- Posts: 1
martinize.py with elnedyn
- dllynch
-
Topic Author
- Offline
- Fresh Boarder
Less
More
11 years 6 months ago #1187
by dllynch
martinize.py with elnedyn was created by dllynch
Hi All,
I downloaded the martinize.py python tool (version 2.1) and generated a top file for my example (1UBQ)
by:
./martinize.py -f 1UBQ.pdb -o system.top -x cg_1UBQ.pdb -ss 1ubq.ssd -p all -ff elnedyn -ef 500 -eu 0.9
I was looking at the Protein_A.itp file this generates and comparing it to the earlier ElNeDyn fortran codes to produce the analogous itp file.
What i am not sure about is:
1) I was planning on using the v2.1 martini ff (not polarizable), but the AC1 and AC2 atoms types are not present in the new Protein_A.itp ( looks like they are C1 and C2 instead). If I don't want to use a polarizable water molecule (at least not yet, I'm thinking of comparing the results with/without the polarizable water), can i get martinize.py to output these atom types as AC1/AC2? I suppose I could just edit (blech) the .itp file itself.
and
2) It looks like the very stiff backbone - backbone (Ca i -> i+1) "bonds" of the original elnedyn are replaced with constraints. Is there a reason for this? I am guessing it has something to do with timestep/stablity issues with such stiff bonds.
Any help is much appreciated!
Thanks,
Diane
I downloaded the martinize.py python tool (version 2.1) and generated a top file for my example (1UBQ)
by:
./martinize.py -f 1UBQ.pdb -o system.top -x cg_1UBQ.pdb -ss 1ubq.ssd -p all -ff elnedyn -ef 500 -eu 0.9
I was looking at the Protein_A.itp file this generates and comparing it to the earlier ElNeDyn fortran codes to produce the analogous itp file.
What i am not sure about is:
1) I was planning on using the v2.1 martini ff (not polarizable), but the AC1 and AC2 atoms types are not present in the new Protein_A.itp ( looks like they are C1 and C2 instead). If I don't want to use a polarizable water molecule (at least not yet, I'm thinking of comparing the results with/without the polarizable water), can i get martinize.py to output these atom types as AC1/AC2? I suppose I could just edit (blech) the .itp file itself.
and
2) It looks like the very stiff backbone - backbone (Ca i -> i+1) "bonds" of the original elnedyn are replaced with constraints. Is there a reason for this? I am guessing it has something to do with timestep/stablity issues with such stiff bonds.
Any help is much appreciated!
Thanks,
Diane
Please Log in or Create an account to join the conversation.
- xavier
-
- Offline
- Admin
Less
More
- Posts: 416
11 years 6 months ago #1188
by xavier
Replied by xavier on topic martinize.py with elnedyn
Dear Diane,
To answer your concerns:
1) The change from a C1/C2 to AC1/AC2 in ElNeDyn has not yet been operated. It was kept as in the original ElNeDyn, at the time is was made this change did not exist. To be consistent with more recent MARTINI version the C1/C2 particle should be replace by a AC1/AC2 particle when interacting with charges. You can do the change by hand in the topology or in the martinize.py in the definition of ELNeDyn.
Thanks for pointing this out. It will be added in the new release of martini.py.
2) The change from a strong bond to a constrain was a natural thing to do and will not affect your system behavior. Note however that in the case of the histidine residue, the presence of a constrain to model the BB-SC bond gives troubles to LINCS to converge. So the use of bonds is numerically most stable but should be used as a constrain.
To answer your concerns:
1) The change from a C1/C2 to AC1/AC2 in ElNeDyn has not yet been operated. It was kept as in the original ElNeDyn, at the time is was made this change did not exist. To be consistent with more recent MARTINI version the C1/C2 particle should be replace by a AC1/AC2 particle when interacting with charges. You can do the change by hand in the topology or in the martinize.py in the definition of ELNeDyn.
Thanks for pointing this out. It will be added in the new release of martini.py.
2) The change from a strong bond to a constrain was a natural thing to do and will not affect your system behavior. Note however that in the case of the histidine residue, the presence of a constrain to model the BB-SC bond gives troubles to LINCS to converge. So the use of bonds is numerically most stable but should be used as a constrain.
dllynch wrote: Hi All,
I downloaded the martinize.py python tool (version 2.1) and generated a top file for my example (1UBQ)
by:
./martinize.py -f 1UBQ.pdb -o system.top -x cg_1UBQ.pdb -ss 1ubq.ssd -p all -ff elnedyn -ef 500 -eu 0.9
I was looking at the Protein_A.itp file this generates and comparing it to the earlier ElNeDyn fortran codes to produce the analogous itp file.
What i am not sure about is:
1) I was planning on using the v2.1 martini ff (not polarizable), but the AC1 and AC2 atoms types are not present in the new Protein_A.itp ( looks like they are C1 and C2 instead). If I don't want to use a polarizable water molecule (at least not yet, I'm thinking of comparing the results with/without the polarizable water), can i get martinize.py to output these atom types as AC1/AC2? I suppose I could just edit (blech) the .itp file itself.
and
2) It looks like the very stiff backbone - backbone (Ca i -> i+1) "bonds" of the original elnedyn are replaced with constraints. Is there a reason for this? I am guessing it has something to do with timestep/stablity issues with such stiff bonds.
Any help is much appreciated!
Thanks,
Diane
Please Log in or Create an account to join the conversation.
Time to create page: 0.089 seconds