grade tutorial flavopiridol in 3blr
 
- In this example we are going to see how grade can be used to create a cif restraint dictionary for flavopiridol and how this can be used to produce a model with better ligand geometry for the flavopiridol bound to pdb entry 3blr. 
- look at the 3blr entry at rcsb
- It is a 2.80Å resolution structure
- The structure of P-TEFb (CDK9 cyclin T1), its complex with flavopiridol and regulation by phosphorylation
- The quoted Rwork 0.176 and Rfree of 0.228 are good.
- molprobity summary for the structure is respectable
Clashscore, all atoms    	14.34
Clashscore percentile    	   96
Rotamer outliers         	 2.42%
Ramachandran outliers    	 0.18%
Ramachandran favored     	95.56%
CB deviations>0.25Angs   	    0
MolProbity score         	 2.26
MolProbity percentile    	   97
Residues with bad bonds  	 0.00%
Residues with bad angles 	 0.91%
REMARK  40  MOLPROBITY OUTPUT SCORES:
REMARK  40  ALL-ATOM CLASHSCORE     :  14.34
REMARK  40  BAD ROTAMERS            :   2.4%   12/495    (TARGET  0-1%)
REMARK  40  RAMACHANDRAN OUTLIERS   :   0.2%    1/541    (TARGET  0.2%)
REMARK  40  RAMACHANDRAN FAVORED    :  95.6%  517/541    (TARGET 98.0%)
- This is very respectable indeed.
- What about the ligand?
Calculate an initial autobuster map.
- create a new directory for this tutorial
mkdir grade-tutorial
cd grade-tutorial
tar xf 3blr_tutorial.tgz
cd 3blr_tutorial
- The directory 3blr contains the pdb and mtz files 3blr/3blr.pdb and 3blr/3blr.mtz
1. Calculate an initial autobuster map
- use refine to calculate a map (time on pentium 4 3.5 minutes)
refine -p 3blr/3blr.pdb -m 3blr/3blr.mtz -M MapOnly -d 1-MapOnly.dir > 1-MapOnly.log &
- and use grade_PDB_ligand to get information about CPB from the rcsb and calculate a grade dictionary.
time grade_PDB_ligand CPB >& grade_PDB_ligand.CPB.log &
- alternatively to save time you can unpack a tar file with both results.
- while this is running move on to 2.
- or unpack 1.tar that contains the results!
2. while the refinement is running, use mogul to look at the ligand geometry in 3blr.
grep CPB 3blr/3blr.pdb > 3blr-CPB.pdb
then load the ligand into mogul.  Ask it about the rings.
mogul-ring-report.png
You might also like to ask mogul's opinion on CPB.grade_PDB_ligand.cif
3. Produce (or download) an omit-map
- load 3blr/3blr.pdb into a text editor and delete all the lines involving CPB or HOH (they're all together, near the bottom).  Save this as 3blr-dry-omit.pdb.  Then do
refine -p 3blr-dry-omit.pdb -m 3blr/3blr.mtz -d dry-omit -L | tee dry-omit.lis
where the -L option runs BUSTER in ligand-chasing mode.
4a.  Try to fit into the omit map using coot's default ligand generator
- Load dry-omit/refine.pdb and dry-omit/refine.mtz into coot
- Obtain the SMILES string for CPB
- Copy-and-paste (or type, if you are feeling masochistic) the string into the File ... Smiles menu in coot.  
- Try to fit the molecule into the right green blob using coot
4b. Try to fit into the omit map using the GRADE dictionary
- Produce a grade dictionary for CPB using the Grade Web Server http://grade.globalphasing.org/ to "Produce dictionary for an existing PDB chemical component" for code CPB
- Start a new coot with dry-omit/refine.pdb and dry-omit/refine.mtz loaded
- Load CPB.grade_PDB_ligand.cif into coot
- Extensions ... Modelling ... Molecule from dictionary will bring you a copy of CPB in a conformation that grade likes
- Turn on torsion restraints (use the R/RC button in the top right corner of the coot window)
- Try to fit the molecule into the right green blob using coot 
5a. Can autobuster tidy up the molecule entirely by itself?
- Try refining 3blr.pdb without giving any dictionary; 
refine -p 3blr/3blr.pdb -m 3blr/3blr.mtz -d no-dictionary
(or download no-dict.tgz and unpack it)
- What does the CPB ligand look like?  What does mogul think of it?
5b. Can autobuster tidy up the molecule if you give a GRADE dictionary?
refine -p 3blr/3blr.pdb -m 3blr/3blr.mtz -d grade-dict -l CPB.grade_PDB_ligand.cif | tee grade-dict.lis
(or download grade-dict.tgz and unpack it)
- What does the CPB ligand look like?  What does mogul think of it?