Multiscale interactome analysis coupled with off-target drug predictions reveals drug repurposing candidates for

Michael G Sugiyama1, Haotian Cui2,3, Dar'ya S Redka4

  • 1Department of Chemistry and Biology, Ryerson University, Toronto, ON, Canada.

Scientific Reports
|December 3, 2021
PubMed

Insights

This study introduces a computational method using graph convolutional networks to find new antiviral drugs for human coronaviruses. It identified capmatinib as a potent antiviral and revealed host protein roles, aiding future drug discovery.

Area of Science:

  • Virology
  • Computational Biology
  • Drug Discovery

Background:

  • The COVID-19 pandemic necessitates novel antiviral therapies for human coronavirus infections.
  • Human coronaviruses cause a spectrum of diseases, from common colds to severe respiratory illnesses.

Purpose of the Study:

  • To develop a computational strategy for identifying new antiviral drug targets and repurposing existing drugs against human coronaviruses.
  • To validate in silico predictions through cell-based experimental assessments.

Main Methods:

  • Utilized graph convolutional networks (GCN) for a computational approach.
  • Performed multiscale host-virus interactome analysis coupled with off-target drug predictions.
  • Conducted cell-based experimental validation of predicted drug candidates.

Main Results:

  • Identified several clinically relevant drug repurposing candidates with predicted antiviral activity.
  • Discovered capmatinib, a MET inhibitor, exhibits potent and broad-spectrum antiviral activity against multiple coronaviruses.
  • Uncovered novel roles for host cell proteins, including IRAK1/4, in facilitating human coronavirus infection.

Conclusions:

  • The computational approach is effective for identifying antiviral drug repurposing candidates.
  • Capmatinib shows promise as a broad-spectrum antiviral agent for coronavirus infections.
  • Understanding host-pathogen interactions, like the role of IRAK1/4, can guide future antiviral drug development.