Bioinformatics methods for the analysis of hepatitis viruses

A new hepatitis B virus e antigen-negative strain gene used as a reference sequence in an animal model

Infection with hepatitis B virus (HBV) e-antigen (HBeAg)-negative strains is increasingly prevalent. Currently, detailed information of the obtained natural HBV strain is not available except for the B genotype and HBeAg-negative. The aim of the present study was to characterize the natural genetic variation of the HBeAg-negative strain and investigate its function. The genic sequence was determined using Sanger sequencing, and compared to related sequences using alignment and phylogenetic analysis. In vivo, virus-specific serum markers were investigated in CBA/CaJ mice. The sequence had a full genome length of 3215 nucleotides. Sites 122, 125, 127, and 160 in S regions were identified as lysine, threonine, proline, and lysine respectively. The main four point variants including A1762T, G1764A, G1896A, and G1899A were detected in the full-length genome. The genotype of the sequence was B, with sub-genotype B2 and serological subtype adw2. The characterize of the natural genetic variation strain showed no reported drug-resistant variant in P region and no reported immune escape site in S region. The strain will increase viral replication and infection for mutations A1762T and G1764A in the basal core promoter region, and mutations G1896A and G1899A in the pre-core region. The G1896A variant resulted in a premature stop codon and abolished HBeAg expression. HBsAg persisted for 26 weeks and HBeAg was still negative in CBA/CaJ mice. The present sequence is representative of the HBeAg-negative genome and may serve as a valuable reference for studying HBeAg-negative strains. The present findings were successfully verified in CBA/CaJ mice, demonstrating good applicability of the sequence.

A computational chemistry perspective on the current status and future direction of hepatitis B antiviral drug discovery

Computational chemical biology, applied to research on hepatitis B virus (HBV), has two major branches: bioinformatics (statistical models) and first-principle methods (molecular physics). While bioinformatics focuses on statistical tools and biological databases, molecular physics uses mathematics and chemical theory to study the interactions of biomolecules. Three computational techniques most commonly used in HBV research are homology modeling, molecular docking, and molecular dynamics. Homology modeling is a computational simulation to predict protein structure and has been used to construct conformers of the viral polymerase (reverse transcriptase domain and RNase H domain) and the HBV X protein. Molecular docking is used to predict the most likely orientation of a ligand when it is bound to a protein, as well as determining an energy score of the docked conformation. Molecular dynamics is a simulation that analyzes biomolecule motions and determines conformation and stability patterns. All of these modeling techniques have aided in the understanding of resistance mutations on HBV non-nucleos(t)ide reverse-transcriptase inhibitor binding. Finally, bioinformatics can be used to study the DNA and RNA protein sequences of viruses to both analyze drug resistance and to genotype the viral genomes. Overall, with these techniques, and others, computational chemical biology is becoming more and more necessary in hepatitis B research. This article forms part of a symposium in Antiviral Research on "An unfinished story: from the discovery of the Australia antigen to the development of new curative therapies for hepatitis B."