VMD-L Mailing List

From: Neena Susan Eappen (neena.susaneappen_at_mail.utoronto.ca)
Date: Thu Sep 28 2023 - 14:34:14 CDT

Thank you so much Pawel and Josh!

I feel so enthusiastic reading both of your answers.
@Paweł Kędzierski<mailto:pawel.kedzierski_at_pwr.edu.pl> I understand there is no single best way
@Vermaas, Josh<mailto:vermaasj_at_msu.edu> I also really like figure 1 from 1st article you shared, and the 5th article is so useful for my thesis research as I characterize peptides in vacuum

A big thank you again! Honestly, no other forum gave me any insights on this question (pymol, gromacs, researchgate).

Regards,
Neena
________________________________
From: Vermaas, Josh <vermaasj_at_msu.edu>
Sent: Tuesday, September 26, 2023 9:06 AM
To: Paweł Kędzierski <pawel.kedzierski_at_pwr.edu.pl>; Neena Susan Eappen <neena.susaneappen_at_mail.utoronto.ca>; vmd-l_at_ks.uiuc.edu <vmd-l_at_ks.uiuc.edu>
Subject: Re: vmd-l: Total H-bond occupancy analysis

You don't often get email from vermaasj_at_msu.edu. Learn why this is important<https://urldefense.com/v3/__https://aka.ms/LearnAboutSenderIdentification__;!!DZ3fjg!80VP13KgLbzm7ux-xBjRx9BJWr4MDDf4-kdYYptjP_1_YfDACO0YYA5YmjGVPDqA7J3r14laOEQHTRmRMXm5GFD1GrN0dYcJjmz0$ >

One thing I’d encourage is to think about the geometry cutoffs you would use for a hydrogen bond. I think the VMD cutoffs by default are too stringent. I really like figure 1 from https://urldefense.com/v3/__https://www.pnas.org/doi/full/10.1073/pnas.1308560110__;!!DZ3fjg!80VP13KgLbzm7ux-xBjRx9BJWr4MDDf4-kdYYptjP_1_YfDACO0YYA5YmjGVPDqA7J3r14laOEQHTRmRMXm5GFD1GrN0dWJTRknQ$ , which argues for a 30 degree cutoff (cos 30 ~ 0.85), and figure 5 from https://urldefense.com/v3/__https://pubs.aip.org/aip/jcp/article/141/21/214507/351929/The-structure-of-water-around-the-compressibility__;!!DZ3fjg!80VP13KgLbzm7ux-xBjRx9BJWr4MDDf4-kdYYptjP_1_YfDACO0YYA5YmjGVPDqA7J3r14laOEQHTRmRMXm5GFD1GrN0dZ3O38Wn$ I think makes the argument of a 3.2 or 3.3 Angstrom heavy atom distance. Once you’ve picked a cutoff you can live with, and measured how many hydrogen bonds there are in the system, you have some choices to make. Pawel’s suggestion is a good one, but will only give you only the enthalpic contributions, since that is what classical force fields give you directly when you are measuring energy. To also account for the loss in entropy that occurs once a hydrogen bond is formed, you would either do a much more complicated simulation (in https://urldefense.com/v3/__https://pubs.acs.org/doi/full/10.1021/acs.jpcb.5b00778__;!!DZ3fjg!80VP13KgLbzm7ux-xBjRx9BJWr4MDDf4-kdYYptjP_1_YfDACO0YYA5YmjGVPDqA7J3r14laOEQHTRmRMXm5GFD1GrN0dQVrLyhT$ , I used TI calculations, which in Figure 7 I can compare directly to the number of hydrogen bonds in equilibrium), or you just rely on “rules of thumb” to get an approximate change in free energy. https://urldefense.com/v3/__https://pubmed.ncbi.nlm.nih.gov/8889177/__;!!DZ3fjg!80VP13KgLbzm7ux-xBjRx9BJWr4MDDf4-kdYYptjP_1_YfDACO0YYA5YmjGVPDqA7J3r14laOEQHTRmRMXm5GFD1GrN0dcjqJ6gq$ is probably the most relevant to your case, and sells you 2.2kcal/mol of stabilization per extra hydrogen bond. Table 1 from https://urldefense.com/v3/__https://www.pnas.org/doi/10.1073/pnas.2133366100__;!!DZ3fjg!80VP13KgLbzm7ux-xBjRx9BJWr4MDDf4-kdYYptjP_1_YfDACO0YYA5YmjGVPDqA7J3r14laOEQHTRmRMXm5GFD1GrN0dREOCoT3$ will tell you that a hydrogen bond is worth ~1.6 or 1.9kcal/mol in solution in specific secondary structural elements.

-Josh

From: <owner-vmd-l_at_ks.uiuc.edu> on behalf of Paweł Kędzierski <pawel.kedzierski_at_pwr.edu.pl>
Organization: Wrocław University of Science and Technology
Date: Tuesday, September 26, 2023 at 8:05 AM
To: Neena Susan Eappen <neena.susaneappen_at_mail.utoronto.ca>, "vmd-l_at_ks.uiuc.edu" <vmd-l_at_ks.uiuc.edu>
Subject: Re: vmd-l: Total H-bond occupancy analysis

Dear Neena,

There is nothing like a single "best way", only a multitude of trade-offs depending on your needs and constraints.

For example, if the accuracy is paramount, the "best" way would be to calculate the interaction energies on adequate quantum level but it would be extremely costly.

The quickest evaluation could be to just measure the H...Y distance in a X-H...Y H-bond and this alone would correlate with the H-bond strength.

In VMD, you may also get the force field approximation of the H-bond interaction energy over a trajectory by using the "NAMD Energy" plugin and using the donor and the acceptor groups as two interacting selections; this is what I would advice to try.

Cheers,

Pawel

W dniu 25.09.2023 o 21:14, Neena Susan Eappen pisze:

Hello VMD users,

This is not a question related to VMD, but I was wondering if someone can help me or guide me to the appropriate resource.

I extracted two types of H-bond occupancy data (donor: Lysine side chain and backbone amide NH, with backbone carbonyl CO as acceptor) from two different trajectories of the same peptide, centroid structures from predominant cluster of each trajectory shown in each row. These structures just differ by dihedral angles across a few residues.

Question: What is the best way to quantify relative strength of total intramolecular hydrogen bonding between these two trajectories?

Many thanks,

Neena Eappen

Graduate Student

Jockusch Lab<https://urldefense.com/v3/__http:/www.chem.utoronto.ca/wp/jockusch/__;!!DZ3fjg!7BQq4_UfJq0ad40jwPnxqL3G06aRbBD6j4Fx_-6-SB-yTf7vn4luF0F2P9KDJbOXZEMOpLXymgYArwst6Lr_uGo_2z-XOFwjY6MG$>, U of T