All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Our reviewers are satisfied with the changes you introduced. I am glad to accept this manuscript for publication.
[# PeerJ Staff Note - this decision was reviewed and approved by Paula Soares, a PeerJ Section Editor covering this Section #]
The revised version of the manuscript can now be accepted for publication.
Additions made to the experimental design improved the overall comprehension of the methods e the design of the computational approach.
The findings are sound and grounded on excellent computational prediction capability.
None.
Our reviewers have some suggestions for the improvement of your manuscript.
[# PeerJ Staff Note: Please ensure that all review and editorial comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. #]
[# PeerJ Staff Note: The review process has identified that the English language must be improved. PeerJ can provide language editing services - please contact us at copyediting@peerj.com for pricing (be sure to provide your manuscript number and title) #]
The manuscript shows the application of the neural network model to predict the resistance of HIV protease inhibitors against different genotypes of the virus. It is clearly presented, with figures and tables providing evidence of the claims. The novelty of the approach is backed up by the state-of-the-art papers presented by the references cited throughout the text.
The method is briefly described, but the model and raw data are provided on a GitHub page, which other researchers can access to evaluate and access the findings. According to the journal's scope, the study can be classified as a biological application. The research question that the author tackled fills a gap in the scientific knowledge of HIV protease inhibitors, taking into account the high level of mutations for HIV.
The model's novelty resides in the analyses of pairs of drug resistance data and the development of a model that can identify potential cross-resistance profiles for unknown chemicals. The model is robust, with internal and external datasets used to validate it. The conclusion offers the significant findings discussed in the manuscript.
The language should be edited to remove some problems scattered along the manuscript, like:
line 242 ...threshold value of fold chance value is selected...
line 272 ... PC1 and PC2 as were able to represent...
line 278 Though same of the external set molecules were not forming clusters, still there were three clusters that contain three or more molecules.
line 296 ...connected to piperazine ring such instead of pyridine...
This is not meant to be an exhaustive list.
The manuscript is well prepared, in general.
However I have some comments and suggestions.
First I suggest that the authors double check English, ideally, with a fluent English speaker.
Second, I ask the authors to change some terms that they used in the manuscript:
- line 87: "genotype-fold" -> "genotype-phenotype"
-
1.Section: "Representation of Isolates"
I suggest that the authors provide detailed description of the binary barcoding technique, probably along with the example of real transformation of a sequence to the set of binary vector. What exactly are x1, x2, x3 etc in the set X = {x1, x2, x3...xn}. In other words, are there any differences between particular changes of an amino acid in each position with a mutation (M184V or M184I). Are specific positions of mutations are taken into account in x1, x2, x3... values?
2. Does obtaining prediction for a pair of protease inhibitors require usage chemical structure descriptors (Morgan fingerprints) for a pair of molecules? If yes then in which way were they generated?
3. I suggest that the authors provide the accuracy of classification of strains into resistant and susceptible based on sequence fingerprints only (i.e. excluding chemical structure fingerprints) and compare the accuracy of prediction for (1) combined descriptors and (2) sequence descriptors only.
In general, there are no significant violations in the findings.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.