Hi Danny,
I have written some code to calculate the flux of water through a channel in the past. Generally what I did was after aligning each frame of a trajectory to a reference set of coordinates of the protein, use 2 planes to split the simulation box into 3 sections (above the channel [1] , in the channel [2], and below the channel [3]). A tracer (in this case measured by the position of the OH2 atom of a water molecule) has made a complete transition if it sequentially passed either passed from [1] to [3] or from [3] to [1] having gone through [2]. This is necessary to ensure that you aren't counting events from [1] to [3] that went through the periodic boundary. The reference has been aligned so that the pore axis parallel to the z-axis.
You might have to modify this scheme based on the exact geometry of your system. Another thing to think about is that in order to accurately capture water motions, you need to carefully consider how frequently you are saving your coordinates to your trajectory output. For the systems I've worked on, this required frames being written every 1 ps, since at lower temporal resolutions you would miscount crossings (if I remember correctly moving to 10 ps sampling the counts were off by approximately 15-20%).
The code I wrote in python currently only has the ability to read Anton/Desmond and DCD formatted trajectories, but could easily be adapted to read netcdf4 from Amber as well. If you or anyone else is interested in the code, feel free to contact me and I'd be happy to work with you to adapt it to your particular system.
Josh
On Jul 22, 2012, at 12:44 PM, Danny Blue wrote:
> Dear Amber community,
>
>
> I'm working on a protein with a water channel. I'm wondering if is there a way to find out which water molecules have gone through the channel along its pore axis based on the trajectory file; and how should I define the pore axis of the channel?
>
> I have no experience in dealing with this kind of problem before, please give me some suggestions. Any suggestion will be great helpful! Thanks in advance!
>
> the best regards,
>
> your sincerely,
> Danny
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Jul 22 2012 - 11:30:03 PDT
