Discussion of relaxation in documentation

[lorenzo-rovigatti]

Discussion of relaxation in documentation

[RodenLuo]

While waiting for it to be in the doc, the following helped me a lot during my learning.

Example input files

(A few weeks ago, I found a NEW_RELAX_PROCEDURE. But the commit was 2 years ago. I have not yet checked the difference between these and the above though.

Section 3.2 Relaxation of initial geometry of this aRxiv pdf. Other sections are also good on other topics. The link given in the Section 2.2 Files is broken. But I somehow found this link back then.

These are also quite helpful in general: oxViewer and oxDNA_analysis_tools NAR paper, A Primer on the oxDNA Model of DNA Frontiers paper, nature protocols paper.

[lorenzo-rovigatti]

I have added a page with some information. You can check it out here. Let me know what you think. Any suggestion is welcome!

[RodenLuo]

Hi Lorenzo, Thanks for putting them into the doc! It's clearer than my understanding before. But still some questions:

For MC relaxation, I have been using analysis/example_input_files/input_relax_MC for quite some time. Compared to the doc, the delta_translation and delta_rotation in the doc are about 5 - 10 times bigger. I guess this just means that if I stay with the example input, the movement will be smaller and the stresses are reduced slower. The max_backbone_force = 100 in the example input vs. max_backbone_force = 5 in the doc makes me unease. What does this entail and how should I choose? The same goes for the fact that max_backbone_force_far is missing in the example input.

As for MD relaxation, I have been using analysis/example_input_files/input_relax_MD for quite some time. I put the difference in the table below. Could you please kindly explain the differences or refer to some materials? Thanks

Parameter	Doc	Example input file
dt	0.002	0.0003
thermostat	langevin	bussi
diff_coeff	0.5	2.5
max_backbone_force	5	10
max_backbone_force_far	10	/

[lorenzo-rovigatti]

Hey Roden,

the meaning of these options are explained in the input file section of the documentation. In general, there is no best set of options: just experiment and see what works for your particular situation. Here are some more specific answers to your queries:

• The delta_translation and delta_rotation keys control the maximum displacement for MC move. There is no definite "best value" here, but I think it makes sense to decrease the value reported in the docs to 0.1.
• The bussi thermostat is a bit more stable, and in fact it's a good idea to use that one instead of the langevin (I'll update the docs!).
• Using max_backbone_force is the maximum force that the FENE potential can exert. I wouldn't expect much difference between 5 and 10.
• dt is the MD time step. Lowering it may make it possible to relax very stressed configurations, at the price of longer simulations.

[RodenLuo]

Hi Lorenzo, Thanks very much! Will read more of the doc.