Systems-level approaches to discover host-targeted repurposable drugs against SARS-CoV-2 infection
Development of candidate anti-infectives identified by Artificial Intelligence (AI).
The UPDDI is participating in a COVID19-related project aimed at discovering potential new treatments for the disease through a collaborative effort by research groups at the University of Pittsburgh Department of Computational and Systems Biology, the University of California, Los Angeles (UCLA), and Cedars Sinai Hospital. Through a series of sophisticated mathematical and artificial intelligence-based methods, the team identified a number of potential candidates that were tested in the laboratory for inhibition of SARS-CoV-2 infection and cellular entry. The most promising candidates are currently being considered for further development into potential COVID19 therapies.
Workflow of the quantitative systems pharmacology approach for selecting compounds for experimental evaluation. (A) We use as input the RNA-seq data from SARS-CoV-2 infected A549 cells and ACE2-overexpressing A549 cells. (B) Up- and down-regulated differentially expressed genes (DEGs) were identified from these data using Wald test with false-discovery rate (FDR) default upper value of 00.05. (C) The antiviral gene signature was identified upon manual curation of GO enrichment results corresponding to the DEGs, using the QuickGO hierarchical annotation. (D) Candidate compounds or repurposable drugs that best reproduced the antiviral signature were extracted from Connectivity Map (CMap). (E) Known and predicted targets of these compounds were identified using QuartataWeb. (F) A host response network composed of four modules related to SARS-CoV-2 infection (called disease modules) was constructed. (G) The compound-target interaction network identified in E and the disease modules constructed in F were used to calculate a network proximity metric for each compound with respect to disease modules, which led to the identification 25 top-ranking compounds (repurposable or investigational drugs) for each module. (H) These were clustered to identify ‘representatives’ from each cluster in different modules and further curated manually to select a list of prioritized compounds that has been experimentally validated (I).
Papers
Chen F, Shi Q, Pei F, Vogt A, Porritt RA, Garcia G, Jr., Gomez AC, Cheng MH, Schurdak ME, Liu B, Chan SY, Arumugaswami V, Stern AM, Taylor DL, Arditi M, Bahar I. A systems-level study reveals host-targeted repurposable drugs against SARS-CoV-2 infection. Mol Syst Biol. 2021;17(8):e10239. PubMed PMID: 34339582. (open access) http://dx.doi.org/10.15252/msb.202110239
Click here and include “SARS” in the Additional Gift Instructions to support research at the UPDDI to apply AI to identify existing therapeutics to treat COVID-19 infections now and more rapidly identify treatments for future infectious agents.
