QM/MM tutorial

QM/MM calculations on thymine dimer repair.

QM/MM setup

After an electron is transferred to the thymine dimer, the total charge of the formerly neutral QM subsystem is decreased to -1. In addition, the excess electron increases the total spin quantum number of the QM subsystem from 1/2 to 1. Thus, instead of singlet we now have a doublet multiplicity. These changes must be added to the QM/MM parameters of the mdp file:

QMMM = yes
QMMM-grps = QMatoms
QMmethod = AM1
QMbasis = STO-3G
QMMMscheme = ONIOM
QMcharge = -1
QMmult = 2

We now perform a short 1 ps QM/MM simulation of the system with the excess electron on the thymine dimer. We use the final frame from the previous QM/MM simulation as the starting structure:

grompp -f qmmm2.mdp -p qmmm.top -n qmmm.ndx -c qmmm1out.gro

./mdrun -v -c qmmm2out.gro -x traj_stepV.xtc

Also this simulation will take approximately 40 minutes. Therefore, you can download the output files: electron1.tar , and uncompress it as the previous tar files.

In the final frame, we see that one of the bonds in cyclobutane ring has broken (Figure 4). To visualize the whole process you can convert the trajectory into pdb format that you can subsequently read in with pymol. Alternatively, you can use VMD.

trjconv -s -f traj_stepV.xtc -n qmmm.ndx -o traj2.pdb

and select "heavyatoms" group when asked.

Thus, upon electron uptake, the bond between the C5 atoms on the dimer (Figure 2) is broken. Considering the timescale at which the break occurs, this process must be barrierless. The reason why the C5-C5 bond breaks and not the bond between C6-C6 is probably the steric repulsion between the methyl groups. The covalent bond between the C6 atoms remains intact on the timescale of the simulation. Cleaving the second bond therefore is an activated process that involves a (small) activation barrier. To overcome this barrier in a short QM/MM MD simulation, we will make use of the chemical flooding approach, that is described in the next session.

back to top

updated 28/10/08