Question on scoring of docked protein-protein complex (omit docking process)

HADDOCK
1

Dear users,
Could I ask a question on scoring of docked protein-protein complex (omit docking process)? Thanks!

I tried two different cfg files for scoring:
(1) molecules = [“score_input1.pdb”, “score_input2.pdb”]
“score_input1.pdb” contains chain A, “score_input2.pdb” contains chain B,
The chain A and chain B are actually split into two pdb files from docked complex.
[topoaa]
[emscoring]
Error: ValueError: Found tokens after a closed string. Invalid TOML.

(2) molecules = [“score_input.pdb”]
“score_input.pdb” contains docked protein-protein complex: chain A and chain B
[topoaa]
[emscoring]
per_interface_scoring = true
It runs successfully. and emscoring_A_B.tsv could be found.

My question is: Why setting (1) is wrong, and Whether setting (2) shows a correct binding score?

Thanks!

2

(2) molecules = [“score_input.pdb”]
“score_input.pdb” contains docked protein-protein complex: chain A and chain B
[topoaa]
[emscoring]
per_interface_scoring = true
It runs successfully. and emscoring_A_B.tsv could be found.

My question is: Why setting (1) is wrong, and Whether setting (2) shows a correct binding score?

Because the scoring modules expect a complex as input and not separated molecules.

In your first scenario, if you replace emscoring by emref it should work

And if you want to score a single structure (a complex), you can also use the command line:

haddock3-score

3

@ amjjbonvin Thank you very much for your quick reply. I really appreciate it.
I think I figured out the setting (1) problem: because of my typo: molecules = [“score_input1.pdb” “score_input2.pdb”], I missed the comma in the list.

However, I am still confused about the scoring of a docked complex.
When I score a chainA-chainB complex using a single pdb file, the output score from emscoring.tsv is a global energy score of the complex instead of binding energy score between chainA and chainB. Is my opinion correct?
If my opinion is correct, which scoring method you would recommend to get the binding energy score between chainA and chainB? Could I use my setting(2) with “per_interface_scoring = true” ? or anything else?
Thanks!

4

However, I am still confused about the scoring of a docked complex.
When I score a chainA-chainB complex using a single pdb file, the output score from emscoring.tsv is a global energy score of the complex instead of binding energy score between chainA and chainB. Is my opinion correct?

The various molecules in your complex MUST HAVE different chainIDs.