Construction of Multilevel Structure for Avian Influenza Virus System Based on Granular Computing

Area of Science:

• Molecular Ecology
• Medical Genetics
• Virology

Background:

• Influenza virus epidemics cause significant global morbidity and mortality.
• Rapid genetic variations in influenza viruses complicate subtyping, drug development, and vaccine design.
• Understanding the evolutionary structure of avian influenza viruses is crucial for public health.

Purpose of the Study:

• To construct the evolutionary structure of the avian influenza virus system.
• To develop an optimization model for determining virus system granularity and exploring subtype relationships.
• To identify effective granular virus signatures for efficient homologous virus detection.

Main Methods:

• Constructed evolutionary structure using hemagglutinin and neuraminidase protein fragments.
• Established an optimization model with a fuzzy hierarchical evaluation index to determine rational granularity.
• Developed an algorithm to extract the rational structure and identify granular virus signatures using a coarse-grained approach.
• Evaluated signature performance using a designed classifier.

Main Results:

• The proposed method successfully identified granular virus signatures that approximate and reflect the entire avian influenza virus system.
• The identified signatures are effective for homologous virus identification.
• The approach demonstrated efficiency in reducing systematic and computational complexity.

Conclusions:

• The developed method provides an effective way to identify avian influenza virus signatures.
• Hierarchical identification of homologous viruses is efficient upon detection of new molecular viruses.
• This research contributes to improved understanding and management of avian influenza.