You are here

Home » Forums » Rosetta 3 » Rosetta 3 - Build/Install

Backrub patch

3 posts / 0 new
Log in or register to post comments
Last post
Backrub patch
#1
Top

Hi,

When I add the rosetta3.1_backrub_patch35535 rosetta can't compile.
It complains about the new SidechainMover.cc:

src/protocols/moves/SidechainMover.cc:158: error: 'const class core::scoring::dunbrack::SingleResidueDunbrackLibrary' has no member named 'get_phi_from_rsd'
src/protocols/moves/SidechainMover.cc:159: error: 'const class core::scoring::dunbrack::SingleResidueDunbrackLibrary' has no member named 'get_psi_from_rsd'
src/protocols/moves/SidechainMover.cc:186: error: 'class core::scoring::dunbrack::DunbrackRotamerSampleData' has no member named 'chi_probability'
src/protocols/moves/SidechainMover.cc:195: error: 'class core::scoring::dunbrack::DunbrackRotamerSampleData' has no member named 'assign_random_chi'
src/protocols/moves/SidechainMover.cc: In member function 'core::Real protocols::moves::SidechainMover::proposal_density(const core::conformation::Residue&, core::Size, const core::chemical::ResidueType&, const utility::vector1 >&) const':
src/protocols/moves/SidechainMover.cc:314: error: 'const class core::scoring::dunbrack::SingleResidueDunbrackLibrary' has no member named 'get_phi_from_rsd'
src/protocols/moves/SidechainMover.cc:315: error: 'const class core::scoring::dunbrack::SingleResidueDunbrackLibrary' has no member named 'get_psi_from_rsd'
src/protocols/moves/SidechainMover.cc:330: error: 'class core::scoring::dunbrack::DunbrackRotamerSampleData' has no member named 'chi_probability'
src/protocols/moves/SidechainMover.cc:336: error: 'class core::scoring::dunbrack::DunbrackRotamerSampleData' has no member named 'chi_probability'

Also, the new backrub.cc seems to use the "devel" namespace rather than "core"

What am I doing wrong and is this an important patch?

Thanks!

Post Situation: 
Unsolved
Wed, 2010-10-27 08:08
enoee
Top
You aren't doing anything wrong; there's a bug in the patch. I've asked someone who fixed it once to remind me how they did it. (Or just tell you directly.)
Wed, 2010-10-27 10:28
smlewis
Top

Here's a reply from a relevant developer:

"It might be that this patch requires another previous patch. If there isn't some other patch that updates the core::scoring::dunbrack namespace, it's probably easiest to just ignore it and tell the user to wait for the next version. There's no reason a user absolutely needs this patch. It just prevents a crash if their resfile doesn't specify any flexible sidechains. If crashes are experienced because of this, the workaround is to add this to the command line:

-backrub::sc_prob 0

You could pass that info along if you'd like. Thanks."

For what it's worth, the "next version" should be coming soon.

Wed, 2010-10-27 14:20
smlewis