tstate definitions for 2D Weighted ensemble

23 views
Skip to first unread message

Praveen Ranganath Prabhakar

unread,
Feb 6, 2025, 5:54:12 PMFeb 6
to westpa...@googlegroups.com
Hi, 

I am running a weighted ensemble simulation of a protein with 2D progress coordinates, I just have a small query on how the input in tstate.file should be. For the recycling of walkers, after the target state is reached, do we have to define the target state with respect to both dimensions  (Or is this up to the user) ?

For example, the tstate.file can have,
 
unfolded 10.0 8.0

or just
 
unfolded 10.0 

In the latter case I believe it will by default take the first progress coordinate. Which one is recommended :)?

Thanks in advance. 

--
Praveen Ranganath Prabhakar
Andricioaei group,
Department of Chemistry, 
UC Irvine

Praveen Ranganath Prabhakar

unread,
Feb 6, 2025, 6:28:11 PMFeb 6
to westpa...@googlegroups.com
Hi, 
 
Sorry missed typing the entire email, 

I am running a weighted ensemble simulation of a protein with 2D progress coordinates, I just have a small query on how the input in tstate.file should be. For the recycling of walkers, after the target state is reached, do we have to define the target state with respect to both dimensions  (Or is this up to the user) ?

For example, the tstate.file can have,
 
unfolded 10.0 8.0

or just
 
unfolded 10.0 

If I am understanding it correctly, in the first case, the recycling would happen only if the first progress coordinate is 10 and the second progress coordinate is 8. In the latter case I believe it will by default take the first progress coordinate and recycling will happen when the first progress coordinate is 10 and the second progress coordinate can be any value.

Thanks in advance. 

Goldberg, Alexander

unread,
Feb 6, 2025, 9:52:49 PMFeb 6
to westpa-users
Hi Praveen,

You will need to define the target state for both dimensions. For the tstate.file containing "unfolded 10.0 8.0", you will recycle anything that enters the bin containing 10.0, 4.0 at the last frame of the WE iteration. The second case likely will not work because you will need to define the value for the second progress coordinate. If you want to label more bins, you can extend your tstate.file by adding additional lines beneath your initial tstate definition: 

unfolded 10.0 8.0
unfolded 10.0 4.0

Best,
Alex

Reply all
Reply to author
Forward
0 new messages