Chiral center rendering

Hello,





For the smiles string...





CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36





When I view this in Daylight Depict, it displays the H's of the two chiral centers and a wedged bond for the H's as well. However, when view in Marvin Sketch, the explicit H's are not displayed and the wedged bonds are in the ring system. My question is why?....and how can I get it to display as rendered by Depict.





Thanks.

Hi,





The given smiles string has no explicit H atoms, only implicit ones.


It seems that Depict first changes some implicit Hydrogens to explicit ones and then depict it.


In Marvin you can only add explicit Hydrogens to all atoms.


You can generate the smiles with explicit Hydrogens:

Code:

molconvert smiles:+H -s "CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36" [H]N1C2=C([H])C([H])=C([H])C([H])=C2C2=C1[C@]([H])(N1C(=O)C([H])([H])N(C(=O)[C@@]1([H])C2([H])[H])C([H])([H])[H])C1=C([H])C2=C(OC([H])([H])O2)C([H])=C1[H]

Or add explicit Hydrogens at the viewer and clean wedged bonds.





In either case the wedges will appear only at the bond to the Hydrogens.





So basically it is not possible to do the same thing what Daylight does.


You should make some programming to do the same thing:


1) Add explicit H to chiral centers only


2) Clean wedges


3) Change some more bond to wedge keeping the chirality.








Andras

Thank you for your reply. Your suggestion indeed works.





However, is there a way to replicate the Daylight depict render using the original smiles string ...make the implicit H's explicit and show only the explicit hydrogens in the structure using molconvert or MSketch?





Thanks again.

If you mean can you do it in Java, this should work for you:

Code:

String smilesIn = "CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36"; MolImporter imp = new MolImporter(); Molecule mol = imp.importMol(smilesIn); mol.hydrogenize(true); // add explicit H atoms mol.stereoClean(); // move wedged bonds to H atoms mol.implicitizeHydrogens(0); // remove unwedged H atoms only String smilesOut = mol.toFormat("smiles");

If you want the positions of the atoms in smilesOut to remain unchanged, use instead:

Code:

String smilesOut = mol.toFormat("mrv");

Almost every Marvin GUI action has a corresponding API command, and vice versa.

Thank you again.





One last follow up.





Is it possible to also show both wedged bonds (solid and dashed) at this chiral center, instead of 3 normal bonds?








Also, how can I achieve the same result with a chiral center that has no H's...for example:





C[C@H]1[C@H]([C@H](C[C@@H](O1)O[C@H]2C[C@@](CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)N)O.Cl





Thanks in advance.

speedlearn wrote:

Is it possible to also show both wedged bonds (solid and dashed) at this chiral center, instead of 3 normal bonds? Also, how can I achieve the same result with a chiral center that has no H's...for example: C[C@H]1[C@H]([C@H](C[C@@H](O1)O[C@H]2C[C@@](CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)N)O.Cl

How much programming are you willing to do to achieve the desired result? You would have to examine the bonds of each atom in the Molecule and decide whether you want to convert any of its regular bonds into bold or hashed bonds, based on the presence of other bold/hashed bonds, the relative orientation of the bonds, and the nature of that bond in the original SMILES string. A lot of work to convert the representation into something that has no more meaning than what you started with. But, yes, it is doable.

We have decided to implement it in the future, but surely not in Marvin 5.0.1.





Andras

When testing the java code suggested by Bobr:

Code:

String smilesIn = "CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36"; MolImporter imp = new MolImporter(); Molecule mol = imp.importMol(smilesIn); mol.hydrogenize(true); // add explicit H atoms mol.stereoClean(); // move wedged bonds to H atoms mol.implicitizeHydrogens(0); // remove unwedged H atoms only String smilesOut = mol.toFormat("smiles");

There seems to be no difference in the resulting molecule. The mol.implicitizeHydrogens(0) which should remove all unwedged H atoms seems to remove all H atoms. Is there a way to make it recognize only the unwedged atoms?





Thanks in advance.

The documentation for MoleculeGraph.implicitizeHydrogens(0) is incomplete. Evidently, it works only on 2D molecules. When you import a SMILES molecule, its dimension is 0, so you need to do a 2D clean before implicitzing the H atoms. This code:

Code:

String smilesIn = "CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36"; System.out.println("starting with " + smilesIn); MolImporter imp = new MolImporter(); Molecule mol1 = imp.importMol(smilesIn); mol1.clean(2, null); // convert to 2D mol1.hydrogenize(true); // add explicit H atoms mol1.stereoClean(); // move wedged bonds to H atoms mol1.implicitizeHydrogens(0); // remove unwedged H atoms only String smilesOut = mol1.toFormat("smiles"); System.out.println("after implicitize H: " + smilesOut);

gives this output:

Code:

starting with CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36 after implicitize H: [H][C@]12CC3=C(NC4=CC=CC=C34)[C@]([H])(N1C(=O)CN(C)C2=O)C5=CC6=C(OCO6)C=C5

Sorry for the earlier misinformation.

Dear All,





MoleculeGraph.implicitizeHydrogens(..) works for molecules in 0, 2, 3 dimension.


