On Tue, Nov 08, 2022, Reim, Thorben via AMBER wrote:
>
>For my PhD project, I need to do MD simulations of protein crystals. I
>found your tutorial where it is recommended to use at least 2 Unit
>Cells. But in the paper by Cerutti et al., (2010, ACS, Simulations of a
>Protein Crystal with a High Resolution X-ray Structure: Evaluation of Force
>Fields and Water Models), to which the tutorial refers, 12 unit cells were
>used.
>
>I am analyzing two different proteins with about 130 residues forming a
>unit cell of six and eight asymmetric units, respectively. Using 12 unit
>cells seems to be too computationally expensive, leading to my question:
>What is the "right" number of unit cells for simulating protein
>crystals? Is there a rule of thumb?
There is no "right" number. A lot depends on what you are trying to learn.
Using more unit cells tends to increase physical realism (since, at the
microscopic level, each unit cell is not identical to the next one at any
give time point.) It also gives you more protein chains to average over, so
you statistics are likely to improve. If you already have 6 to 8 chains per
unit cell, then you are correct in noting that you don't need as many unit
cells.
It can take a long time (such as hundreds of nanoseconds) to really
equilibrate a crystal simulation, even starting from the experimental
structure. So, try to use a system small enough to allow you to carry out
microsecond-length simulations, if possible. With as many chains per unit
cell as you have, even a single unit cell may well be sufficient.
....good luck....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 09 2022 - 08:00:02 PST
