WORK-IN-PROGRESS

Content:


Introduction

6Y7M: Zhang, L., Lin, D., Hilgenfeld, R. (2020). Crystal structure of the complex resulting from the reaction between the SARS-CoV main protease and tert-butyl (1-((S)–3-cyclohexyl-1-(((S)4(cyclopropylamino)–3,4-dioxo-1-((S)-2-oxopyrrolidin-3-yl)butan-2-yl)amino)-1-oxopropan-2-yl)-2-oxo-1,2-dihydropyridin-3-yl)carbamate. To be published.


Data

This is a mmCIF-only structure - so we need to do some small detour: 1

  wget https://files.rcsb.org/download/6Y7M.cif
  sed -i "s% AAA % A %g" 6Y7M.cif
  gemmi convert 6Y7M.cif pdb6y7m.ent
  gzip pdb6y7m.ent
  fetch_PDB 6Y7M

The restraint dictionary for the ligand can be generated via

  grade_PDB_ligand OEW

or using our GRADE webserver.


Initial (re-)refinement

A series of "standard" refinement steps:

  cd 6Y7M

  refine \
    -p 6y7m.pdb \
    -m 6y7m.mtz \
    -RB \
    -l OEW.grade_PDB_ligand.cif \
    -d 01 | tee 01.lis

 refine \
    -p 01/refine.pdb \
    -m 6y7m.mtz \
    -M TLSbasic \
    -l OEW.grade_PDB_ligand.cif \
    -d 02 | tee 02.lis

 refine \
    -p 02/refine.pdb \
    -m 6y7m.mtz \
    -M TLSalternate -TLS \
    -WAT \
    -l OEW.grade_PDB_ligand.cif \
    -d 03 | tee 03.lis

gives us

aB_refine.01_summary.png

and going from

Ramachandran outliers =   0.33 %   
              favored =  96.71 %   
Rotamer outliers      =   3.41 %   
C-beta deviations     =     0   
Clashscore            =   3.78   
RMS(bonds)            =   0.0145   
RMS(angles)           =   1.94   
MolProbity score      =   1.78   

to

Ramachandran outliers =   0.00 %   
              favored =  98.03 %   
Rotamer outliers      =   3.03 %   
C-beta deviations     =     1   
Clashscore            =   2.10   
RMS(bonds)            =   0.0113   
RMS(angles)           =   1.56   
MolProbity score      =   1.35   

Looking explicitly at the ligand (using buster-report - see also here 2 ): below are some color-coded plots highlighting particular geometries that the CSD thinks (via Mogul) are suspect ('bad' = purple, 'poor' = violet, 'ok' = lime, 'good' = green, 'unknown' = gray).

Stage deposited (OCC=0.80) initial (re-)refined (OCC=0.80)
bonds MapOnly.01.report_ligand_pictures_A_401_2dschematic_mogul_bond_small.png aB_refine.01_03.report_ligand_pictures_A_401_2dschematic_mogul_bond_small.png
angles MapOnly.01.report_ligand_pictures_A_401_2dschematic_mogul_angle_small.png aB_refine.01_03.report_ligand_pictures_A_401_2dschematic_mogul_angle_small.png
torsion MapOnly.01.report_ligand_pictures_A_401_2dschematic_mogul_tors_small.png aB_refine.01_03.report_ligand_pictures_A_401_2dschematic_mogul_tors_small.png
density 3 MapOnly.01.report_ligand_pictures_A_401_electrondensity_rotated.png aB_refine.01_03.report_ligand_pictures_A_401_electrondensity.png

Full buster-report output and files

NoteThe remaining difference density after refinement is still being investigated.

(Re-)fitting the ligand

To get a second opinion about the fit of the ligand to density, we can use rhofit after stripping th ligand out of the latest model

  grep -v "OEW" aB_refine.01/03/refine.pdb > aB_refine.01/03/refine_noOEW.pdb

and using our ligand-detection feature in BUSTER:

  refine \
    -p aB_refine.01/03/refine_noOEW.pdb \
    -m 6y7m.mtz \
    -TLS \
    -L \
    -M MapOnly \
    -d aB_refine.01/minusL.01 | tee aB_refine.01/minusL.01.lis

The resulting electron density map can be used to fit the compound in:

  rhofit \
    -l OEW.grade_PDB_ligand.cif \
    -m aB_refine.01/minusL.01/refine.mtz \
    -p aB_refine.01/minusL.01/refine.pdb \
    -d rhofit.01 | tee rhofit.01.lis

The final PDB file (rhofit.01/merged.pdb) needs two small changes:

  • adding the correct LINK card back in (could also be done in Coot) since this is a covalent ligand with a bond between SG of Cys-A145 and the C57 carbon of OEW
  • setting the occupancy of all OEW atoms to 0.99 (as a marker for pdb2occ for occupancy refinement

We can now also create a correct restraints dictionary for the covalent linkage using aB_covalent_ligand:

  aB_covalent_ligand rhofit.01/merged_edit.pdb

which produces the required restraints and also shows how to use them in subsequent BUSTER refinements.

Combining all this into a new refinement command:

  pdb2occ -p rhofit.01/merged_edit.pdb -o rhofit.01/merged_edit.occ

  refine \
    -p rhofit.01/merged_edit.pdb \
    -m 6y7m.mtz \
    -TLS \
    -WAT -M WaterUpdate2ndShell \
    -l OEW.grade_PDB_ligand.cif \
    -Gelly rhofit.01/merged_edit.occ \
    MakeLINK_LinkagesFile=CYS-SG_OEW-C57.dat \
    -l CYS-SG_OEW-C57.dic \
    RunBusterDuplicatesOverride=CYO \
    -d aB_refine.01/04 | tee aB_refine.01/04.lis
 

And looking again at the ligand geometry (in comparison with the initial refinement):

Stage initial (re-)refined (see above) (re-)fitted and refined (OCC=0.66)
bonds aB_refine.01_03.report_ligand_pictures_A_401_2dschematic_mogul_bond_small.png aB_refine.01_06.report_ligand_pictures_A4000_2dschematic_mogul_bond_small_rotated.png
angles aB_refine.01_03.report_ligand_pictures_A_401_2dschematic_mogul_angle_small.png aB_refine.01_06.report_ligand_pictures_A4000_2dschematic_mogul_angle_small_rotated.png
torsion aB_refine.01_03.report_ligand_pictures_A_401_2dschematic_mogul_tors_small.png aB_refine.01_06.report_ligand_pictures_A4000_2dschematic_mogul_tors_small_rotated.png
density aB_refine.01_03.report_ligand_pictures_A_401_electrondensity.png aB_refine.01_06.report_ligand_pictures_A4000_electrondensity_crop.png

Files

 
 

Footnotes

  • *1: we will need to simplify this with better mmCIF support in future BUSTER releases
  • *2: since 11/2019 this is also part of the wwPDB validation
  • *3: the orientation of the view is defined automatically by buster-report and can be different for different refinements - due to small conformation changes