You are here

Regarding snugdock Ag-Ab score

8 posts / 0 new
Last post
Regarding snugdock Ag-Ab score
#1

Hi There,

I am running snugdock locally for an Ag-Ab complex.  Program is running fine but I am getting all score value 0.00.

SCORE: total_score         rms CAPRI_rank       Fnat       I_sc       Irms   Irms_leg atom_pair_constraint                cbeta              cen_rms chainbreak complex_normalized           dG_cross dG_cross/dSASAx100 dG_separated dG_separated/dSASAx100 dSASA_hphobic dSASA

SCORE:       0.000       4.622      1.000      0.250     -9.228      3.186      2.599                0.000              169.078                2.973      1.684              0.662              0.000              0.000       -2.243                 -0.203       557.632  1105SCORE:       0.000       5.659      1.000      0.370    -14.032      2.651      2.143                0.000              168.115                5.569      1.478              0.687              0.000              0.000       35.761                  1.931       832.898  1851

SCORE:       0.000       4.820      1.000      0.272    -19.858      3.393      3.089                0.000              168.160                4.229      1.468              0.775              0.000              0.000       -8.163                 -0.533       731.121  1532SCORE:       0.000       6.191      1.000      0.217    -10.966      3.229      2.928                0.000              167.774                5.073      1.677              0.576              0.000              0.000       -9.448                 -1.014       475.903   931SCORE:       0.000       4.848      2.000      0.304    -14.027      3.482      2.459                0.000              168.533                3.780      1.627              0.601              0.000              0.000       -0.605                 -0.044       659.559  1362

SCORE:       0.000       9.305      0.000      0.033    -16.859      6.064      6.220                0.000              167.724                7.687      1.420              0.884              0.000              0.000        9.396                  0.598       859.740  1572SCORE:       0.000       7.578      1.000      0.315    -11.290      3.413      2.792                0.000              168.437                5.140      1.603              0.650              0.000              0.000       11.870                  0.718       751.594  1652

Program is generating 100  structures but I am not sure how to interpret this data. Shall I use I_sc only?

 

I also use Rosie server but it is giving an error message. Here is the link for job ID:  http://rosie.rosettacommons.org/snug_dock/viewjob/56601

Thank you

SS

(screenshot is attached)

Category: 
Post Situation: 
Mon, 2018-12-10 14:47
ss

We are working on getting the full log for the ROSIE failure.  I can read that structure just fine in Rosetta. 

Can you post your full cmd-line input to snugdock?

Mon, 2018-12-10 15:21
jadolfbr

Locally it is running fine. I use two different trials

1.

mpiexec -np 8 $ROSETTA3/bin/snugdock.mpi.linuxgccrelease -in:file:s HLChain_B.pdb -ex1 -ex2 -partners LH_B -docking:dock_rtmin -nstruct 100 > log

 

2. and also with flag

### JOB IO
-s HLChain_B.pdb
#-native HLChain_B.pdb # if available
-out:prefix DOCKING_
-out:file:scorefile score_out_snugdock.sc
#-out:path:pdb models_path_out/
-out:pdb_gz # gzip models for space
-nstruct 100 # adequate sampling
-ignore_zero_occupancy false
-ignore_unrecognized_res
-multiple_processes_writing_to_one_directory


### Docking
-partners HL_B
-spin
-dock_pert 3 8

### do not filter for h3, constraint takes care of this geom
-h3_filter false

### loop modeling settings
#
# KIC

-kic_rama2b
-loops:ramp_fa_rep
-loops:ramp_rama

-kic_omega_sampling
-allow_omega_move true

-loops:refine_outer_cycles 3
-loops:max_inner_cycles 80

# CCD (exit with ptr error)
#-antibody:centroid_refine refine_ccd
#-antibody:refine refine_ccd
#-loops:outer_cycles 5

#more standard settings, for packages used by antibody_H3
-ex1
-ex2aro
-score:weights talaris2014
-corrections::restore_talaris_behavior true

# Disulphide
-detect_disulf false

 

 

But in both run snugdock score is 0.000 and rest of the print is fine.

Thanks for your prompt reply.

S

Tue, 2018-12-11 08:47
ss

Ok, so - its definitely a bug.  The code was recently refactored and a stage in it AFTER docking is inadvertently clearing the pose energies cache.  Simple fix, but it requires some coding to do - as basically the pose needs to be re-scored.  Thanks for bringing this to our attention.  I'll be able to get a pull request up tomorrow and it should be released next week.  Once I have the code working, I can walk you through changing it locally if you don't want to wait for the next weekly release. 

 

Cheers,

-Jared

Thu, 2018-12-13 14:25
jadolfbr

Happy New Year

 

Any update?

-S

Thu, 2019-01-03 09:08
ss

Jared,

Any idea so far?

Thanks

S

 

Wed, 2018-12-12 14:06
ss

Nothing out of the ordinary from your flags.  I'll be able to dive into the code later this week. BTW, did you try the pre-pack locally?  It seems that is what was failing for snugdock on ROSIE.  Should have something on the scoring by Friday. 

Wed, 2018-12-12 15:43
jadolfbr

 

Predock is working fine. I think ROSIE is failing because of occupancy cols are zero.  If you use this flag  -ignore_zero_occupancy then it will work.

$ROSETTA3/bin/docking_protocol.linuxgccrelease -in:file:s HLChain_B.pdb -ex1 -ex2 -partners LH_B -docking:dock_rtmin -ignore_zero_occupancy false

 

Working PDB file is also attached.

Thanks a lot for your time.

File attachments: 
Thu, 2018-12-13 11:30
ss