Contact map for interaction of 2 proteins

User discussions
1

GROMACS version: 2020
GROMACS modification: No
Here post your question

Dear all,

I am looking at 2 protein interactions. I am interested to see which specific residue in protein A interacts with residues in protein B (a broad range targeting as I do not know which residue protein A is interacting) and the evolution during the simulation time. I saw that using gmx mdmat can give you the contact map, but it is only for single protein residue, and I also saw that using gmx select, but it is only giving me 0 - 1 (not naming specific residue). Therefore, I would like to ask is there any plugin software or can someone give me some idea on how to achieve my idea?

Best regards,

Ben

2

Hi,
you can try gmx mindist. It does not provide a contact map, but computes the distance between one group and a number of other groups. Among other output, the number of contacts (and the list of contact) within a given distance are written to two separate output files.
see gmx mindist gmx mindist — GROMACS 2021.1 documentation
best regards
Alessandra

3

HI Alessandra,

Thanks for your answer. However, I am less interested the number of contacts, but more interested which residue in A is binding with residue in B. gmx mindist cannot give me that.

Best,

Ben

4

Hi,
if you are interest in hbond bridging, maybe gmx hbond helps.
Alessandra

5

Hi,

Thanks again for your answer. However, I am only interested to find out which residue is interacting and not the average HBond interaction.

Ben

6

Maybe you can use g_contacts to get this information.

7

I do not understand this claim. mdmat analyzes the entire group you present it with and produces a matrix of residues in each protein (with quadrants for intra-protein as well as inter-protein residue distances). It does not show you anything over time, but can indeed show you the overall pattern of interactions, which can get you started rather easily.

8

Hi,

Thank you very much on the script. I will try and use that.

Best,

Ben

9

Hi Justin,

I see, you mean I should extract it as a whole i.e. protein, instead of separate it as “chain A” and “chain B”. Thank you for that suggestion.

Edit:

I have used the gmx mdmat and choose the option 1: protein. Upon seeing the graph, it doesn’t give the residue name on the X and Y axis. That is what I want to get out of the graph.

Best,

Ben

10

Hi Ben,

Check out our contact map software called “conan”, it runs gmx mdmat for you and interprets its output. It is a little wasteful, as it works on an (N+M)^2 matrix rather than NxM, but it gets the job done most of the time. It plots contact maps and computes various quantities but also gives you plain-text output that you can post-process if you like.

Or the paper (including the SI which is way longer) to see some of its functionality:

Let me know if you have any more questions.

Csaba

edit: we also compute time-dependent ie animated contact maps by using “gmx mdmat … -frames”