March 14, 2006

The reason why I am getting different .pdb files/inpcrd/prmtop is because of
the residue libraries. They have to have exactly the same ordering...

22:10 PM

I have updated amber9 and compiled the programs. I have created the prmtop
and inpcrd files for ggcc and igigicic. They have exactly the same .pdb
files (except the residue names. and dummy atom and transformed hydrogen
names, etc.). The inpcrd files are exactly the same. I have done a 10 fs
sim, and the sim looks fine.

March 15, 2006

RESULTS 1:

1. Min with CPU=2 works fine, but does not work with CPU=4 or CPU=8.

2. ntc_1_ntf_1_icfe_1_ntb_1 and ntc_1_ntf_1_icfe_2_ntb_1 works fine with
CPU=4.

4. ntc_1_ntf_1_icfe_1_ntb_2 works fine too. In the input file, I set ntp=1 or
2.

5. ntc_2_ntf_2_icfe_1_ntb_1 does not work. When SHAKE is turned on, the
system on both runs have diff. initial temperatures. So, take the .rst file
from ntc_1_ntf_1_icfe_1_ntb_1 and tell the comp. to use this .rst as inpcrd
file, so that SHAKE wont assign arbitrary initial temp.s to the system.

6. In my system I dont have only bonds involving hydrogens, but also dummy
atoms (which are DH, transformations from H to DH, or from DH to H). How can
I include those bonds to be included in SHAKE?

RESULTS 2:

March 16, 2006

1. ntc_1_ntf_1_icfe_1_ntb_1 with time=100 ps simulation has a nice
trajectory.

2. ntc_1_ntf_1_icfe_2_ntb_1 with time=100 ps simulation has a nice
trajectory.

3. ntc_2_ntf_2_icfe_1_ntb_1_with_rst with time=2 ps does not work (I used the .rst file of
ntc_1_ntf_1_icfe_1_ntb_1). There is a definitive problem with SHAKE. I think
that the reason is because SHAKE is contraining the bond involving hydrogens
(H), not the dummy atoms (DH).

4. ntc_2_ntf_2_icfe_1_ntb_1 has the same problem. SHAKE does not work with
TI approach.