Comprehensive Evaluation of End-Point Free Energy Techniques in Carboxylated-Pillar[6]arene Host-Guest Binding: IV. The QM Treatment, GB Models and the Multi-Trajectory Extension
Round 1
Reviewer 1 Report
The authors evaluate the performance of the QM/GBSA approach in its one-trajectory and three-trajectory protocols by using different semiempirical Hamiltonians to describe the QM region to compute the binding free energy in a series of host-guess complexes. They found that for the one-trajectory approach the inclusion of a QM description does not significantly improve the results, while the for the three-trajectory protocol the use of DFTB is better than using a classical MM/GBSA approach.
I found the paper very interesting, and the results are useful for the computational chemistry community which employes the widely extended MM(QM)GBSA approach to compute binding free energies, especially for ligand/protein systems. In addition, I recognize the great effort performed by the authors to have a good sampling, especially for the free host. However, I think the discussion of the results is very short and should be extended a bit to have a larger number of interesting and useful conclusions.
My major concern is that the results shown in Tables 2 and 3 are barely discussed. I think that further comparison can be made. For example, comparison among solvent models, or the performance of DFTB2 vs DFTB3 or the effect of including the DH+ correction in the PM6 and AM1 models.
Minor concerns:
1. The conclusions should be shortened. The first two paragraphs are not needed, or they can be moved to the introduction, although most of the content is already in the introduction. The conclusions could start in the third paragraph “Int this work, in order …”.
2. In the section of “Model Construction and Configurational Sampling”, it would be good to explicitly show the systems that have been used for the calculations. It is said that they are WP6 hosts and 13 guests (G1 to G13) from a GitHub repository, but they are never shown in the manuscript.
3. It would be nice to know also whether the GPU or CPU version of pmemd has been used.
4. In the results section, before the discussion of the different quality metrics, it would be useful for the reader to explain the meaning of all of them, especially, the Kendall coefficient, the Peerlman’s predictive index and the Pearson correlation coefficient.
5. The conclusion in page 5 “we generally do not recommend applying QM/GBSA in the popular single trajectory end-point free energy calculation” is, in my opinion, a bit strong since the authors have performed calculations only for a small set of systems. I agree that the approach is likely not good for class of systems investigated here, but this cannot be generalized.
Only minor corrections are needed.
Author Response
see pdf response
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The manuscript focuses on the fundamental question how the applicability of the end-point free energy technique can be improved to obtain reliable information on the thermodynamics of inclusion complex formation. The modifications of the calculation methods were tested using carboxylated pillar[6]arene host and various guests. The paper is well-written. The assessment of the predictive power of the various calculation methods is likely to be of interest to members of wide scientific community. The prediction of the relative binding affinities of various guests to a host would be important for the rational design of tailor-made inclusion complexes in various applications. I recommend the publication of the manuscript in the journal Liquids after the Authors considered the following remarks:
1. The structural formula of the 13 guests and WP6 host should be given in the manuscript or in the supporting information.
2. It would be worth including a comment whether the conclusions are expected to be valid for the inclusion complexes of the much less flexible cucurbit[n]uril hosts.
Author Response
pdf response letter
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
I am happy with the modifications made by the authors. My only minor concern is that the text with the energies in Figure 1a in very small and also the molecular structures are made with very thin lines. Other than that, the manuscript deserves to be published.
Author Response
We've adjusted Fig. 1a to enlarge the font size and thicken the line width.
Reviewer 2 Report
The manuscript has been sufficiently improved to warrant publication in Liquids.
Author Response
Thank you for your review.
