Issue in asymmetric run

[accibio]

Hi. I have a receptor-ligand complex and I'm trying to generate plots to visualize the various types of interactions between the receptor and ligand proteins. The complex has 166 residues in total with residues 1-106 belonging to the receptor and residues 107 to 166 belonging to the ligand. First I ran Conan in the symmetric mode to visualize both the intermolecular and intramolecular interactions. But since I'm interested only in the intermolecular interactions (i.e. between the receptor and the ligand) I used the asymmetric mode by adding the following lines:

START_X 1
NRES_X 60
NTERM_X 107
START_Y 61
NRES_Y 106
NTERM_Y 1

But the two plots show different interactions between the receptor and ligand. This is the output from the symmetric run.

And this is from the asymmetric run. I expected the same interactions from the bottom right part of the symmetric run to show up here. But a lot of new interactions are formed. I'm unable to figure out the issue. Please help.

[dadayqsim]

Hey there,

The NTERM_X and NTERM_Y refer to plotting purposes only (for example, you could set NTERM_X to 1001 and it would give you the same plot, only with the x-axis shifted). The actual selection is done by START_X and START_Y. In this case, it sounds like you are looking for:

START_X 107 <--- the x-axis starts at 107
NRES_X 60 <--- the number of residues (nres) on the x-axis is 60
START_Y 1 <--- the y-axis starts at 1
NRES_Y 106 <--- the number of residues on the y-axis is 106.

This should give the bottom right corner.

I should also mention that these colored contacts are only based on the "r_min" distance criterion, so they are more like first guesses. It could be the case, for example, that two carbon atoms are in contact, but because both residues are polar, conan guesses it must be an H-bond. Similarly, obviously also aliphatic residues can form H-bonds (through their backbones), but conan will think they are hydrophobic interactions. However, if the contacts are not due to the secondary structure, then these guesses are usually okay.

Hope this helps!

Csaba

[accibio]

Thanks for the explanation. I edited those lines but the new plot still doesn't match the bottom right corner.

[dadayqsim]

Hey,

That is strange. I assume this means this "interaction type" plot is broken in asymmetric mode. The pattern seems similar to the bottom right corner but some interactions are missing, others are changed. I think it gets the sequence data wrong. I have not tested this particular plot much because it is an approximation anyway. If you think it is useful to you, I suggest you extract the data from the full plot. Unfortunately I have been locked out of this repository and I could not change it even if I had the time. :(

Csaba