VAPPD: Visual Analysis of Protein Pocket Dynamics
, Li Feng 1, Chuanbao Shi 1, Lina Cao 1, Yu Li 1, Yanfen Wang 3,* and Ximing Xu 4
Round 1
Reviewer 1 Report
The authors try to present their research on the dynamic analysis of binding pockets in proteins, but their presentation including English is bad and hard to read through (they use imperative forms many times but it is not appropriate for a paper). Their introduction of how to calculate a correlation (Eq.(5)) is awkward because it is a reverse tree, that is, we can not understand it until we read the later equation or explanation. They say, in the abstract, " in order to improve the accuracy of pocket similarity calculation" , but we don't know which part is improved. They use "accurate" several times but we don't know how accurate their method is. There are many methods or softwares to find binding pockets, so if the authors cannot figure out the points of improvement, it should be not published.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This manuscript explores the dynamic characteristics of protein pockets based on molecular dynamics (MD) simulation trajectories to analyze the intrinsic dynamic features of protein pockets. They use a representation method of dynamic pocket shape combined with topological features and propose a novel high-dimensional pocket similarity calculation method. After that, a visual analytics system is proposed to study the characteristics of high-dimensional dynamic pockets. Two case studies evaluate the efficacy of the approach. This work is novel and has important implications for biomolecules fields. Before formal acceptance, the authors should try to resolve the following issues:
1. There are many claims in the first paragraph of the introduction, which lack significant supports from the literature, which should be cited.
2. Lack details about the collaborating expert team: who are they? What are their expertise? And What are the limitations of the existing practices?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
1. Citation is missing in several places. e.g., two datasets used without citation.
2. Line 642, When comparing, different methods and different data are used for combination- its too generic
3. Include the objectives of the work
4. Line 741 -When the mouse clicks the arrow in the upper right corner of other feature views. where this click operation is performed?
5. Line 761-experts in this field evaluate - who are the experts? detail it.
6. Some images don't have clarity.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Their manuscript seems to be much improved but there still remains concerns about their method’s superiority. I said they should show their method is better than the previous ones and what they suggest seems to be tables 2 and 3. My questions are
- What is the ground truth (that is, experiment)? They compared their results with other computational methods and without ground truth or experiment, we cannot judge which is right or improved.
- Their comparison is focused on the method called “D3Pockets”, but is this the best method for binding pocket finding? What about other methods? For example, if this “D3Pockets” is not so good as the other better methods, it would not be a fair comparison. The authors should compare with the “best” previous methods.
They should show their method’s superiority more clearly in the manuscript, otherwise this paper should not be published in any form or in any journal.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
