3D cleaning with MarvinView

User dd0280721c

27-07-2011 15:09:15

Dear ChemAxon support team and users,


I have been using the tools using command line in particular. I
have a problem: starting from a protonated 3D structure, exploring its
tautomers results in 2D structures, where information about isomers has been
lost. Can you suggest a way of retaining the isomer information, so that I may
end up with 3D tautomers? I've come across the "clean:3" option
available with the standardizer script. Is this an acceptable way of restoring
the original isomer? If so, how do I access this standardizer script?


Many thanks and kind regards,


Juliette

ChemAxon e08c317633

02-08-2011 14:33:28

Hi Juliette,


Try the following workflow:


3D structures ---[2D clean]---> 2D structures ---[generate tautomers]---> tautomers ---[generate 3D]---> 3D tautomers


Applications:


2D clean: Standardizer or Molconverter
generate tautomers: cxcalc
generate 3D: Standardizer, Molconverter, or cxcalc


Zsolt

ChemAxon e08c317633

02-08-2011 14:46:21

Commad line example (using unix pipes):


standardize -c clean 3dtauexample.sdf -f sdf | cxcalc tautomers -f sdf | standardize -c clean:3 -f sdf -o out.sdf

Input: 3dtauexample.sdf
Output: out.sdf


Zsolt

User dd0280721c

15-08-2011 16:35:25

Dear Zsolt,


Thank you very much for your response. I unfortunately do not have access to standardizer, so I have managed to work around the issue by using molconvert to calculate all possible isomers for each tautomer.


I have a further question: when I run the dominant tautomer calculation on the command line with no further options:


./cxcalc dominanttautomerdistribution -H 7.4 ligands.sdf > ligands_tautomers.sdf


the sdf file containing the resulting 2D tautomers does contain hydrogen atoms, but these have all coordinates as 0. My question is, how does Marvin know which tautomers are present if these hydrogens are not even "attached" to atoms in 2D?


Also, when I try and run the same calculation but with the added option of including pH effects (adding the keyword -H 7.4), the resulting 2D tautomers have had all hydrogen atoms removed, so cleaning in 3D then does not result in hydrogenated structures. Is this a bug? Or is there an extra option to add so that all hydrogens are retained?


Thank you for your help in this matter.


Kind regards,


Juliette

ChemAxon e08c317633

17-08-2011 10:55:07

Could you attach some examples (input and output files)? I cannot reproduce the error.


Which version of Marvin/JChem are you using?


Zsolt

User dd0280721c

17-08-2011 14:17:22

Dear Zsolt,


Here are my files:


- ligand.sdf (a single ligand, in 3D and with all hydrogens defined)


- get_tautomers.pl (the script I use to make calls to cxcalc)


- tautomers.sdf (what results from the dominant tautomer distribution calculation when no pH effects are considered - these are 2D structures with all H atoms with 0 coordinates)


- tautomers_pH_effect.sdf (what results from the dominant tautomer distribution calculation when pH effects are considered - these are 2D structures but all H atoms are gone)


- stereoisomers.sdf (what results from taking tautomers.sdf, and calculating stereoisomers, including converting back to 3D --> all hydrogens are now present)


- stereoisomers_pH_effect.sdf (what results from taking tautomers_pH_effect.sdf, and calculating stereoisomers,
including converting back to 3D --> still all hydrogens are gone)


I hope that you can find where I'm going wrong!


Juliette

ChemAxon e08c317633

22-08-2011 15:37:35










pradon wrote:

Here are my files:


- ligand.sdf (a single ligand, in 3D and with all hydrogens defined)


- get_tautomers.pl (the script I use to make calls to cxcalc)


- tautomers.sdf (what results from the dominant tautomer distribution calculation when no pH effects are considered - these are 2D structures with all H atoms with 0 coordinates)



It (0 coordinates) was a bug, but it has been already fixed in the latest version of Marvin.









- tautomers_pH_effect.sdf (what results from the dominant tautomer distribution calculation when pH effects are considered - these are 2D structures but all H atoms are gone)



It's a limitation. The explicit hydrogens are removedĀ  when microspecies are generated. Please use Generate3D with "prehydrogenize" option.









- stereoisomers_pH_effect.sdf (what results from taking tautomers_pH_effect.sdf, and calculating stereoisomers,
including converting back to 3D --> still all hydrogens are gone)



Same here. Explicit hydrogens are removedĀ  when microspecies are generated.


Zsolt