Rerefined Cocrystal Structures: Default and Tight

Dear Grand Challenge 2 Participant,

Corrected crystal structures are now available on the Grand Challenge 2 page. We apologize that this has taken so long. A summary of the issues we encountered when making the corrections is provided below, and Dr. Stephen Burley, D3R's crystallography lead, and director of the RCSB PDB, will discuss these further at the March 27 Webinar. We are rerunning the RMSD calculations, but preliminary calculations indicate that using the corrected structures changes the pose RMSDs by at most ~0.1 Angstrom.

Additionally, instructions for your Stage 2 submissions are now available via the Grand Challenge 2 page or directly here. The D3R site will open for submissions on January 18, and the deadline is February 1.

With best regards,
The D3R Team

Correction and re-refinement of the FXR co-crystal structures
There were six structures with clear errors in the initial answer set: 1lrpw/ FXR_5, 1xxar/FXR_11, 1hgaf/FXR_13, 1hvih/FXR_15, 1rfrr/FXR_19, and 1ifba/FXR_25. These resulted primarily from incorrectly assigned hybridization states, e.g., sp3 vs sp2 carbons, which were easily fixed. However, additional issues surfaced when we looked more closely at the structures in the course of making these corrections. First, the bromine atoms proved to be at reduced occupancy, presumably due to radiation damage. Second, in several structures, the electron density forces one to trade off among allowing: a) poor fit to the density; b) nonplanarity of a nominally planar ring system; or c) distortion of the bond angle at an sp3 carbon. We explored several possible refinement procedures before arriving at a single approach that was used to re-refine all of the structures, to generate the new set of answers. These are in the Main-Default folder. In the course of exploring the planarity issue, we additionally re-refined all of the structures with stronger planarity restraints. (For the crystallographers among you: the restraints were tightened from the default standard deviation of 0.02 Angstroms to a reduced value of 0.005 Angstroms.) Strengthening these restraints had a negligible effect on most of the structures, but in a few cases (1ymrc, 1ytut, 1sjpr, 1xxar, 1hvih, and maybe 1hoia) , it caused several atoms to shift by up to ~0.1 Angstrom, and visibly improved planarity. The Supplementary Information (SI-Tight) folder provides the full set of structures refined with the tightened planarity restraints.