Dear Amber Mailing List,
I'm trying to estimate the free energy barrier of SN2 reaction in
solution using thermodynamic integration approach implemented within
sander, where bromine is substituted by iodine in 2-bromoethylbenzene.
Topology files were created in such a manner, that 2-bromoethylbenzene
and the unbound iodine placed in the close vicinity of the rest of the
system were solvated using standard tleap procedure (20 Å box). Then,
the first prmtop file was saved. Next, the bond between carbon and
bromine was deleted and replaced by a bond between carbon and iodine,
what somehow describes the reaction. Charges were manually assigned to
match the latter molecule configuration. Additional frcmod file was
then loaded to incorporate data for missing bonds. Such system was
then saved as the second topology file.
Then, prior to TI calculation, minimization, heating and equilibration
were carried out, using one of the obtained topology files. Next, TI
calculations were run with the following input files:
#TISANDER
&cntrl
imin=0, ntx=5, irest=1,
cut=10.0, ntb=2,
pres0=1.0, ntp=1, taup=2.0,
ntpr=1000, ntwx=1000, ntwr=10000, ntave=5000, nscm=2000,
ntt=3, gamma_ln=5.0,
tempi=313.15, temp0=313.15,
nstlim=500000, dt=0.001,
ifqnt=1,
iwrap=1,
icfe=1,
clambda=0.11270,
/
&qmmm
qmmask=':1',
qmcharge = -1,
spin = 1,
qmshake = 0,
qm_ewald = 1, qm_pme = 1,
qm_theory='PM3',
/
I ran 5 calculations, where different values of clambda were assigned
accordingly to AMBER manual and available tutorials.
So, the problem is, that those calculations end with values of DV/DL
equal to 0, what would mean that my Hamiltonians are exactly the same,
even though both bonding and charges of the systems are different.
Could anyone provide me with any feedback, what may be wrong with my
calculations, where the mistake have been made and what can I do in
order to calculate free energy of the reaction?
Additionally, since in the both of the systems there are unbound atoms
(first iodine, then bromine) I manually change the number of atoms in
the first residue to match all of the atoms I want to treat quantum
mechanically. Without doing so, simulations do not run since the
number of atoms in the whole system do not match the number of atoms
within defined residues. Is this fix correct, or should I just assign
the unbound molecule to another, single-atom residue?
Thank you deeply for your time,
Szymon Żaczek
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Mar 26 2017 - 13:30:02 PDT
