Fast and robust genotyping of infectious bursal disease virus (IBDV) via alignment-free machine learning

Infectious bursal disease virus (IBDV) remains a major threat to the global poultry industry due to its immunosuppressive effects, economic impact, and extensive genetic diversity. Although genotype-based classification systems based on VP2 and VP1 sequencing have improved epidemiological understanding, their routine application in diagnostic laboratories is often limited by the need for high-quality sequences, curated reference datasets, and time-consuming phylogenetic analyses. In this study, we developed and validated an alignment-free, machine learning-based approach for rapid IBDV genogroup classification using commonly sequenced regions of segments A (VP2) and B (VP1). A curated dataset of 5,765 segment A and 1,503 segment B sequences, previously genogrouped by phylogenetic analysis, was used for training and validation. Sequences were encoded as 6-mer frequency profiles and analyzed using Random Forest (RF), Artificial Neural Network (ANN), and Support Vector Machine (SVM) classifiers. Model performance was assessed through repeated cross-validation and independent test datasets. RF and ANN achieved near-perfect performance (accuracy ≥ 0.997; Cohen's κ ≥ 0.995; AUC = 1.000), outperforming SVM. Robustness analyses showed that RF and ANN maintained high accuracy despite sequence degradation, tolerating up to 10% ambiguous bases and gaps, with RF showing greater stability under extreme conditions. Sliding-window analyses highlighted reduced performance outside validated genomic regions, emphasizing the importance of standardized targets. To facilitate practical application, the RF model was implemented in a web-based platform (GumboType; https://apps.maps.unipd.it/sequence/index.php ) enabling rapid genogroup assignment directly from FASTA files, without alignment or phylogenetic inference. Overall, this approach provides a fast, robust, and reliable alternative to traditional phylogenetic methods for IBDV genotyping, particularly suited to routine diagnostics and surveillance.

Authors

Institutions

Publication Details

Journal
Veterinary Research
Published
2026-11-18
DOI
https://doi.org/10.1186/s13567-026-01890-z
Primary Topic
Animal Virus Infections Studies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Fast and robust genotyping of infectious bursal disease virus (IBDV) via alignment-free machine learning

Matteo Legnardi, Giovanni Franzo, Mattia Cecchinato, Francesca Poletto et al.
Veterinary Research
Animal Virus Infections Studies
article

Fast and robust genotyping of infectious bursal disease virus (IBDV) via alignment-free machine learning

Matteo Legnardi, Giovanni Franzo, Mattia Cecchinato, Francesca Poletto, Claudia Maria Tucciarone, Cristina Andolfatto
article en

Abstract

Infectious bursal disease virus (IBDV) remains a major threat to the global poultry industry due to its immunosuppressive effects, economic impact, and extensive genetic diversity. Although genotype-based classification systems based on VP2 and VP1 sequencing have improved epidemiological understanding, their routine application in diagnostic laboratories is often limited by the need for high-quality sequences, curated reference datasets, and time-consuming phylogenetic analyses. In this study, we developed and validated an alignment-free, machine learning-based approach for rapid IBDV genogroup classification using commonly sequenced regions of segments A (VP2) and B (VP1). A curated dataset of 5,765 segment A and 1,503 segment B sequences, previously genogrouped by phylogenetic analysis, was used for training and validation. Sequences were encoded as 6-mer frequency profiles and analyzed using Random Forest (RF), Artificial Neural Network (ANN), and Support Vector Machine (SVM) classifiers. Model performance was assessed through repeated cross-validation and independent test datasets. RF and ANN achieved near-perfect performance (accuracy ≥ 0.997; Cohen's κ ≥ 0.995; AUC = 1.000), outperforming SVM. Robustness analyses showed that RF and ANN maintained high accuracy despite sequence degradation, tolerating up to 10% ambiguous bases and gaps, with RF showing greater stability under extreme conditions. Sliding-window analyses highlighted reduced performance outside validated genomic regions, emphasizing the importance of standardized targets. To facilitate practical application, the RF model was implemented in a web-based platform (GumboType; https://apps.maps.unipd.it/sequence/index.php ) enabling rapid genogroup assignment directly from FASTA files, without alignment or phylogenetic inference. Overall, this approach provides a fast, robust, and reliable alternative to traditional phylogenetic methods for IBDV genotyping, particularly suited to routine diagnostics and surveillance.

Veterinary ResearchVol. 57(1)
University of Padua (IT)
Good health and well-being, Zero hunger
Animal Virus Infections Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon