Vaxign2: the second generation of the first Web-based vaccine design program using reverse vaccinology and machine...

Area of Science:

• * Bioinformatics
• * Computational Biology
• * Vaccinology

Background:

• * Vaccination remains a cornerstone of modern medicine.
• * Reverse vaccinology (RV) is a genomic approach to identify potential vaccine candidates.
• * Existing RV tools require enhancement for improved prediction accuracy and user efficiency.

Purpose of the Study:

• * To update and enhance Vaxign2, a web-based vaccine design program.
• * To integrate a novel machine learning-based prediction method (Vaxign-ML) alongside the existing filtering approach.
• * To provide advanced post-prediction analysis tools for refining vaccine candidates and reducing analysis time.

Main Methods:

• * Development of Vaxign2, a comprehensive web server integrating predictive and computational workflows.
• * Implementation of Vaxign-ML, a machine learning model for predicting vaccine candidates.
• * Inclusion of post-prediction analysis modules for rational vaccine design.
• * Precomputation of results for approximately 1 million proteins across 36 pathogens.

Main Results:

• * Vaxign-ML demonstrated superior prediction performance compared to existing RV tools in benchmarking tests.
• * Vaxign2 provides enhanced capabilities for users to refine vaccine predictions based on specific rationales.
• * The server significantly reduces the time required for analyzing prediction results.
• * Successful application of Vaxign2 in analyzing SARS-CoV-2, the virus responsible for COVID-19.

Conclusions:

• * Vaxign2 offers a robust and comprehensive framework for rational vaccine design.
• * The integration of machine learning and advanced analysis tools improves vaccine candidate identification.
• * Vaxign2 supports accelerated and more effective vaccine development efforts.
• * The publicly accessible Vaxign2 platform facilitates broader research in vaccinology.