New approaches for unravelling reassortment pathways

Area of Science:

• Virology
• Computational Biology
• Epidemiology

Background:

• Influenza outbreaks pose significant public health risks due to recurring pandemics.
• Reassortant influenza strains, with composite genomic architectures, evade host immune systems, potentially causing pandemics.
• Identifying reassortant strains is vital for understanding viral evolution and reassortment pathways.

Purpose of the Study:

• Develop robust algorithms to detect reassortant influenza strains.
• Differentiate topological inconsistency caused by reassortment from phylogenetic error.
• Enhance the identification of pandemic-potential influenza subtypes.

Main Methods:

• Implemented two novel algorithms: MLreassort (maximum likelihood) and Breassort (Bayesian approach).
• Algorithms utilize significance of difference between phylogenetic trees and subtree pruning/regrafting operations.
• Focused on reassortment events identified by both methods.

Main Results:

• Successfully implemented and tested MLreassort and Breassort algorithms.
• Validated methods on simulated datasets and real viral data from Hong Kong (1999).
• Demonstrated the ability to robustly identify reassortment events.

Conclusions:

• Developed algorithms effectively identify reassortment events in small viral datasets.
• Methods are applicable to influenza and other segmented viruses.
• Further development is needed to address computational demands for larger datasets.