Ab-initio warning

HADDOCK
1

Hello, I am getting this error while running ab-initio docking on linux.

TOTAL NUMBER OF DISTANCE RESTRAINTS FOR RIGID BODY DOCKING IS ZERO!
CONTROL YOUR PARAMETER SETTINGS AND RESTRAINT DEFINITIONS
STRUCTURE NUMBER 1

I have made following changes in the run.cns file.

1: Number of structures to dock = 10000,400,400
2: analysis and clustering = RMSD, 7.5
3: Do you want to randomly exclude a fraction of the ambiguous restraints (AIRs)? noecv=true
4: epsilon_0 (it0) =10.0 and epsilon_0 (it1) =10.0

Please tell me what I am doing wrong.
Thank you

2

You will have to enable either the center of mass restraints (cmrest=true) or the random AIRs (ranair=true) in run.cns for ab-initio docking

3

Thanks for reporting this warning @gaurav.sharmapsit.

There is information about ab-initio docking in our Best Practices Guide, please have a look at the ab initio section: How to use information about interactions in HADDOCK? – Bonvin Lab

In this guide you will also find an explanation about the different types of ab initio docking supported by HADDOCK - Random interaction restraints, Surface contact restraints and Center of mass restraints.

The best one to use will depend on your scenario, please have a look at the guide and the tutorials that it references. Good luck!

4

Hello,

Thanks for the reply. One more question. where is the run.cns file is saved. If I want to make modifications.

Best

5

When using the server, you can download all pre-processed file before submission.

And otherwise you can find it in the full archive of the run provided in the results pages.

But you can’t upload a modified run.cns file to the server

6

Okay. Thanks for the explanation.

I am running ab initio docking in with 10000, 400 and 400 docking steps. However, in first 10000 steps I am getting the following error:
Structure 8056: running
Structure 8057: running
------------------------------------------------------------
HADDOCK has detected an error
Check the FAILED file in /home/abwer/Downloads/haddock2.4-2022-01/examples/CD44_d24_blind/run1
Stopping…

  • Cleaning up the run directory …*
    Cleaning up HADDOCK run directory
    Only files for structure #1 will be kept…
    QUEUE_DOWN
    gzip: CD44_d24_blind_run1_it0_refine_8048.out: No such file or directory

Queue command failed, retrying in 15 seconds
QUEUE_DOWN
gzip: CD44_d24_blind_run1_it0_refine_8047.out: No such file or directory

Queue command failed, retrying in 15 seconds
##############################################################################
Finishing HADDOCK on: 2023-06-20 14:44:17
Au revoir.
Tot ziens.
Bye bye.
I can fix it by deleting the FAILED file and running haddock again. However, what is the actual reason and how to prevent this error from happening again and again.

Thanks you so much for your guidance
Gaurav

7

Check the content of the FAILED file if any.

You are running a local version, which one?

This file is generated usually if no restraints are defined. What mode are you using for ab-initio docking?

8

The FAILED file was empty. I am using Haddock 2.4 version January release. For distance restraint I have only changed the {===>} cmrest=true;

Following are the distance restraint in run.cns
{===>} noecv=true;
{===>} ncvpart=2;
{===>} hbonds_on=false;
{===>} ranair=false;
{===>} cmrest=true;
{===>} cmtight=true;
{===>} kcont=1.0;
{===>} surfrest=false;
{===>} ksurf=1.0;

Finally, I am not sure about the mode. But these are the two “mode” I found in run.cns file.

{* number of semi-flexible segments for molecule 1 (-1 for automated mode) }
{ Note that current max is 10 (edit the run.cns to add more segments *}
{===>} nseg_1=-1;

{* Rebuild missing atoms in the context of the complex? (refinement mode) *}
{+ choice: true false +}
{===>} rebuildcplx=false;

Thank you for your help

9

Ok - nothing special

In any case you can try:

  1. to delete the FAILED file

  2. run haddock-clean from the tools directory to clean the output files

  3. restart haddock in the run directory. It will try to complete the missing models (i.e. do not delete the PDB files or start a new run)

And eventually you can change the CNS code in the protocols directory not to write those FAILED files in case of problems…
You might get bad models (or undocked ones) in that case. There are two instance of writing the FAILED file in the refine.inp CNS script.
Search for FAILED and in particular the

evaluate ($errfile = "RUN:FAILED”)

statements. And change RUN:FAILED to e.g. RUN:WARNING