HADDOCK2.4 Local installation error

HADDOCK
1

I have installed Haddock2.4, CNS1.3, with a local Python2.7 and few other additional packages for Haddock.
Python2 is included in the Path too.
Since its a local desktop, there are no queues.
Installation and sourcing works fine.
When I start running a test job with my own structures, the inputs are created and the job runs till it1 refinement.
In between, I get the following error and the job waits for queue.

calculating structure 6
  queue command:
 ------------------------------------------------------------

/bin/csh haddock_trials_runfr320_4w.job
Structure 5 10: running
Structure 6 4: running

Queue command failed, retrying in 1920 seconds
QUEUE_DOWN
Traceback (most recent call last):
File “./tools/make_contacts.py”, line 86, in
_calculate_contacts(executable, struct, cutoff)
File “./tools/make_contacts.py”, line 32, in _calculate_contacts
p = Popen([executable, pdbfile, d_cutoff], stdout=PIPE)
File “/usr/lib/python2.7/subprocess.py”, line 394, in init
errread, errwrite)
File “/usr/lib/python2.7/subprocess.py”, line 1047, in _execute_child
raise child_exception
OSError: [Errno 8] Exec format error
foreach: No match.

I haven’t changed anything in run.cns apart from reducing the number of structures to evaluate.
I thought the error is due to Python and so I tried with both python2.7.17 and python2.7.18. I even tried on a different desktop.
The error still persists.

Any help would be very useful…

Sangeetha

2

Try to do a make clean followed by make all in the haddock installation directory.

It looks like the archive contains already an executable for some tools that needs to be compiled for your system

3

Thank you for solving the issues.
Now the installation works fine.

I have another question regarding the definition of passive resides.
I am working with an IDR and I know the active residues based on a chemical shift experiment.
However, since the protein is only ~40aa long, all residues have solvent accessibility >40%.

I have gone through the literature and I am well aware that the definition of passive residues only helps the docking process and does not directly affect the scoring.
Also, it is not usually advisable to define the whole protein as passive.

Any guidance on how to define my passive residues?
Being a small protein, is it ok to include all residues as passive or can I entirely ignore defining any passive residues?

4

It all depends on how confident you are with you active residue definition from NMR.

The more passive you define the more computational time it will cost.

You could consider including only a few neighbouring residues.

Also remember that HADDOCK will not fold your peptide if this is what is happening…

5

I have an ensemble of conformations that I use for docking so I am not expecting the peptide to fold during docking.

I still have an issue with defining the active/passive residues.

I am certain of few residues expressing chemical shifts after complex formation and so I define them as active residues for one of the proteins. I do not know any actives on the interacting partner.
Also, I would like to see which structure from the ensemble would form a complex by interacting through the known actives.

In simple words, I would like to do a blind docking to see which complex satisfies the experimental data.

Is it possible to perform blind docking with Haddock?

Can I ignore defining any active residues and define all of them as passive and that could be a blind docking?

6

On the server you can easily define all solvent accessible residues as passive for the second molecule (there is a checkbox for that).

And use your few active residues for the first molecule.

7

Thank you for all your suggestions.