Dear Haddock Community,
I encountered the following issue and was wondering if someone is able to provide some hints.
I was trying to reproduce an X-ray protein-ligand complex by redocking the ligand (in its crystallographic pose) to the protein using the template-based approach
I performed only water refinement dockings using the Haddock2.2 guru interface (following the guidelines here: Advanced refinement of molecular complexes – Bonvin Lab). Everything worked fine and I was able to reproduce the crystallographic binding pose.
However, I repeated some dockings with the Haddock2.4 interface and was observing different results, i.e. I was not able to reproduce the crystallographic binding pose whereas Haddock2.2 did! (following the guidelines here: Advanced refinement of molecular complexes – Bonvin Lab).
In addition to the above mentioned options I also changed some settings according to the recommendations within the ligand docking tutorial. This was done manually (Haddock2.2) or automatically by the new Haddock2.4 interface.
I compared both run.cns files but I did not detect any significant changes. Nevertheless, I repeated the docking with Haddock2.4 and adjusted those parameters. Unfortunately, I got the same result and was not able to reproduce results from Haddock2.2.
Later on, I had a more detailed look on respective docking poses and observed that in poses produced by Haddock2.4 more water molecules were placed within the binding pocket of the molecule. However, I was not able to figure out why.
Does anyone has some suggestions on how to adjust my parameters in order to reproduce the results from Haddock2.2 in Haddock2.4?
Thanks in advance for your feedback,
Nicolas
Dear Nicolas
Did you try to use the refinement interface instead?
https://bianca.science.uu.nl/haddock2.4/refinement/1
What is happening with your ligand in 2.4 then?
Some settings must be off.
Especially if you only perform refinement I would not expect the ligand to really move much.
Hello again,
I tried using the refinement interface and received the following error message:
The following error occurred when processing your refinement run: [Errno 2] No such file or directory: ‘/home/enmr/csb_webserver/data/process/haddock24/9140627022/9c52905f4083f2fa9338d224236a07857b0615dd/9c52905f4083f2fa9338d224236a07857b0615dd/ligand.par’
For the previous tests with Haddock2.2 and 2.4 both run.cns files only the subsequent parameters are different (I did not change them and thus I am guessing this is by default):
“hbond_firstit”: 1 (Haddock2.2); 0 (Haddock2.4)
“water_tokeep”: 0.25 (Haddock 2.2); 0.5 (Haddock2.2)
“dnap_water_tokeep”: 0.25 (Haddock2.2); 0.75 (Haddock2.4)
As mentioned beforehand I rerun the docking by adjusting those parameters but that did not help (which was expected from my side)
The ligand is not able to reproduce the binding pose, i.e. is not able to reproduce every hydrogen bond in the X-ray structure (unlike Haddock2.2). Some hydrogen bonds are now water mediated which was not the case in Haddock2.2. Therefore, I was inspecting the amount of water placed within the binding pocket.
Hi Nicolas
The following error occurred when processing your refinement run: [Errno 2] No such file or directory: ‘/home/enmr/csb_webserver/data/process/haddock24/9140627022/9c52905f4083f2fa9338d224236a07857b0615dd/9c52905f4083f2fa9338d224236a07857b0615dd/ligand.par’
And you managed to submit via the regular interface the same PDB files? Strange.
Can you share via email your input PDB file(s) and also a link to a previous run?
For the previous tests with Haddock2.2 and 2.4 both run.cns files only the subsequent parameters are different (I did not change them and thus I am guessing this is by default):
“hbond_firstit”: 1 (Haddock2.2); 0 (Haddock2.4)
“water_tokeep”: 0.25 (Haddock 2.2); 0.5 (Haddock2.2)
“dnap_water_tokeep”: 0.25 (Haddock2.2); 0.75 (Haddock2.4)
These parameters have nothing to do with a refinement setup…
What are you exactly doing / changing.
Comparing the run.cns file is not so simple since they are quite different.
Thanks for sharing this error, it was caused by a bug in the refinement and it has been fixed now! 