AMBER Archive (2003)

Subject: gibbs run

From: Xiang, Tian-Xiang (txian2_at_email.uky.edu)
Date: Mon Jan 27 2003 - 11:05:06 CST

Dear Sir:

I am trying to calculate free energy for mutating a dipeptide to a monpeptide (e.g. ALA-ALA -> ALA) using gibbs. According to the manual, for decoupling electrostatic constribution, two runs are needed: : electrostatic
LAMBDA = 1 -> 0 with LAMBDA(vdw) = 1, followed by
van der Waals LAMBDA = 1 -> 0.

I am not sure how these two runs can be set up. My understanding is that:
in the first run, set ielper=1, almda=1.0, all others are set the same as a regular run without decoupling;
in the second run, set ielper =-1,almda=1.0, all others are set the same as a regular run witht decoupling.

After the first run is finished, copy the restart file to be the inintial coordinate file and start the second run.

 Are the above procedures correct?

T.X. Xiang