Understanding the host-pathogen evolutionary balance through Gaussian process modeling of SARS-CoV-2

Area of Science:

• Computational Biology
• Genomics
• Epidemiology

Background:

• Host-pathogen interactions are complex and dynamic, influenced by viral evolution.
• Understanding spatial-temporal mutational events is crucial for predicting disease dynamics.
• Existing methods may lack the predictive power for early detection of emerging viral threats.

Purpose of the Study:

• To develop a machine learning approach for tracking host-pathogen balance.
• To apply Gaussian process-based spatial covariance (GP-SCV) for analyzing viral genome evolution.
• To establish an early warning anomaly detection (EWAD) system for variants of concern (VOCs).

Main Methods:

• Utilized machine learning with Gaussian process (GP)-based spatial covariance (SCV).
• Applied GP-SCV to analyze daily, genome-wide SARS-CoV-2 variant patterns and pathology.
• Integrated GP-SCV with genome-wide co-occurrence analysis for anomaly detection.

Main Results:

• Demonstrated GP-SCV's ability to track spatial-temporal mutational impacts on host-pathogen balance.
• Showcased EWAD system's capability to anticipate changes in viral spread and pathology weeks ahead.
• Successfully identified signatures of emerging variants of concern (VOCs) using GP-based analyses.

Conclusions:

• GP-based SCV offers a novel method for monitoring viral evolution and host-pathogen dynamics.
• The EWAD system provides a powerful tool for early detection of significant viral variants.
• This approach advances our understanding of evolutionary paths and natural selection in biological systems.