Reviewer 2 comments

[labarba]

The authors present an implementation for a numerical solver specifically design to tackle continuum electrostatics models. Specifically, the authors present a solver for the Poisson-Boltzmann equation for a set of charges found in biomolecules. The manuscript is well written, and the technical and scientific hypothesis are sound therefore it is worthy of publication; however a few concerns were raised by its current form:

Major concerns

• [x] 1. In the introduction, the authors claim that simulations of viruses are limited to a few elite researchers; this is simply not true. While large-scale simulations of entire viruses at atomistic remain a niche field, several groups have been able to conduct state-of-the-art virus research that have revealed novel biology using XSEDE resources and campus clusters. There seems to be fine line in the computational virology community in terms of what studies constitute a computational benchmarks and what studies reveal novel biology. There are several reviews available on this topic from various authors.

• [x] 2. The simulations presented in the introduction serve as computational benchmarks (e.g., Arkhipov 2007, Durrant 2020), but the biological impact of these papers is questionable. There are several computational studies on the biology of viruses that are far more relevant to the present manuscript and to the virology field at large, none of which are cited.

• [x] 3. To the casual reader the present manuscript does not communicate why another Poisson-Boltzmann solver is needed by the community – it is clear to this reviewer that this is a useful and much needed implementation. However, the computational biophysics field has trusted APBS for several years due to its ease of use, reliability, and performance. There is not a direct comparison between the results from the presented software and APBS. Furthermore, the authors seem to use the APBS suite themselves as they use the PDB2PQR tool to generate the partial charges as the input in their software.

• [x] 4. Like the previous point the determination of the molecular surface in the proposed approach is dependent on another external package, namely NanoShaper. Although NanoShaper is poised for FEM calculations of electrostatics; APBS includes its own molecular surface determination package that is widely used. More on this below.

• [x] 5. To address both 3 and 4, the authors should make sure that the needed binding to these external packages is part of their software.

• [x] 6. The authors use as benchmarks crystallographic and cryoEM structures. Another important technique for structure determination is NMR. The latter has the advantage of yielding ensembles of structures that are statistically independent. The authors should perform solvation free energy calculations on each structure of the ensemble, to establish whether their formulation is sensitive enough to small changes in the structure of the protein which in often cases result in large changes in free energy of solvation.

Minor concerns:

• [x] 1. The figure in 4 has two different size scales for a same sized particle. It would be more visually appealing if the two renders of the virus had the same size.
• [x] 2. The units in the scale of the potential in Figure 4b are missing.
• [x] 3. Details about the molecular surface are lacking. These are important as these parameters determine the solvent exposed area of the system.

[labarba]

item 1

In the introduction, the authors claim that simulations of viruses are limited to a few elite researchers; this is simply not true. While large-scale simulations of entire viruses at atomistic remain a niche field, several groups have been able to conduct state-of-the-art virus research that have revealed novel biology using XSEDE resources and campus clusters. There seems to be fine line in the computational virology community in terms of what studies constitute a computational benchmarks and what studies reveal novel biology. There are several reviews available on this topic from various authors.

We appreciate the reviewer pointing us to literature on coarse grained (CG) models, which also play an important role in making virus-level simulations more accessible. We added the corresponding references (and other) on the topic in the introduction, when we review MD simulations for viruses, highlighting the role of CG models.

• fixed in bf78986

See references listed by the editor in https://github.com/barbagroup/bempp_exafmm_paper/issues/13

Reddy, Sansom 2016; Huber et al. 2021: reviews coarse-grained approaches for virus. We have added this in the introduction. Tarasova, Nerukh 2018; Hadden, Perilla 2021: reviews all-atom MD, they need large supercomputers

[labarba]

item 2

The simulations presented in the introduction serve as computational benchmarks (e.g., Arkhipov 2007, Durrant 2020), but the biological impact of these papers is questionable. There are several computational studies on the biology of viruses that are far more relevant to the present manuscript and to the virology field at large, none of which are cited.

The reviewer here is asking us to highlight computational studies with biological impact, which is a fair point, even though our paper does not focus on the biological aspects. We now mention this in the introduction, with the corresponding citations.

• fixed in e00591f and 67b9b6c

[labarba]

item 3

To the casual reader the present manuscript does not communicate why another Poisson-Boltzmann solver is needed by the community – it is clear to this reviewer that this is a useful and much needed implementation. However, the computational biophysics field has trusted APBS for several years due to its ease of use, reliability, and performance. There is not a direct comparison between the results from the presented software and APBS. Furthermore, the authors seem to use the APBS suite themselves as they use the PDB2PQR tool to generate the partial charges as the input in their software.

We have included a new set of results comparing with the community-wide trusted codes APBS and MIBPB, as validation. As the reviewer points out, our implementation stands out from other PB solvers in its ease of use and flexibility, as it runs directly on a Jupyter notebook. We make this point in the discussion section.

[labarba]

item 4

Like the previous point the determination of the molecular surface in the proposed approach is dependent on another external package, namely NanoShaper. Although NanoShaper is poised for FEM calculations of electrostatics; APBS includes its own molecular surface determination package that is widely used. More on this below.

In the most recent versions, the molecular surface determination libraries in APBS are MSMS and Nanoshaper. Both of these packages are widely used for Poisson-Boltzmann calculations.

[labarba]

item 5

To address both 3 and 4, the authors should make sure that the needed binding to these external packages is part of their software.

• APBS: We now include comparisons, as requested in item 3. We cannot add bindings to run APBS (which is not a dependency of our computational workflow) seamlessly in our platform.
• Nanoshaper: Our approach to reproducibility is to deposit the final mesh in a data archive, and share that with readers. We added to our repository a script that runs Nanoshaper, as an example.

We have added the script to generate surface meshes using Nanoshaper in our repro-pack, see /bempp_exafmm_paper/repro-pack/bempp_pbs/scripts/generate_mesh.py.

[labarba]

item 6

The authors use as benchmarks crystallographic and cryoEM structures. Another important technique for structure determination is NMR. The latter has the advantage of yielding ensembles of structures that are statistically independent. The authors should perform solvation free energy calculations on each structure of the ensemble, to establish whether their formulation is sensitive enough to small changes in the structure of the protein which in often cases result in large changes in free energy of solvation.

The Zika virus structure is not available as an NMR ensemble, so this review request is not feasible to execute. In the case of the small molecule (5PTI), an NMR ensemble is available, but in this case, the result is a grid convergence analysis, which should be done with a unique structure. The quantification of uncertainty in the solvation energy due to the variations in an ensemble of structures would be an interesting study. It is not in the scope of this paper, which is focused on the computational methods and workflow, rather than biological or chemical observations. Our workflow, on the other hand, would facilitate such a study, where an identical computation could be automatically executed on an ensemble of structures, programmatically and reproducibly.

[labarba]

minor concern 1

The figure in 4 has two different size scales for a same sized particle. It would be more visually appealing if the two renders of the virus had the same size.

We disagree. The colored-chain visualization has coarse detail, and it would waste too much space to make this figure much larger. The surface potential, on the other hand, has much finer detail, and a large figure is needed to visualize it. The comparison of the two is easy to make, despite the difference in display scale.

[labarba]

minor concern 2

The units in the scale of the potential in Figure 4b are missing

• Fixed in 2c08b5c and a3f7df0.

[labarba]

minor concern 3

Details about the molecular surface are lacking. These are important as these parameters determine the solvent exposed area of the system.

We provide the actual meshes for reproducibility, but we can add/mention the total surface area, and probe size used to generate the mesh (mesh density is already mentioned).

• Fixed in 055a398.