Giulio,
I don't think there is anything strange in your case. Cancellation error
tends to make Delta G correlates poorer than G in general.
All the best,
Ray
==========================================
Ray Luo, Ph.D.
Dept Molecular Biology & Biochemistry
University of California, Irvine, CA 92697
USPS: PO Box 3900
Phone: (949) 824-9528
Email: rluo.uci.edu
Web: http://rayl0.bio.uci.edu/
==========================================
-----Original Message-----
From: owner-amber.scripps.edu [mailto:owner-amber.scripps.edu] On Behalf Of
Giulio Rastelli
Sent: Friday, May 11, 2007 7:48 AM
To: amber.scripps.edu
Subject: AMBER: correlation between PBTOT and GBTOT values in a MD
trajectory
Dear all,
we are analysing 2800 snapshots taken from a 14 ns MD simulation of a
protein-ligand complex in water using Amber9, and comparing MM-PBSA
(Amber PBSA, no Delphi) and MM-GBSA (igb=1) performance in binding free
energy prediction by analysing the free energy of each snapshot (i.e.
not printing only the average value)
We found that the PBTOT and GBTOT values of the 2800 snapshots correlate
well for complex,
receptor and ligand, while the complex - receptor - ligand (delta)
values of PBTOT and GBTOT do not correlate at all.
Complex: PBTOT vs GBTOT R2=0.88
PBCAL vs GB R2=0.96
Receptor: PBTOT vs GBTOT R2=0.90
PBCAL vs GB R2=0.97
Ligand PBTOT vs GBTOT R2=0.98
PBCAL vs GB R2=0.90
Delta: PBTOT vs GBTOT R2=0.036
PBCAL vs GB R2=0.41
I wonder if the lack of correlation of the binding free energy
differences (delta) is due to the uncertainties in the calculation of
the single contributions, which are sometimes high, or to the much lower
absolute values of the free energy contributions in delta with respect
to complex, receptor and ligand?
Average values:
Complex: GAS= -1817+/-128
PBSUR= 128+/-1.8
PBCAL= -4197+/-106
GB= -4334+/-107
Receptor: GAS= -1790+/-129
PBSUR= 130+/-1.8
PBCAL= -4224+/-105
GB= -4346+/-107
Ligand: GAS= 53+/-5
PBSUR= 4.2+/-0.1
PBCAL= -37.9+/-2
GB= -34.3+/-1.9
Delta: GAS= -80+/-7.6
PBSUR= -6.3+/-0.2
PBCAL= 64.9+/-7.5
GB= 46.5+/-6
Comments are really aprreciated.
Regards,
Giulio
--
Prof. Giulio Rastelli
Dipartimento di Scienze Farmaceutiche
Facolta' di Bioscienze e Biotecnologie
Universita' di Modena e Reggio Emilia
Via Campi 183
41100 Modena - ITALY
-------------------------------------
tel 0039-059-2055145
fax 0039-059-2055131
www.scifar.unimore.it/on-line/Home/Gruppidiricerca/Prof.GiulioRastelli.html
-------------------------------------
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun May 13 2007 - 06:07:36 PDT
