Next Article in Journal
Phylogenetic and Timescale Analysis of Barmah Forest Virus as Inferred from Genome Sequence Analysis
Next Article in Special Issue
Elimination of Hepatitis B in Highly Endemic Settings: Lessons Learned in Taiwan and Challenges Ahead
Previous Article in Journal
Beyond Cholera: Characterization of zot-Encoding Filamentous Phages in the Marine Fish Pathogen Vibrio anguillarum
Previous Article in Special Issue
Molecular, Evolutionary, and Structural Analysis of the Terminal Protein Domain of Hepatitis B Virus Polymerase, a Potential Drug Target
 
 
Article
Peer-Review Record

In Silico Prediction of Human Leukocytes Antigen (HLA) Class II Binding Hepatitis B Virus (HBV) Peptides in Botswana

Viruses 2020, 12(7), 731; https://doi.org/10.3390/v12070731
by Wonderful Tatenda Choga 1,2, Motswedi Anderson 1, Edward Zumbika 3, Bonolo B. Phinius 1, Tshepiso Mbangiwa 1,2, Lynnette N. Bhebhe 1, Kabo Baruti 1,4, Peter Opiyo Kimathi 5, Kaelo K. Seatla 1,6, Rosemary M. Musonda 1,7, Trevor Graham Bell 8, Sikhulile Moyo 1,7, Jason T. Blackard 9 and Simani Gaseitsiwe 1,7,*
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Viruses 2020, 12(7), 731; https://doi.org/10.3390/v12070731
Submission received: 12 June 2020 / Accepted: 3 July 2020 / Published: 6 July 2020
(This article belongs to the Special Issue Hepatitis B Virus: From Diagnostics to Treatments)

Round 1

Reviewer 1 Report

The authors have carefully revised the manuscript and provided enough descriptions for all figures and tables.

Reviewer 2 Report

Choga et. el. used in-silico methods to infer 15-mer peptides, based on HBV genome sequences extracted from GeneBank, that can bind to HLA class II alleles commonly expressed in the Botswana population, and studied them in detail.

This revised version of the manuscript is well written and have addressed all my previous concerns. I have no further comments.

Reviewer 3 Report

In recent years, bioinformatics has greatly facilitated vaccine design as it is safer, more convenient, more efficacious, and less expensive than traditional vaccines selection. The current study is of such kind, in which the author selected HBV surface antigen and HLA class II for peptide design. The author concluded in the end that they have identified antigenic and genotype-specific peptides suitable for vaccine as well as diagnostic kits development. However, to the reviewer, the study is still preliminary, and the conclusion not only poorly supported by the results but also failed to fulfill the stated aim.

 

First, the paper is not well written. Poor logic flow, language redundancy and multiple grammar mistakes made the reading very difficult.

 

The major scientific concerns include:

  1. It is not clear why surface antigen and HLA II were selected for the study (why not core protein or B cell epitope or HLA I)? It looked as if the author wanted to focus on improving the design of preventive vaccine. However, aa mutation in the context of failed HBV clearance which mediated mainly by CD8 were mentioned a lot, making the rationale of the study design unclear.
  2. The author should provide immunogenicity predication of the predicted epitopes.
  3. The toxicity analysis of the predicted epitope is insufficiently based on BLAST search of NCBI protein database (mainly sequence match). More advanced method should be used as the recognition of host tissue by the specific immune response induced by the epitope involves physical-chemical properties of the two.
  4. Regarding the identified candidate vaccine, how is their molecular docking to HLAII like? Can the author verify experimentally if they can form stable HLA-peptide complex in vitro?
Back to TopTop