Temperature coupling

User discussions
1

GROMACS version:2020.7
GROMACS modification: Yes/No

Hello All,

Excuse my novice experience in GROMACS MD simulation.
What is the approach in setting up proper control of temperature coupling. I have system with a metalloprotein(Zinc Ions) and DNA scaffold.
My logic in approaching this situation is as follows
I will create an index file, creating the below naming scheme.
in .mpd files, tc-grps = SOLUTE SOLVENT

SOLUTE = will consist of protein, DNA, Zinc ions and also counterions(NA+)
SOLVENT = water artoms

Is this correct ?, or this flawed. Also should I define the bulk counterions as their own thermostat group?
Below are the components of system(Cysteine amino residues have their own RES ID, so i believe standard tc-grps = PROTEIN NON-PROTEIN will not work for my system)

Reading structure file
Going to read 0 old index file(s)
Analysing residue names:
There are: 130 Protein residues
There are: 20 Other residues
There are: 38 DNA residues
There are: 28 Ion residues
There are: 11632 Water residues
Analysing Protein…
Analysing residues not classified as Protein/DNA/RNA/Water and splitting into groups…
Analysing residues not classified as Protein/DNA/RNA/Water and splitting into groups…
Analysing residues not classified as Protein/DNA/RNA/Water and splitting into groups…

0 System : 38348 atoms
1 Protein : 2052 atoms
2 Protein-H : 1014 atoms
3 C-alpha : 130 atoms
4 Backbone : 390 atoms
5 MainChain : 522 atoms
6 MainChain+Cb : 638 atoms
7 MainChain+H : 654 atoms
8 SideChain : 1398 atoms
9 SideChain-H : 492 atoms
10 Prot-Masses : 2052 atoms
11 non-Protein : 36296 atoms
12 Other : 164 atoms
13 CY1 : 10 atoms
14 CY2 : 10 atoms
15 CY3 : 10 atoms
16 CY4 : 10 atoms
17 CY5 : 10 atoms
18 CY6 : 10 atoms
19 CY7 : 10 atoms
20 CY8 : 10 atoms
21 ZN1 : 1 atoms
22 ZN2 : 1 atoms
23 CY9 : 10 atoms
24 1CY : 10 atoms
25 2CY : 10 atoms
26 3CY : 10 atoms
27 4CY : 10 atoms
28 5CY : 10 atoms
29 6CY : 10 atoms
30 7CY : 10 atoms
31 ZN3 : 1 atoms
32 ZN4 : 1 atoms
33 Na+ : 28 atoms
34 DNA : 1208 atoms
35 CY1 : 10 atoms
36 CY2 : 10 atoms
37 CY3 : 10 atoms
38 CY4 : 10 atoms
39 CY5 : 10 atoms
40 CY6 : 10 atoms
41 CY7 : 10 atoms
42 CY8 : 10 atoms
43 ZN1 : 1 atoms
44 ZN2 : 1 atoms
45 CY9 : 10 atoms
46 1CY : 10 atoms
47 2CY : 10 atoms
48 3CY : 10 atoms
49 4CY : 10 atoms
50 5CY : 10 atoms
51 6CY : 10 atoms
52 7CY : 10 atoms
53 ZN3 : 1 atoms
54 ZN4 : 1 atoms
55 Na+ : 28 atoms
56 Ion : 28 atoms
57 CY1 : 10 atoms
58 CY2 : 10 atoms
59 CY3 : 10 atoms
60 CY4 : 10 atoms
61 CY5 : 10 atoms
62 CY6 : 10 atoms
63 CY7 : 10 atoms
64 CY8 : 10 atoms
65 ZN1 : 1 atoms
66 ZN2 : 1 atoms
67 CY9 : 10 atoms
68 1CY : 10 atoms
69 2CY : 10 atoms
70 3CY : 10 atoms
71 4CY : 10 atoms
72 5CY : 10 atoms
73 6CY : 10 atoms
74 7CY : 10 atoms
75 ZN3 : 1 atoms
76 ZN4 : 1 atoms
77 Na+ : 28 atoms
78 Water : 34896 atoms
79 SOL : 34896 atoms
80 non-Water : 3452 atoms
81 Water_and_ions : 34924 atoms

Any suggestions or tips would be greatly appreciated.
Thanks

2

Nowadays there are extremely few systems that require multiple temperature coupling groups. I would suggest to use a single group.

3

Okay, thanks for the reply.

Wow, you are right. I done short experimentation:
exp_1 ::( tc_grps = non_water(protein+DNA+Zn+NA(counterions) : Water atoms
exp_2 ::( tc_grps = Solute(protein+DNA+Zn): Solvent (Water and NA)
exp_3 :: (tc-grps - System(whole system)

Ran a MD simulation on each experimentation, it seems to be case oddly enough the experiment 3 produced the most stable MD analysis(RMSD calculations of the trajectory analysis of the protein).

Is it fine Hess, to use the whole system for my tc-grps for my .mdp file. I am actually suprised this worked out the way it did, I was expecting a different result. because on this page : GROMACS MDP File Parameters: integrator, thermostat, barostat, and cutoffs explained - Compchems
states that " You can create separate groups for the thermostat with the tc-groups parameter. It is generally recommendable to create at least two groups, consisting of the protein (Protein) and the rest of the system (Non-protein)."

4

Very likely your results are just the statistical variation between simulations. I would not expect any measurable difference between those settings.

The advice for separate T-coupling groups is relevant for GROMACS versions of 15 years ago.

1 Like