If the molecule is in 0D then it has no wedged H atoms, so it will remove all H atoms.


(Also the ones which would get wedge in 2D.) This also applies to Hydrogens in 3D.





I also would like to call your attention to the return value of mol.stereoClean(), which is false in case of 0D molecules.





Andras

volfi wrote:

Dear All, MoleculeGraph.implicitizeHydrogens(..) works for molecules in 0, 2, 3 dimension. If the molecule is in 0D then it has no wedged H atoms, so it will remove all H atoms. (Also the ones which would get wedge in 2D.) This also applies to Hydrogens in 3D. Andras

This code (same as above, but with clean() commented and some reporting added):

Code:

String smilesIn = "CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36"; System.out.println("starting with " + smilesIn); MolImporter imp = new MolImporter(); Molecule mol1 = imp.importMol(smilesIn); // mol1.clean(2, null); // convert to 2D mol1.hydrogenize(true); // add explicit H atoms System.out.println("After hydrogenizing:\n " + mol1.toFormat("smiles")); mol1.stereoClean(); // move wedged bonds to H atoms System.out.println("After moving stereo bonds to H:\n " + mol1.toFormat("smiles")); mol1.implicitizeHydrogens(0); // remove unwedged H atoms only String smilesOut = mol1.toFormat("smiles"); System.out.println("after implicitize H: " + smilesOut);

generates this output:

Code:

starting with CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36 After hydrogenizing:  [H]N1C2=C([H])C([H])=C([H])C([H])=C2C3=C1[C@]([H])(N4C(=O)C([H])([H])N(C(=O)[C@@]4([H])C3([H])[H])C([H])([H])[H])C5=C([H])C6=C(OC([H])([H])O6)C([H])=C5[H] After moving stereo bonds to H:  [H]N1C2=C([H])C([H])=C([H])C([H])=C2C3=C1[C@]([H])(N4C(=O)C([H])([H])N(C(=O)[C@@]4([H])C3([H])[H])C([H])([H])[H])C5=C([H])C6=C(OC([H])([H])O6)C([H])=C5[H] after implicitize H: CN1CC(=O)N2[C@H](CC3=C(NC4=CC=CC=C34)[C@H]2C5=CC6=C(OCO6)C=C5)C1=O

Note the presence of H atoms in the 0D molecule after hydrogenize(). Note the movement of the wedge bonds to H after stereoClean(). Note the removal of ALL H atoms after implicitizeHydrogens(0). Note that when the clean() line is uncommented, implicitizeHydrogens(0) works as advertised. Your statement is inconsistent with the results.

Thanks bobgr for all the help in troubleshooting this.





Best regards.

Hi Andras,

When do you expect to implement this? Also, why doesn't the following standardizer config achieve the desired result?

addexplicitH..wedgeclean..removeexplicitH

Moses

volfi wrote:

We have decided to implement it in the future, but surely not in Marvin 5.0.1. Andras

We have already implemented this request.

see H add Hydrogens at

http://www.chemaxon.com/marvin/help/sci/cleanoptions.html#2d

Here is a simple example:

molconvert -2:H mrv -s 'CN1CC(=O)N2[C@@H](C1=O)CC3=C([C@H]2C4=CC5=C(C=C4)OCO5)NC6=CC=CC=C36' |mview -

Also, why doesn't the following standardizer config achieve the desired resultaddexplicitH..wedgeclean..removeexplicitH

removeexplicitH also removes the wedged H so you get back your original molecule.

 

Andras

Hi Andras,

I guess my question is, isn't the default behavior of the removeexplicitH standardizer action *not* to remove Hs with wedge bonds? If that's not the default, then why provide the option to do something like this:

addexplicitH..wedgeclean..removeexplicitH:wedged

Moses

volfi wrote:

Also, why doesn't the following standardizer config achieve the desired resultaddexplicitH..wedgeclean..removeexplicitH removeexplicitH also removes the wedged H so you get back your original molecule.   Andras

Hi,

Actually it is possible to do it in API level, but I let Zsolt to answer the standardizer part, as he is the expert.

 

Andras

moses wrote:

I guess my question is, isn't the default behavior of the removeexplicitH standardizer action *not* to remove Hs with wedge bonds? If that's not the default, then why provide the option to do something like this: addexplicitH..wedgeclean..removeexplicitH:wedged Moses

Moses, you are right, "removeexplicitH" action by default does not remove wedged bonds, but wedged bonds exists only in 2D. See Andras's earlier comment:

volfi wrote:

If the molecule is in 0D then it has no wedged H atoms, so it will remove all H atoms. (Also the ones which would get wedge in 2D.) This also applies to Hydrogens in 3D.

Examples:

1. In 0D (SMILES) there are no wedged bonds, so the explicit hydrogens are removed by Standardizer.

$ standardize -c removeexplicitH "[H][C@](F)(Cl)Br"
F[C@@H](Cl)Br

2. If 2D clean is performed before explicit hydrogen removal, then the explicit hydrogens with wedged bonds are not removed by Standardizer.

 $ standardize -c clean..removeexplicitH "[H][C@](F)(Cl)Br"
[H][C@](F)(Cl)Br

I hope this helps.

Zsolt