minirosetta: how to reserve (inter-chain) disulfides after relax ?

(I would appreciate it very much if any of my questions can be answered.)

Dear friends,
I have been trying to reserve the disulfides of my target protein during the minirosetta comparative modelling. However, though I have gone through many posts, I still could not work it out after using several disulfide related flags.

My situation is:

in the target protein:
Light chain (LC): residue 1-214
Heavy chain (HC): residue 215-442

in the template Fab protein 4KMT:
residues 23 88 (light chain disulfide)
residues 134 194 (light chain disulfide)
residues 236 310 (heavy chain disulfide)
residues 356 412 (heavy chain disulfide)
residues 214 432 (inter-chain disulfide)

In the target Fab protein C226S:
residues 23 88 (light chain disulfide)
residues 134 194 (light chain disulfide)
residues 236 310 (heavy chain disulfide)
residues 358 414 (heavy chain disulfide)
residues 214 434 (inter-chain disulfide)

During the relax of minirosetta (option attached), the disulfide (especially the inter-chain disulfide) in target protein randomly exist. So I want to use some flags to keep them fixed.

PS: the last two pairs of disulfide are not of identical pose numbering between template and target.
PS: the last disulfide is inter-chain disulfide
PS: after minirosetta, the chain ID of "L" and "H" both change to "A"

The following is the options for my minirosetta and my unsuccessful experience for the disulfide related flags.

-----------------------------------------------------------------------------------------------------------------
-run:protocol threading
-in:file:alignment /home/lanselibai/Cheng/20150121_minirosetta/input/C226S.4KMT.aln
-cm:aln_format general
-frag3 /home/lanselibai/Cheng/20150121_minirosetta/input/aaC226S_03_05.200_v1_3
-frag9 /home/lanselibai/Cheng/20150121_minirosetta/input/aaC226S_09_05.200_v1_3
-in:file:fasta /home/lanselibai/Cheng/20150121_minirosetta/input/C226S.fasta
-in:file:fullatom
-loops:frag_sizes 9 3 1
-loops:frag_files /home/lanselibai/Cheng/20150121_minirosetta/input/aaC226S_09_05.200_v1_3 /home/lanselibai/Cheng/20150121_minirosetta/input/aaC226S_03_05.200_v1_3 none
-in:file:psipred_ss2 /home/lanselibai/Cheng/20150121_minirosetta/input/C226S.psipass2
-in:file:fullatom
-out:nstruct 10000
-in:file:template_pdb /home/lanselibai/Cheng/20150121_minirosetta/input/4KMT_FullAtom_missing_residue_in_HC.pdb
-loops:extended
-loops:build_initial
-loops:remodel quick_ccd
-loops:refine refine_ccd
-silent_decoytime
-random_grow_loops_by 4
-select_best_loop_from 1
-out:file:fullatom
-out:output
-out:pdb
-out:file:scorefile threaded_model.fasc
-overwrite
-loops:relax fastrelax
-relax:default_repeats 8

#-in:fix_disulf /home/lanselibai/Cheng/20150121_minirosetta/input/disulfide_file
#-norepack_disulf
#-disulfide_file
#-detect_disulfide_before_relax

#-disulf_filter true
#-rebuild_disulf false
#-detect_disulfides

#-constraints:cst_fa_file /home/lanselibai/Cheng/20150121_minirosetta/input/inter_chain.cst
#-constraints:cst_file /home/lanselibai/Cheng/20150121_minirosetta/input/inter_chain.cst
------------------------------------------------------------------------------------------------------------------
Flag: -in:fix_disulf /home/lanselibai/Cheng/20150121_minirosetta/input/disulfide_file

This seems to be the most related one (e.g. https://www.rosettacommons.org/node/3207).
However, by referring to "main\source\src\core\io\raw_data\DisulfideFile.hh", I tried to used the below in the "disulfidefile":

23 88
(a new blank line)

But I got:

"Segmentation fault (core dumped)" and no record was in the log file.

The link below contains all my files.

https://copy.com/GOaBvrz3tBb4JSz2

When the "-in:fix_disulf" is disabled, it works all fine.

Can I ask why? I was trying if it can work for just one disulfide.

------------------------------------------------------------------------------------------------------------------
Flags not available:

-norepack_disulf
-disulfide_file

I got the error message:

ERROR: ERROR: Option matching -norepack_disulf not found in command line top-level context
ERROR: ERROR: Option matching –disulfide not found in command line top-level context
------------------------------------------------------------------------------------------------------------------
Flag: -detect_disulfide_before_relax

With this flag, the disulfide bond still randomly exist after the minirosetta. What is this flag supposed to do?
------------------------------------------------------------------------------------------------------------------
Flags that I still do not know:

-disulf_filter true
-rebuild_disulf false
-detect_disulfides

These flags appeared in some posts. But what do they really do?

------------------------------------------------------------------------------------------------------------------
How to filter outputs based on disulfide?

According to Dr R Moretti said on https://www.rosettacommons.org/node/3207 , I think he refers to "dslf_fa13" in the score_jd2 method. However, as I rank the score based on "dslf_fa13" for 76 outputs ("score.sc" attached), some of the high dslf_fa13 score ones (e.g. 3.634) are also of all the disulfide while some of the low dslf_fa13 score ones (e.g. -2.877) are not of all the disulfide. Therefore, we cannot simply filter out the outputs with all the 5 disulfide based on "dslf_fa13". As a result, the rule of "the lower of the dslf_fa13, the more disulfide exist" does not exist.

------------------------------------------------------------------------------------------------------------------
What is the proper constraint to use?

I think SiteConstraint should be the most related one as the residue 214 in LC should interact with HC by the inter-chain disulfide. So I write the below in the "inter_chain.cst" file and apply "-constraints:cst_fa_file" and "-constraints:cst_file" in the option file (https://www.rosettacommons.org/docs/latest/constraint-file.html):

SiteConstraint CA 214L H SIGMOID 5.0 2.0

However, I got "Segmentation fault (core dumped)" and the log file is attached.

I also do not understand the meaning of "AmbiguousConstraint" as smlewis suggested on https://www.rosettacommons.org/node/3207.

------------------------------------------------------------------------------------------------------------------
I would be very grateful if someone can help me for any of the problems I have encountered. Thank you very much.

Yours sincerely
Cheng