AMBER Archive (2004)

Subject: AMBER: GB parameters for Calcium

From: cailliez (
Date: Wed Jun 23 2004 - 07:39:38 CDT

Dear AMBER users,

First I would like to thank David Case for his answer to my first
question on calcium parameterization
few days ago.
I now have two more questions.
I tried to make the same experience as in the paper of Gohlke and Case
(J Comp Chem 25: 238-250, 2004)
plotting results of GB calculations for snapshots taken from MD (almost
200 snapshots) and PB calculations
for the same snapshots.
I used Tsui and Case parameterization for Born radii.
1) Reading the paper I understood that these parameters lead to an
overestimation to electrostatic with GB.
     With my protein (without calcium ions for the moment), this is the
opposite. Did I make any mistake or is it
     possible that, for some systems, GB (with these parameters) leads
to an underestimation of the electrostatic
     part of solvation free energy ?
2) I tried few parameters for calcium : three radii (1.14, 1.5, and
1.71) and three screen factors (0.7, 0.8, 0.9).
     It seems that screen factor does not have a strong influence on the
results unlike radii.
     Anyway for all the cases I made, results are quite similar with
correlation coefficient of 0.98 between GB and PB.
     For all cases, we thus can write E(PB)=A*E(GB)+B, with A=0.96-0.97.

     The big difference is for B. which vary from 300 to 660 kcal/mol,
depending which Born radius I take.
      How can I know which choice is the best (if there is one...) ?

Thanks in advance for your answers,
Sincerely yours,

Fabien Cailliez                         Tel : 01 58 41 51 63
Laboratoire de Biochimie Théorique      e-mail :
IBPC    13, rue Pierre et Marie Curie
75005 Paris

----------------------------------------------------------------------- The AMBER Mail Reflector To post, send mail to To unsubscribe, send "unsubscribe amber" to