FES from reweighting with two CVs OPES simulation

615 views
Skip to first unread message

BMK

unread,
Aug 3, 2022, 12:00:28 PM8/3/22
to PLUMED users

Dear all,

I tried to perform the OPES tutorial using both the x and y positions as CVs. The FES that I obtained with the FES_from_State.py script (fes_40.png) looks reasonable, but the FES from FES_from_Reweighting.py with sigma values from the last KERNELS entry looks quite weird (fes-rew.png). Can someone help me explain what I did wrong?

The command that I used to obtain the latter FES was: “../FES_from_Reweighting.py --kt 1 -s 0.067,0.046 --blocks 5 --deltaFat 0 --cv p.x,p.y”

The plumed input/output and a crude plotting script are attached to this post.

Regards, Bas

plumed.dat
plot_fes2d.py
fes-rew.png
opes_output.tar.gz
fes_40.png

Michele Invernizzi

unread,
Aug 3, 2022, 3:58:05 PM8/3/22
to plumed...@googlegroups.com
Dear Bas,

You did everything correctly, it is just a problem of visualization. The FES from reweighting has no upper bound, but you should only visualize it below a reasonable value, e.g. 16, since that is the region you actually care about. You can for example add the following line to your plotting script: Z[Z>16] = 16

I also suggest removing an initial transient before the reweighting, it will get rid of the small artifacts around x=0. You just have to add to FES_from_Reweighting.py the following option: --skiprows 100. You can try a few different values (50, 1000, ?) and see how the FES changes, it should be pretty stable.

Best,
Michele


--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/0f17aef1-43ad-4f7d-af11-cb1146776a7fn%40googlegroups.com.

BMK

unread,
Aug 4, 2022, 5:23:46 AM8/4/22
to PLUMED users
Hi Michele,

Thank you for the quick reply!

Changing the upperbound to 20 instead of 16 resulted in a bit more detail in the FES, and skipping the first 50 lines was enough to remove the artifacts.

Regards, Bas

Op woensdag 3 augustus 2022 om 21:58:05 UTC+2 schreef inve.m...@gmail.com:
Reply all
Reply to author
Forward
0 new messages