alchemistry / alchemical-best-practices

Best practice document for alchemical free energy calculations going to livecoms journal
Creative Commons Attribution 4.0 International
63 stars 18 forks source link

Suggestions #91

Closed yunhuige closed 4 years ago

yunhuige commented 4 years ago

Hi,

I read the preprint on bioRxiv (https://arxiv.org/abs/2008.03067) and have following suggestions on this version:

1/ It seems Fig 9 and 12 are identical. Based on the captions, I believe Fig 9 is not the supposed figure.

2/ Some figures are far from where they are referred in the text. (e.g., Fig 5-7). I personally feel annoying if I have to scroll up and down multiple times to check figures when I read them in the paper.

3/ Is that possible to talk more about hybrid topology? Or a figure to explain how does that work? I kind of know the answer but just think people may be interested in it.

4/ Is that possible to include an example of the practical single/dual topology files as part of SI? As a GROMACS user I'm particularly interested in this. Maybe just for a transformation of benzene to benzyl alcohol as illustrated in Fig 3. The detailed parameters may not be necessary but it will be good to show the actual format differences between single and dual topologies. I think it will benefit people (at least GROMACS users) in learning how to perform these simulations in practice.

I really like this paper as it includes so many detailed explanations and suggestions on this technique. Above are my personal feelings after reading the current version. Some may not be feasible but I hope any of these could contribute to a better version of this paper.

Thank you!

ppxasjsm commented 4 years ago

Hey,

1/ It seems Fig 9 and 12 are identical. Based on the captions, I believe Fig 9 is not the supposed figure.

Thanks for the spot. This is due to the way the arxiv submission doesn't allow pdf uploads. I'll fix this on the arxiv.

2/ Some figures are far from where they are referred in the text. (e.g., Fig 5-7). I personally feel annoying if I have to scroll up and down multiple times to check figures when I read them in the paper.

I am afraid this cannot easily be fixed due to the fact that the paper is written in LaTeX and has automatic figure placement. I can try and improve this with proofs for a published version.

3/ Is that possible to talk more about hybrid topology? Or a figure to explain how does that work? I kind of know the answer but just think people may be interested in it.

Yes! Ideally if you have a pragraph you would like to add please suggest it and a place in the document where it can go and I'll happily add it.

4/ Is that possible to include an example of the practical single/dual topology files as part of SI? As a GROMACS user I'm particularly interested in this. Maybe just for a transformation of benzene to benzyl alcohol as illustrated in Fig 3. The detailed parameters may not be necessary but it will be good to show the actual format differences between single and dual topologies. I think it will benefit people (at least GROMACS users) in learning how to perform these simulations in practice.

I like this suggestion. At the moment we don't really have a SI and the paper itself is already a monster. I'll add it to a list of ideal future improvements, but probably wont' fix it immediately.

I really like this paper as it includes so many detailed explanations and suggestions on this technique. Above are my personal feelings after reading the current version. Some may not be feasible but I hope any of these could contribute to a better version of this paper.

Thank you for all your feedback! Much appreciated.

ppxasjsm commented 4 years ago

I will add you to the acknowledgements. Is it ok for you to appear as Yunhui Ge?

yunhuige commented 4 years ago

Thank you for your reply!

I am afraid this cannot easily be fixed due to the fact that the paper is written in LaTeX and has automatic figure placement. I can try and improve this with proofs for a published version.

This is a minor issue and basically my personal feeling. No worries!

Yes! Ideally if you have a pragraph you would like to add please suggest it and a place in the document where it can go and I'll happily add it.

Well, to be honest I don't have much experience in this area. But I recently joined the Mobley lab as a new postdoc and I did some literature review. One of the two work cited in the hybrid topology paragraph is Dr. Mobley's work and I happened to read it and had some ideas about how it worked. So I'm not sure if I am the right person who can contribute more to this topic but I'd like to try my best. Maybe I should ask Dr. Mobley and see if he has any suggestions first.

I like this suggestion. At the moment we don't really have a SI and the paper itself is already a monster. I'll add it to a list of ideal future improvements, but probably wont' fix it immediately.

Sounds good! If it can be added in the final version, I think it will be very helpful as people like me always have similar questions when I start to learn how to perform alchemical FEP simulations using GROMACS.

I will add you to the acknowledgements. Is it ok for you to appear as Yunhui Ge?

Yes, that is fine. You really don't have to do this since I only gave some suggestions.

mrshirts commented 4 years ago

Yes, that is fine. You really don't have to do this since I only gave some suggestions.

Acknowledgements is exactly appropriate for these sorts of suggestions for LiveCoMS papers!

yunhuige commented 4 years ago

I talked to Dr. Mobley and I will prepare a draft talking about the hybrid topology idea (and figures if necessary). Then I will pass it to Dr. Mobley for edits. Sound good? @ppxasjsm

ppxasjsm commented 4 years ago

Yes sounds great! Thanks for this.

ppxasjsm commented 4 years ago

@yunhuige, how are you getting on with this? We've had revisions back for version 1.0 and it would be great to also incorporate your edits.

yunhuige commented 4 years ago

@ppxasjsm I prepared my draft and have sent it to Dr. Mobley. I will ask him for any updates today.

ppxasjsm commented 4 years ago

Great! Thank you.

davidlmobley commented 4 years ago

We've actually realized that this is super confusing in the literature, and there may not be a "correct" terminology, which has slowed me down on this point. I'll put it into the queue with the other issues you've assigned me.

davidlmobley commented 4 years ago

OK, so @yunhuige and I discussed the issue of single/hybrid/dual/separated topologies some more and I dug in to the literature a bit and have revised the relevant section (7.1.1) to clarify/make it more correct (for now on Overleaf). I think mainly we were using a term "hybrid" that I've not seen used except for by Gapsys, de Groot etc., to refer to something I called "separated topologies". So I've added some text clarifying what "hybrid" means very briefly, then clarified/mentioned "separated topologies" and otherwise tried to clarify that section.

@ppxasjsm do you want to save item (4) in a separate issue and we can close this one?

ppxasjsm commented 4 years ago

Yes that would be great thanks!