Thank you Noel, the -a option solved my issue. I made a program that fragments molecular structures by fragmenting the SMILES string instead of the actual structure. Therefore, the resulting SMILES fragments can be a bit messed up, still I want them to match the original structure. Furthermore, the SMILES that I use as input are already processed by a chemical platform (called Vega), so for consistency sake I should not modify the information.
Il giorno dom 19 feb 2023 alle ore 18:32 Noel O'Boyle <baoille...@gmail.com> ha scritto: > It would be useful to know what problem you are trying to solve here. > > OB does not support canonical Kekule SMILES, if you expect different > resonance forms to give the same canonical Kekule SMILES. Of course, you > can just write out an canonical aromatic SMILES, read it back in, and then > write it out in Kekule form (no need for canonical option). > > Regarding the second question, I don't know where you got that SMILES > from, but if you go to https://www.simolecule.com/cdkdepict/depict.html, > and paste the SMILES into the SMILES box, and the SMARTS pattern into the > SMARTS box, you will not see a match either. Both software by default apply > the Daylight aromaticity model (as best they can) leading to the bridging O > and C being aromatic. > > $ obabel -:"O=C(NCCN(C)C)c1cccc2C(=O)c3ccccc3(Oc12)" -osmi > O=C(NCCN(C)C)c1cccc2c(=O)c3ccccc3oc12 > > Note the lowercase 'o' and 'c' - this is why the SMARTS won't match. I > don't recommend it unless you know what you're doing, but OB can preserve > whatever aromaticity is in the input using the "a" input option to SMILES: > > $ obabel -:"O=C(NCCN(C)C)c1cccc2C(=O)c3ccccc3(Oc12)" -osmi -aa > O=C(NCCN(C)C)c1cccc2C(=O)c3ccccc3Oc12 > > Note that the uppercase 'O' and 'C' is preserved. Here's proof that it > matches the SMARTS: > > $ obabel -:"O=C(NCCN(C)C)c1cccc2C(=O)c3ccccc3(Oc12)" -s > "c1cccc2Cc3ccccc3(Oc12)" -osmi > 0 molecules converted > $ obabel -:"O=C(NCCN(C)C)c1cccc2C(=O)c3ccccc3(Oc12)" -s > "c1cccc2Cc3ccccc3(Oc12)" -osmi -aa > O=C(NCCN(C)C)c1cccc2C(=O)c3ccccc3Oc12 > > Regards, > Noel > > > > On Fri, 17 Feb 2023 at 18:50, Thomas <odioidenti...@gmail.com> wrote: > >> Thank you Noel. >> I wanted to get the canonical SMILES, without changing the aromaticity of >> the input SMILES: >> >> mol = pybel.readstring('smi', >> 'O=C(NCCN(C)C)C1=CC=CN2C(=O)c3ccccc3(N=C12)') >> mol.write(opt={"k": True, 'c':True}) >> 'CN(CCNC(=O)C1=CC=CN2C1=NC1C=CC=CC=1C2=O)C\t\n' >> mol.write(opt={'c':True}) >> 'CN(CCNC(=O)c1cccn2c1nc1ccccc1c2=O)C\t\n' >> >> Furthermore, can you explain me this: >> >> mol = pybel.readstring('smi', "O=C(NCCN(C)C)c1cccc2C(=O)c3ccccc3(Oc12)") >> sma = pybel.Smarts("c1cccc2Cc3ccccc3(Oc12)") >> sma.obsmarts.Match(mol.OBMol, True) >> False >> >> Thank you again >> Thomas >> >> Il giorno ven 17 feb 2023 alle ore 18:37 Noel O'Boyle < >> baoille...@gmail.com> ha scritto: >> >>> Hi Thomas, >>> >>> OB does not sanitize molecules when reading from SMILES (or any other >>> format). By default it writes aromatic SMILES though, but it sounds like >>> you want Kekule SMILES - see the obabel -Hsmi for the list of options. In >>> this case you want 'k': >>> >>> $ obabel -:"O=C1C=COC(=C1(O))C" -xk -osmi >>> O=C1C=COC(=C1O)C >>> >>> In Python, this is something like mol.write(opt={"k:" True}). >>> >>> Neither does it add Hs. A SMILES string exactly specifies the number of >>> Hs on each atom; this is preserved on reading/writing. If you could provide >>> information on a specific case, we could explain what's happening more >>> clearly. >>> >>> Regards >>> Noel >>> >>> >>> On Fri, 17 Feb 2023 at 16:18, Thomas <odioidenti...@gmail.com> wrote: >>> >>>> Is there an option to avoid sanitization of a molecule when reading >>>> from SMILES? >>>> For example I'd like the SMILES to remain unchanged if I read and write >>>> it: >>>> >>>> mol = pybel.readstring('smi', 'O=C1C=COC(=C1(O))C') >>>> mol.write() >>>> O=c1ccoc(c1O)C >>>> >>>> Beside kekulization issues, other unwanted sanitizations are the >>>> addition of Hs if I generate the molecule from SMILES fragments (partial >>>> SMILES) >>>> >>>> Thank you >>>> Thomas >>>> _______________________________________________ >>>> OpenBabel-discuss mailing list >>>> OpenBabel-discuss@lists.sourceforge.net >>>> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss >>>> >>> _______________________________________________ >> OpenBabel-discuss mailing list >> OpenBabel-discuss@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss >> >
_______________________________________________ OpenBabel-discuss mailing list OpenBabel-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
