[gmx-users] Entropy

[Yulian]

Dear Gromacs users,
There is a problem with entropy calculations with g_anaeig.

I have 3 runs 100 ns each.
Entropy for each run (Protein-H) is round 31000 J/mol K (100 ns)
If I use trjcat to connect two trajectories it is  26000 J/mol K (200 ns)
For three connect  trajectories it is round 19000 J/mol K


I tried to do some checking:
For example, I used 50 ns from one traj and 50 ns from the other in
different combinations but it does not change the entropy value
significantly. It seems that three runs are not so different.

Also there is a problem with calculating of entropy on different machines.
The result can be different for different machines (+/-  several hundreds of
J/mol K for this system)

I use gromacs-4.5.6.
For all runs I  use "trjconv  -pbc nojump" before analysis.
Thank you.

 Yulian




--
View this message in context: http://gromacs.5086.x6.nabble.com/Entropy-tp5008547.html
Sent from the GROMACS Users Forum mailing list archive at Nabble.com.