You are here

Ineffective Constraints

2 posts / 0 new
Last post
Ineffective Constraints

Hi all,

I am attempting to use Rosetta 3.8 (2017.29.59598) AbinitioRelax with some constraints I have defined, but I get the feeling that the constraints are not being properly implemented. In particular, I attempt to implement a single harmonic constraint using a constraint file with this line:

AtomPair CA 1 CA 41 HARMONIC 55.0 .1

My flags are:

-in:file:fasta /home/safa/RosettaWork/Hunt/Cterm.fasta                  
-in:file:frag3 /home/safa/RosettaWork/Hunt/aat000_03_05.200_v1_3
-in:file:frag9 /home/safa/RosettaWork/Hunt/aat000_09_05.200_v1_3
-database /home/safa/Downloads/rosetta_src_2017.29.59598_bundle/main/database                                                
-nstruct 16                                       
-out:pdb true                                                    
-out:path /home/safa/RosettaWork/Hunt
-abinitio:increase_cycles 10
-abinitio:rg_reweight 0
-abinitio:rsd_wt_helix 0.5
-abinitio:rsd_wt_loop 0.5
-constraints:cst_weight 10.0
-constraints:cst_fa_weight 10.0
-constraints:cst_file /home/safa/RosettaWork/Hunt/constraints
-constraints:cst_fa_file /home/safa/RosettaWork/Hunt/constraints

During the run,  I see the score output always says "atom_pair_constraint        10.000       0.000       0.000", and the final structures do not seem to be using the constraint I have given. However the final structure file does contain the energy breakdown table, and this table typically has an extremely high (>10,000) energy for the constraint. I do not understand how this constraint only seems to gain energy when the final structure is scored.

I know I am probably just missing some key parameter but I cannot figure out what. This is very confusing because I used Rosetta 2.5 and 2.4 about 2 years ago and as best I can remember, everything I am showing now would work with them. Does anyone have any advice on how I can fix this problem?

Thanks a lot,


Post Situation: 
Tue, 2017-09-05 19:36

I don't think this will expose a problem, but let me suggest that you make your constraint deliberately corrupted - call it an AbomPair constraint or something, with a deliberate misspelling - and let's see when and if it rejects that?  


Fri, 2017-09-08 11:05