Genotyping hepatitis B virus dual infections using population-based sequence data.

Area of Science:

• Virology
• Computational Biology
• Genetics

Background:

• Hepatitis B virus (HBV) genotypes A-H influence disease progression and treatment sensitivity.
• Existing computational genotyping methods struggle with HBV dual infections, particularly those with different genotypes.
• Ambiguous sequence positions complicate accurate HBV genotype classification.

Purpose of the Study:

• To develop and validate a computational model for identifying and genotyping inter- and intragenotype HBV dual infections.
• To improve the accuracy of HBV genotyping in clinical settings, especially for complex coinfection cases.
• To enhance the optimization of antiviral therapy regimens through precise HBV genotype identification.

Main Methods:

• Development of a novel computational model for dual HBV infection detection.
• Verification of the model using synthetic data, assessing accuracy and sensitivity.
• Screening of 241 patient sera samples for dual HBV infections.
• Confirmation of predictions using clonal experiments on patient-derived material.

Main Results:

• The model achieved 100% accuracy for intergenotype dual infections on synthetic data.
• Sensitivity for intragenotype dual infections was 36.4% on synthetic data.
• Eight putative intergenotype (e.g., A-D, A-G, D-G) and four intragenotype (e.g., A-A, D-D, E-E) dual infections were identified in patient sera.
• Clonal experiments confirmed three out of three predicted dual infections.

Conclusions:

• The developed computational model effectively identifies and genotypes HBV dual infections, including inter- and intragenotype cases.
• The method enhances the precision of HBV genotype classification, addressing limitations of previous approaches.
• Integration into the geno2pheno([hbv]) web service provides a valuable clinical tool for optimizing HBV antiviral therapy.