Did marvin draw these two smiles appropriately?

I came across these two structures:


FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)\C4=C/C5=CC=CN5)=C3)=C2)=C1


FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)C4=CC5=CC=CN5)=C3)=C2)=C1


in the attached file, for each structure, on the left side is daylight drawing, on the right side is marvin drawing. The daylight drawing makes more sense to me.

Hello, the Marvin and the Daylight drawings are representing identical molecules but using different aromatization representation. If you want to have aromatized mols (circles) use the Edit/Bonds/Aromatize menu. Or do you want to get these pics from API?

Sorry I was not clear in my first post. The first smiles specified double bond configuration and the second smiles does not, however the 2nd marvin picture draws a pointed bond while the first one does not. Daylight puts a -H to specify the configuration for the first smiles and no pointed bond is drawn for the second smiles.

These two look like the same structure, the two smiles should be equivalent (ChemDraw treats it that way).





Daylight seems to treat them differently by showing H for the structure where chirality is specified.





If they are idential, jchem has a bug in treating them as different structure. If they are different, the wedge should be shown for the smiles with chirality specified, but Jchem shows the wedge for the smiles that doesn't specify chirality.

All right, I overlooked the bond and was concentrating the aromatization, (that is not relevant). As an example in a molecule represented by the SMILES string NC(=O)C=CC1=CC=CN1 Marvin will show the bond that it can be cis OR trans (undecided and can not be derived from the string only). If you do not want to have the bond to be represented by this, you can generate the picture with the 'e' option:





http://www.chemaxon.com/marvin/doc/user/cleanoptions.html#2d





e Verbose either flag removal





If it is not possible to decide whether


the double bond is CIS or TRANS, the double


bond either flag is removed





So with molconvert you can have something like:


echo "NC(=O)C=CC1=CC=CN1"|molconvert -2:O2e mrv - -o foo.mrv

The questions are:





1)


Whether FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)\C4=C/C5=CC=CN5)=C3)=C2)=C1


and


FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)C4=CC5=CC=CN5)=C3)=C2)=C1


are the exact structure or not?





We have a compound registered in our database using the first smiles, then another compound registered using the second smiles was considered a diffierent structure (based on Jcart) and assigned a different ID. However, ChemDraw consider them as the same structure and should obtain the same structure identifier.





2)


Regarding plotting, if one cannot derive cis/trans info from FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)C4=CC5=CC=CN5)=C3)=C2)=C1 , why a wedge was drawn?

Hi guys,





I am replying to the questions related to the database part.

yzhou wrote:

1) Whether FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)\C4=C/C5=CC=CN5)=C3)=C2)=C1 and FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)C4=CC5=CC=CN5)=C3)=C2)=C1 are the exact structure or not?

These two structures are different when comparing with PERFECT search (the duplicate checking in the database.) As you earlier pointed out, the first one is a Z isomer, and for the second the double bond stereo is not specified. It is also possible to run a PERFECT search using custom search option to ignore double bond stereo checking. With this, you could check duplicates ignoring double bond stereochemistry from your application prior to insertion.


(Search option for jc_compare: "t:p doubleBondStereo:N")





http://www.chemaxon.com/jchem/doc/guide/cartridge/cartapi.html#jc_compare width="90%" cellspacing="0" cellpadding="3" border="0" align="center"> yzhou wrote: We have a compound registered in our database using the first smiles, then another compound registered using the second smiles was considered a diffierent structure (based on Jcart) and assigned a different ID. However, ChemDraw consider them as the same structure and should obtain the same structure identifier. I should say that this is a bug of ChemDraw to treat the two structures identical, but you can get the same behaviour in your application using JCart with the above mentioned custom search options.

yzhou wrote:

2) Regarding plotting, if one cannot derive cis/trans info from FC(F)(F)C1=CC=CC(NC(=O)NC2=CC=CC(NC3=CC=C4C(NC(=O)C4=CC5=CC=CN5)=C3)=C2)=C1 , why a wedge was drawn?

It is a standard way of showing an unspecified E/Z isomer. Otherwise, the atom coordinates would implicitly specify E/Z stereochemistry. See this IUPAC recommendation for stereochemical notation:


"The most common convention for depicting unspecified double bonds in compounds of all sizes uses a wavy bond adjacent to the double bond."


http://www.iupac.org/publications/pac/2006/pdf/7810x1897.pdf


(Pages 1960-1961)





Kind regards,


Szabolcs

Hi,





We have just found a bug in using PERFECT search with double bond stereo switched off. Currently you can only do it by switching off all stereo matching ("t:p stereoSearch:n") or at least exact stereo matching ("t:p doubleBondStereo:N exactStereoMatching:n"). We will fix it asap.





Szabolcs

Thanks, Szabolcs





That answers the questions.

The bug of using options "t:p doubleBondStereo:n" together is fixed in JChem 3.2 which has recently been released.