Update metal containing entries #58
Conversation
|
I meant they should have been sorted in this branch. This needs to be confirmed. |
|
There don't seem to be any angle restraint for the CL1 atom in 4IR which causes distortions in refinement (testcase 3pk2) |
|
Regarding the point made in the wiki: leaving out "placeholder" atoms is risky for compounds where not always the same atoms leave. You could have a look at cisplatin for that as an example. |
|
We will need to explore alternative options. |
|
So far the cases I looked at seem to be an improvement, sometimes very big. AG1 presents an interesting problem: if you look at PDB entry 2jld, you can see that the chirality of C6 is wrong, albeit consistent with the definition. In the associated paper you see that the main text has the chirality in the model, but the supplemental data about the synthesis has the inverted chirality. Changing the chirality to 'both' in the restraint file allows you to improve the model. |
|
@drlemmus I think one of the next steps should be validation of all pdb entries. Only after we have finished this stage: establish procedures and update dictionaries. We are close to that stage. |
Some manual works were needed. Note from Garib: SXC: acedrg with options --c0 and --c1, metalCoord using PDB. Still angle in the model are odd. But dictionary looks fine NXC: acedrg with options --c0 and --c1, metalCoord using PDB. HB1: removed H1 and changed the ring - added -1 charge, changed bond orders and added one H atom R5N: added a double bond on Si. acedrg with options --c0 and --c1, metalCoord using PDB. metalCoord with --cif B13: in metalCoord -cl square-pyramid to force this coordination geometry. OBL: in metalCoord -cl square-pyramid to force this coordination geometry. SVP: not clear how to proceed V4A: Angles are fine now. But it took several rounds of servlcat update to get the current version O39: in mealCoord -cl octahederal to force this coordination geometry MP1: Manually removed C1 from the plane. in metalCoord -c 4 --cl square-planar to force this geometry KYT: Bonds between Ru1 and benzene ring were missing. Added missing bonds. --cif added. Order of atoms in the PDB and monomer cif are different E5O: removed BR1 from the file. -c 4 (for atom coordination) and default gave simila results (default gave square-prism with CD2 from the protein as a false connection) NYN: N atoms had -1 charge (S atoms also). Removed charges. There are still prolems. N1 becomes with -2 charge. Not good. I need to add H atoms
|
Updated again. I think now this branch is ready for review. Please report any issues. |
GaribMurshudov
left a comment
GaribMurshudov
left a comment
There was a problem hiding this comment.
Hopefully these are good. At least they are better than current metal containing monomers
Finally we will have reasonable restraints for metal containing entries. See the wiki article https://github.com/MonomerLibrary/monomers/wiki/Dealing-with-metals
Probably we will update all again before merging.
Should sort #23 and #36.