Ontological and Bioinformatic Analysis of Anti-Coronavirus Drugs and Their Implication for Drug Repurposing against COVID-19

Abstract

Coronavirus-infected diseases have posed great threats to human health. In past years, highly infectious coronavirus-induced diseases, including COVID-19, SARS, and MERS, have resulted in world-wide severe infections. Our literature annotations identified 72 chemical drugs and 27 antibodies effective against at least one human coronavirus infection in vitro or in vivo. Many of these drugs inhibit viral entry to cells and viral replication inside cells or modulate host immune responses. Many antimicrobial drugs, including antimalarial (e.g., chloroquine and mefloquine) and antifungal (e.g., terconazole and rapamycin) drugs as well as antibiotics (e.g., teicoplanin and azithromycin) were associated with anti-coronavirus activity. A few drugs, including remdesivir, chloroquine phosphate, favipiravir, and tocilizumab, have already been reported to be effective in treating COVID-19. After mapping our identified drugs to three ontologies ChEBI, NDF-RT, and DrON, many features such as roles and mechanisms of action (MoAs) of these drugs were identified and categorized. For example, out of 35 drugs with MoA annotations in NDF-RT, 34 have MoAs of different types of inhibitors and antagonists. Two clustering analyses, one based on ChEBI-based semantic similarity, the other based on drug chemical similarity, were performed to cluster over 60 drugs to new categories. Moreover, PCA analysis of anti-coronavirus drugs found differences in physicochemical properties between those inhibiting viral entry and viral replication. A total of 137 host genes were identified as the targets of 47 anti-coronavirus drugs, resulting in a network of 370 interactions among these drugs and targets. Chlorpromazine, dasatinib, and anisomycin are the hubs of the drug-target network with the highest number of connected target proteins. Many enriched pathways such as calcium signaling and neuroactive ligand-receptor interaction pathways were identified. These findings may be used to facilitate drug repurposing against COVID-19.

CIDO, a community-based ontology for coronavirus disease knowledge and data integration, sharing, and analysis

“…In pursuit of that aim, CIDO can logically define relations between drugs and roles or mechanisms of action -distinct hierarchies in CIDO -and so support advanced analysis of potential drugs used to treat COVID-19, as well as the quick query of drugs having specific roles or mechanisms of action potentially useful as treatments. Such application of CIDO for ontology-based integration and analysis of anti-coronavirus drugs is shown in our recent preprint paper 17 . Using literature mining we identified 72 chemical drugs and 27 monoclonal or polyclonal antibodies that have anti-coronavirus effects in experimental studies in vivo or in vitro.…”

“…1 provides only a high-level overview of some CIDO resources, more details, such as specific signature genes in some cells of the lung that are susceptible to be up-or down-regulated in patients with COVID-19 will be added to the CIDO as new knowledge is acquired. Such systematic modeling and representation of the host-coronavirus interaction mechanisms would facilitate rational design of anti-coronavirus drugs and vaccines 17,18 .…”

COVID Term: A Bilingual Terminology for COVID-19

“…Liu et.al. completed an ontology towards coronavirus named Coronavirus Infectious Disease Ontology (CIDO) and made related analysis [18].…”

COVID Term: A Bilingual Terminology for COVID-19

“…Liu et.al. completed an ontology towards coronavirus named Coronavirus Infectious Disease Ontology (CIDO) and targeting at drug identi cation and repurposing [22].Plenty of techniques (e.g. arti cial intelligence, virtual reality, internet of things (IoT), internet of medical things (IoMT), etc.)…”