A Structure-Informed Atlas of Human-Virus Interactions.

Area of Science:

• Virology
• Structural Biology
• Bioinformatics

Background:

• Understanding virus-host protein-protein interactions (PPIs) is crucial for virology and disease research.
• High-throughput methods for identifying PPIs are limited in scalability for many viruses.

Purpose of the Study:

• To develop and implement a computational framework, P-HIPSTer, for predicting pan-viral human PPIs using structural information.
• To identify novel viral-human interactions and understand shared and unique mechanisms across human-infecting viruses.

Main Methods:

• Implemented an in silico computational framework, P-HIPSTer (pathogen host interactome prediction using structure similarity).
• Utilized structural information to predict viral-human PPIs.
• Experimentally validated a subset of predicted interactions.

Main Results:

• Predicted approximately 282,000 pan viral-human PPIs with a ~76% experimental validation rate.
• Discovered shared and unique molecular machinery across human-infecting viruses.
• Identified specific PPIs related to viral replication (ZIKV-ESR1), oncogenic potential (HPVs), and evolutionary pressure on the human proteome.

Conclusions:

• P-HIPSTer effectively predicts viral-human PPIs, overcoming scalability limitations of experimental methods.
• The framework provides insights into virus-host interactions, cellular circuits, and evolutionary dynamics.
• P-HIPSTer aids in studying experimentally intractable viruses and identifying novel biological mechanisms.