- Posts: 8
how to apply topology exclusions
- irewood
- Topic Author
- Offline
- Fresh Boarder
I'm trying to map a molecule that have a rigid region and a floppy one. I'm looking at cholesterol structure and I saw exclusions but I do not know how to apply them..
please, if any one could help me I will very grateful!
Please Log in or Create an account to join the conversation.
- djurre
- Offline
- Admin
- Posts: 272
1 3 4
3 4
means: atom 1 is excluded from atoms 3&4 but not from atom 2. Atom 3 is excluded from atoms 4 but not from atoms 2. Atom 3 is also excluded from atom one which you don't have to define again on the second line.
Please Log in or Create an account to join the conversation.
- irewood
- Topic Author
- Offline
- Fresh Boarder
- Posts: 8
then, in the particular case of cholesterol exclusions are described for every bead BUT NOT for bead 8 (wich is the floppy tail), that in fact are already defined as bond or angle related interactions between them. Based on this I do not understand the objective of exclusions (if exclusions are automatically generated with grompp when beads are related by means bonds or angles or dihedrals) added manually.
In the case of my molecule the mapping scheme from AA to CG yields
7 -- 6 -- 4 -- 2 -- 1
| |
| |
5 -- 3
[bonds]
; i j
1 2
4 6
6 7
[constraints]
; i j
2 3
3 4
2 4
3 5
4 5
[angles]
; i j k
1 2 3
1 2 4
5 4 6
4 6 7
6 4 3
[dihedrals]
; i j k l
1 3 4 2
2 3 5 4
7 2 4 6
7 5 4 6
6 5 3 4
So, the exclusions could contain
[exclusions]
1 5 6 7 (being are not defined above) or 1 7 (being 5 and 6 1-4 separated from 1)
??
Thanks!
Please Log in or Create an account to join the conversation.
- xavier
- Offline
- Admin
- Posts: 416
The cholesterol was a difficult molecule and the exclusions were explicitly written for all bead to avoid confusion.
You topology is a bit confusing as we do not know to what real molecule it correspond and thus how the different bead should behave one relative to the other. As a general rule you should avoid to use too many constrains especially if they form cycles and link the cycle ... but it should be fine. Dihedrals angle are also delicate as you should make sure three consecutive beads won't align, that would make your molecule extremely numerically unstable.
irewood wrote: Thank you!
then, in the particular case of cholesterol exclusions are described for every bead BUT NOT for bead 8 (wich is the floppy tail), that in fact are already defined as bond or angle related interactions between them. Based on this I do not understand the objective of exclusions (if exclusions are automatically generated with grompp when beads are related by means bonds or angles or dihedrals) added manually.
In the case of my molecule the mapping scheme from AA to CG yields
7 -- 6 -- 4 -- 2 -- 1
| |
| |
5 -- 3
[bonds]
; i j
1 2
4 6
6 7
[constraints]
; i j
2 3
3 4
2 4
3 5
4 5
[angles]
; i j k
1 2 3
1 2 4
5 4 6
4 6 7
6 4 3
[dihedrals]
; i j k l
1 3 4 2
2 3 5 4
7 2 4 6
7 5 4 6
6 5 3 4
So, the exclusions could contain
[exclusions]
1 5 6 7 (being are not defined above) or 1 7 (being 5 and 6 1-4 separated from 1)
??
Thanks!
Please Log in or Create an account to join the conversation.
- irewood
- Topic Author
- Offline
- Fresh Boarder
- Posts: 8
As a first approach we have chosen trp (tryptophan) such its structure bears indole ring, the ring is defined by constraints and only 1-2 bond is defined with 5000 as force constant. In this way we defined similarly another two bonds: 4-6 and 6-7. Not using exclusions the molecule (angles and bonds distances was modified after minimization). Using exclusions as cholesterol topology (7 2 3 4 5; 6 2 3 5) was not modified. However, when we introduce molecule parameterized and vacuum minimized by these two approaches in a lipid bilayer, the system crashes by lincs warning at NPT conditions only when we use exclusions (most of the warnings affecting ring beads). We have previously tested trp alone and sequentially adding beads 6 and 7, it's working at membrane just without or only with bead#6, but these structures are not completes.
We are thinking in modify our topology in other ways: avoiding constraints and using stiff bonds, including exclusions explicitly defined as cholesterol, combining 2 previous strategies.
someone could help me?
Please Log in or Create an account to join the conversation.
- mnmelo
- Offline
- Admin
- Posts: 89
Since you have a rigid frame you will probably get more stability if you define one of those 4 atoms as a virtual site, as a function of the other three.
Check section 4.7 of the manual, particularly the virtual sites of type 3 or 3fd.
You'll probably want to define a triangle of constraints (between atoms 4-3-6, for example) and atom 5 would be a virtual site defined as a function of those three.
Also, the manual recommends setting lincs-order=6 for large timesteps with virtual sites involving constraints, so you may try that to improve stability.
Please Log in or Create an account to join the conversation.