilator and high oxygen supplement), and high-(ECMO and invasive ventilator) from enrolment until discharge, death or the end of follow-up. By using a simulated two-arm controlled study, the remdesivir treatment group as opposed to the control group led to a statistically significantly 29% (95% CI: 22-35%) reduction of death from COVID-19. The treated group also revealed a 33% (95% CI 28-38%) significantly higher odds of discharge than the control group. The median time to discharge for the treated group (5.5 days, 16.5 days, and 29.5 days for low-, medium-, and high-risk state, respectively) was around half of those of the control arm. Our results with a simulated two-arm controlled study have not only corroborated the efficacy of remdesivir but also made great contribution to designing a further large-scale randomized controlled trial. They have significant implications for reducing transmission probability and infectious time of COVID-19 patients when contacting with susceptible health care workers during hospitalization.