Repository navigation
Wondering if it's possible how to use the GALigand workflow with RNA in the mix? #385
Replies: 2 comments
|
Hello, Rosetta is designed to handle RNA, so this issue might be caused by a bug. Could you please share your input file with us? This will help us reproduce the error and investigate potential bugs. |
|
Hi Guangfeng @gfzhou. Sorry to raise this issue again. I'm still trying to set up GALigandDock with RNA in the mix. I'm wondering if you could give me a conceptual pointer. The fundamental problem for me, again, is that the RNA backbone is being sampled for whatever reason by the protocol and i'm not sure how to disable this. I've had it recommended to me multiple times to "just freeze" the RNA backbone so it's not sampled, but it's not clear to me how do this in the context of GALigand since i'm not finding any I tried to set up a MoveMapFactory but again, am pretty sure it's not doing anything by itself. To find out as to why this is happening in the first place, I tried taking a look at the protocol source itself ( First of all I couldn't figure out why this flag doesn't apply to the RNA: The torsion sampling mechanism seems to be invoked in two key places: In GALigandDock.cc, during initial structure generation: And then it goes to LigandConformer.cc: That's at least as far as is intelligible to me (i'm not really familiar with Rosetta's codebase), but begs the question as to why the RNA is included in the LigandConformer in the first place. My thinking was that i have underspecified my binder/receptors so the protocol is trying to do minimize the whole system.. So i tried specifying the ligand with the But now for whatever reason i'm getting the errors associated with that ligand not being a "polymer", ostensibly from the time when the protocol only services peptides (and not small molecules). This was even more baffling to me because i have already successfully used GALigandDock in a situation where the ligand was a small molecule and the receptor was purely protein. To sum up:
Here is a repo with source data, the scripts i refer to reproduce the error if need be.. Thank you kindly for any input you may have. |
Uh oh!
There was an error while loading. Please reload this page.
TLDR:
How can I disable Ramachandran torsion sampling for RNA in GALigandDock?
Is GALigandDock in general fit to handle protein+rna ligand docking tasks?
I want to dock various ligands to the eIF4a protein which frequently feautures a short strand of rna at the binding site (ex. structure is 6XKI). We came across this wonderful paper by DiMaio's lab https://www.nature.com/articles/s41467-024-52061-7 and i want to try their method.
The problem is basically docking a ligand to a known binding site which is composed of BOTH RNA and protein. The protocol worked wonderfully in the context where it was just ligand<-->protein docking, but the moment i switched to ligand<-->protein+rna binding problem i keep getting the following error:
As far as I understand this occurs because the protocol is trying to apply backbone torsion preferences to RNA, for which no appropriate table exists (in general).
I tried "disabling" Rama module by weighing it down to 0, but it seems to have no effect (they are still being sampled, i suppose):
My question is whether it's still too early to use GenFF + GALigandDock in this hybrid RNA+Protein <--> Ligand docking context. I realize that historically there has been less support and facilities for rna in rosetta, but i heard opinions to the effect of "Rosetta should be able to handle it" basically. If so, is it possible to reconcile the GALigandDocking approach with the presence of RNA or should I look at building my own RosettaScript to circumvent this?
I thank you kindly in advance for anything you could share towards this issue. I'll attach my options and scripts files.
galigand.xml:
dock.options:
All reactions