Hello Rosetta devs,
I'm not sure that this is the right place to post a bug. If not, please let me know where to post.
I compiled both Rosetta 3.10 (and then 2019.07 release), and faced the following error when trying to predict cyclic peptides with terminal disulfide bonds using the simple_cycpep_predict protocol.
To be more specific, my peptide has a non-terminal CYS residue at the C-term, and I was using the
-in:file:native
flag. When computing without a native structure, no error occured.
The backtrace pointed to the set_up_terminal_disulfide_cyclization_mover function (see attached file).
Looking at the source code, I found that the last residue index was sent to the set_up_terminal_disulfide_cyclization_mover function instead of the last CYS residue index. As a dirty fix, I added the following line
last_disulf_res = find_last_disulf_res( native_pose );
at line 2481 and changed the next line from
set_up_cyclization_mover( termini, native_pose, true, last_res );
to
set_up_cyclization_mover( termini, native_pose, true, last_disulf_res );
in the main/source/src/protocols/cyclic_peptide_predict/SimpleCycpepPredictApplication.cc file.
This solved my issue but I haven't further tested this fix (especially, I haven't tested the other cyclization types).
HTH,
Florent
Attachment | Size |
---|---|
![]() | 3.93 KB |
Hi, Florent. If you can send me your inputs at vmulligan@flatironinstitute.org, I'll fix that bug permanently. Thanks!
Hello, Vikram
Have you made any progress on that issue?
Regards,
Florent
Hi,
I've made some additional tests, and actually, the problem only happens when (1) you ask for a terminal_disulfide cyclization, (2) a pdb file is provided as input and (3) the sequence and input file does NOT end with a cysteine.
When the last residue is a cysteine, everything goes fine...
Hope this help,
Florent