Disabling SpanFileIO unit test suite for now which has an instability on the mac gcc platform

• Summary

Ooops! Disabling VectorizedEtable unit test suite which is not ready for prime time.

• Summary

Added pilot app for scoring protein-ligand interactions. Separately for each ligand in the pose, calculates the weighted two-body interaction energy between all residues and the given ligand. The output consists of a sorted list of ligand-residue pairs and their weighted two-body interaction energies (total, as well as per score-term), with the residues that contribute most to ligand-binding listed first. No test changes expected.

• Summary

Merging branches master and aleaverfay/vectorize_lj: 10% speed up for the etable evaluation. I'm changing the way the etable calculates the "cubic splines" that it uses to smoothly ramp the potential when it turns off. The new code evaluates the same spline polynomials, but doesn't use the cubic-spine formulation (which requires 8 additions, 13 multiplications and 1 division) but rather a cubic-polynomial formulation (which requires 3 additions and 3 multiplications). (The branch name is deceiving: the etable calculation is not being vectorized) Numeric changes result in 89 integration test trajectory changes. However, the score-function-fingerprint benchmarks do not change.

• Summary

Updating ASCII art in a doxygen tag. No changes.

• Summary

Increasing timeout for integration.debug to 60 minutes

• Summary

Rework the loops in symmetric precomputed rotamer-rotamer energy calculation to handle unusual setups. This change improves handling of strange symmetries where the independent monomer is not first in sequence. It also looks to reduce the memory footprint a bit, and may speed up packing, since it short-circuits rotamer-rotamer calculations if the score_multiply factor for that residue pair is 0. It does not apply to use of the linmem interaction graph, only the (default) precomputed energies interaction graph. Two unit tests (zinc homodimer and beta strand homodimer) show logfile changes since the interaction graph is now smaller.

• Summary

Big merge from rhiju/swm_rigid_body to master. Critical last step for stepwise monte carlo of RNA -- enables identification and sampling of docking of different chunks, a generalization of 'floating_base' move in original stepwise assembly code. Integration tests for SWA/SWM change. Numerous additional bug fixes, and clean-up of enums that were triggering warnings.

• Summary

Remove warnings from phenix interface code. Remove temp ifdefs from EllipsoidalRandomizationMover.

• Summary

Accidentally missed a file in the previous commit

• Summary

Updating boost version from 1.46 to 1.55. Removes boost 1.46 files, adds boost 1.55 files. Updates the import header script to reflect the parts of boost we are using. Two new flags are added to the default build, BOOST_ERROR_CODE_HEADER_ONLY and BOOST_SYSTEM_NO_DEPRECATED. Much thanks to Doug. With integration tests there are 8 changes. 6 are related to boost::unordered_map: - nonideal_rtmin, mr_protocols, hybridization --- via cart_bonded - rs_loophash, match --- via the 6D hasher - sequence_tolerance There are two other integration test changes: - sweep_respair_energies -- the binary .db3 files change; not replicated on other machines - noe_assign -- due to a small bugfix

• Summary

Revert "Fix issues with Gasteiger atom typing" Until I can understand the test server changes. This reverts commit cab6655af046ce82e656c1c35b9fda7c7cc9fa33, reversing changes made to b3b9b5d86a4cc4253f7e3ac9aa8e3d36429a4081.

• Summary

make reduce radii default sasa radii for SasaCalc2/IAM + IAM bugfixes NEWS: ====Sasa==== * Make the reduce radii the default for SasaCalc, and subsequently IAM as discussed at MiniCon2014 * Add calculation of sidechain sasa within SasaCalc, make it accessable to SasaCalculator2 pose metric ====IAM Bugfixes==== * Implement packstat bug workaround to keep IAM from crashing on pymol selection printing of packstat residues. This bug in packstat makes the per-residue values returned not equal the total number of pose residues. You see the problem. I am not sure what causes this to happen, but it may be from the SimpPDB that is used in packstat. For now, we do not want IAM to crash on output of this selection. * Fix taskfactory bug that would cause a crash on poses with interface disulfides if -pack_separated option was used. This stemmed from pose separation repack needing a new packertask due to call to detect_disulfides. ===Tests ==== All integration tests that use IAM will change values slightly.

• Summary

Warnings Removal: hopefully fixing error with command-line options in GCC 4.7

• Summary

RosettaMembrane Framework Updates ========================================== - Updated MembraneConformation with total polymer residue and num polymer chains - Removed unfinished scoring code - Updated protocols library level for top level movers Updated 3/6/14

• Summary

ResidueType: improving error messages for debugging This makes it easier to know how things failed upon loading in a pose. Former output example: ERROR: unrecognized residue: Glc New output examples: ERROR: No match found for unrecognized residue at position 3 Looking for upper-terminal residue with 3-letter code: Glc or: ERROR: No match found for unrecognized residue at position 1 Looking for lower-terminal residue with 3-letter code: Glc etc.

• Summary

Legend: