Re-refining 1pmq for redeposition to the PDB

Chemaxon http://www.chemicalize.org/ is exactly what is needed. Start with the PDB chemical components SMILE
CCCn1c(nc(c2ccc(Cl)c(Cl)c2)c1c3ccnc(NC4CCCCC4)n3)C5CCNCC5

Chemicalize says the ring is protonated:

In 1pmq the piperidine ring contacts phosphate of ANP modelled next to it - so the cationic piperidine makes sense.

The piperidine nitrogen atom is N19 and its existing hydrogen is H19 Lets call the extra hydrogen atom H192 position a pointer in coot and use to create 880_cation_initial.pdb. Conver to mol2 for grade use:

obabel 880_cation_initial.pdb -O880_cation_initial.mol2
grade -in 880_cation_initial.mol2 -ocif 880_cation.grade.cif -opdb 880_cation.grade.pdb -charge 1 -resname 880

Use resulting dictionary 880_cation.grade.cif for further work.

In coot markup alterate occupancy for atoms C35 C37 H37 C38 CL45 C39 CL46 C40 H40 C41 H41 using Measures, residue info, setting the occupancy for the heavy atoms to 0.6

tried in coot do a general torsion along C4 C35 to get an initial model of the B conformer but this did not work. So instead in an editor extract the alternate A atoms and modify to B change occupancy of heavy atoms to 0.4. 1pmq_03_cation880-altB.pdb. In coot rotate the ring 180 degrees around C4 C35 save coordinates.

In editor merge the B alternate coordinates. Results 1pmq_03_cation880-twoalts_init.pdb

03_occupancy_refine.Gelly needed some manual adjustment as we have zero occupancy hydrogen atoms

script: 1pmq_04_ab_cation_two_alts.sh

results:

GLU 111 A appears to be in two alternates. Tried modelling using coot split but do by modelling just and other and editing into a single pdb. Water 639 made alternate A and half occupancy as i it is where the B alternate is 1pmq_05_dataAnsisotropyServer-edit111.pdb.

Add a water molecule (HOH B 1) full occupancy H bonding between OE1 ASP 111 A conformer and NZ LYZ 93

Go through ED peaks (top already dealt with by B 1)

coordinates for re-refine 1pmq_05_dataAnsisotropyServer-edit111-coot-0-edit.pdb

the six new waters fine

ASP 111 alternates look good

N terminus 45 and 46 better than before. ED at 1 sigma. Blob could be waters or disorder protein. pusing it to interpret

progress table - keep 01, 02 and the last two .BUSTER re-refinements (up to 06)

52 GLU alter to mm-40 13% and real space fit.

55 ASP stub to CB

117 CYS and 118 LYS stub to CB

211 ALA and 212 ARG residues before an existing break. No ED at 1.0 rmsd remove both residues.

217 SER + 218 PHE weak ED at 0.8 sigma leave alone

219 MET sidechain no ED - stub to CB

220 MET weak ED at 0.8 sigma leave alone

221 THR + 222 PRO + 223 THR + 224 VAL no ED at 0.8 sigma remove all four residues.

225 VAL stub to CB

266 ARG stub to CB

288 LYS stub to CB

291 GLN stub to CB

295 ARG stub to CB

300 GLN stub to CB

301 ARG stub to CB

303 LYS stub to CB

319 PRO no density at 1.0 rmsd remove residue completely, extend to 320 ALA

321 ASP stub to CB

372 PRO and 373 PRO weak density at 0.8 rms leave alone

374 GLN remove residue no ED at 0.8

379 to 383 ED is weak remove these 5 residues

396 LYS stub to CB

400 ASN stub to CB

redo the run having restored these two side chains 1pmq_06_rebuild1-coot-4.pdb

Number of ATOM/HETATMs in this file is 2815 compared to 2977 before (removed ~5% of the structure).

place water B 1 A 1007 close to 619 HOH and 3.2 angs from N 351.

place water B 2 A 1008 between O 240 and N 240.

place water B 3 A 1009 between OG1 THR 103 and O 218

place water B 4 A 1010 3.0 Angs from N 261 (will need checking).

place water B 5 A 1011 3.0 Angs from N 50 (will need checking).

place water B 6 A 1012 close to OD2 ASP 352 and slightly far away from N 349

place water B 7 A 1013 between O 256 and O 159

place water B 8 A 1014 between N 161 and HOH 1013 (one before). Slightly close to CE MET 166

182 SER had accidentally deleted the OG atom! put it pack.

place water B 9 A 1015 between OD1 ASP 162 and NH1 ARG 165 (ED clearer than the ARG)

place water B 10 A 1016 between O 340 and O 627

Adjacent water HOH 681 is a big blob in 2Fo-Fc and has 4.7 rms difference density makes distant contacts with NH groups and has a LYS near. Likely to be anion, crystallization conditions in Scapin et al 2013 mention NaCl and MgCl₂ so try it as a chloride. CL A 2001 shape is a bit tetrahedral but there is no phosphate mentioned

51 VAL alter to m 20 % (weak ED)

127 LEU change to mt 59% - difference density for change

307 LEU change to mt 59%

388 GLU stub to CB side chain weak

137 GLU stub to CB no ED for side chain

201 CYS leave alone

Is this reasonable? Quick conquest search for NH close to Cl^- shows distances N to Cl of 3.0 to 3.5 Angstroms. So these distances are long. Conclusion leave as element type Cl but mark as UNX "unknown atom or ion"

Molprobity poor rotamers are A 51 VAL and A 201 CYS. No fix possible

373 PRO ED has appeared to allow improvement, real space refine and manuall tweak into difference density.

Coot check waters. 621 and 622 have distant contacts and difference density Mark up as UNX rather than waters

Contour at 0.9 rms, ED now looks interpretable and the sugar ring looks as if it could be tweaked into density.

load BUSTER distributed ANP.cif into coot from $BDG_home/tnt/data/common-compounds/ANP.cif

turn on torsion restraints in coot using R/RC button

coot had glitch on direct real space refine for 502 so copy it as a separate fragment.

Real space refine and drag into density:

mauve lines: start, orange/atom color after tweak. The sugar ring now better fits ED particularly for atom C5' joined to the phosphate. The adenine ring is also better placed in density. Coordinates for just the ANP after 1pmq_08_rebuild3-coot-1_remodelANP.pdb

replace ANP with the new conformation

To mark up CL A 2001 as UNK UNX atom to show ambiguity (follow format in http://www.rcsb.org/pdb/files/1AQN.pdb)

Move HOH 621 and 622 as to be UNK UNX CL type atoms

For 1pmq.pdb report is Report for 1pmq
(downloaded from
http://www.rcsb.org/pdb/d/file?b=validation&n=1pmq_full_validation.pdf.gz)

For 09 structure use the wwPDB validation server at
http://wwpdb-validation.wwpdb.org/validservice/
result report for 09 re-refined

before:

after:

Use the wwPDB validation server at
http://wwpdb-validation.wwpdb.org/validservice/
result Full wwPDB X-ray Structure Validation report for 10 re-refined

Note that the "ligand geometry bond angle outliers" in the report are questionable as they are the result of running Mogul with generalization on.

1pmq.pdb before:

model 10: