I just re-ran using ChemBL's canonical SMILES as input, and the resulting
.mol2 file looks much more reasonable. I now get 13 rotors, and the
conversion process is much faster.
Is InChi less mature?
I cloned the repository and rebuilt just now and got a similar result: 3
rotatable bonds.
Is having consistent NumRotors are reasonable litmus test for --gen3d?
Thanks,
Aaron
On Tue, Mar 24, 2015 at 4:48 AM, Fredrik Wallner <fred...@wallner.nu> wrote:
> Hi,
>
> That must be a but, the molecule from gen3d is a very weird molecule
> indeed, the sugar has been completely mangled. I do believe that the
> 3d-generation code is currently under revision so it might be worth it to
> recompile with the current head. Mine was a bit older than yours (~1 month)
> and had the same problem.
>
> Kind regards,
> Fredrik
>
> 24 mar 2015 kl. 06:49 skrev Aaron Schweiger <aaron.schwei...@gmail.com>:
>
> I am writing because I ran into an example where use of --gen3d apparently
> changed the number of rotatable bonds from 12 to 3.
>
> Is that to be expected or is it a possible bug? (If this indicates that
> the compound is less flexible than might be inferred from the InChi or
> SMILES that would be great. On the other hand, if the number of rotatable
> bonds can increase with --gen3d, then that is less good.)
>
> The compound is CHEMBL1096885
> <https://www.ebi.ac.uk/chembl/compound/inspect/CHEMBL1096885>. ChemBL
> has the number of rotatable bonds at 12, as does NumRotors() based on the
> InChi key.
>
> Here's is my test: (hopefully one can copy, paste and reproduce it)
>
> echo
> 'InChI=1S/C34H36F3NO13/c1-4-5-9-21(40)49-13-20(39)33(47)11-16-24(19(12-33)51-22-10-17(27(41)14(2)50-22)38-32(46)34(35,36)37)31(45)26-25(29(16)43)28(42)15-7-6-8-18(48-3)23(15)30(26)44/h6-8,14,17,19,22,27,41,43,45,47H,4-5,9-13H2,1-3H3,(H,38,46)/t14-,17-,19-,22-,27+,33-/m0/s1'
> > tmp.inchi; obabel -i inchi tmp.inchi --gen3d -o mol2 -O out.mol2 ;
> python -c 'import pybel; import sys; import itertools; [
> sys.stdout.write(str(mol.OBMol.NumRotors())+"\t"+str(mol)) for mol in
> itertools.chain(pybel.readfile("inchi","tmp.inchi"),pybel.readfile("mol2",
> "out.mol2"))];'
> 1 molecule converted
> 12 CCCCC(=O)OCC(=O)[C@@]1(Cc2c([C@H](C1)O[C@H]1C[C@@H]([C@@H]([C@H
> ](C)O1)O)N=C(C(F)(F)F)O)c(c1c(C(=O)c3cccc(c3C1=O)OC)c2O)O)O
> 3 C123CCCC(=O)OCC(=O)[C@
> @]4(Cc5c([CH](C4)(O[CH]46C[CH]([CH]7([CH](C([H]6)[H]7)(O4)[H]1)O)(/N=C(\C(F)(F)F)/O)[H]2)[H]3)c(c1c(C(=O)c2cccc(c2C1=O)OC)c5O)O)O
> *****
>
> My version information:
>
> obabel -V
> Open Babel 2.3.90 -- Feb 26 2015 -- 00:48:45
>
> Also, what does the "*****" mean from pybel's str(mol)?
>
> Aaron Schweiger
>
> p.s. I sent this to Geoff directly become becoming aware of the proper
> etiquette; apologies if already discussed.
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now.
> http://goparallel.sourceforge.net/_______________________________________________
> OpenBabel-discuss mailing list
> OpenBabel-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
>
>
>
------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
OpenBabel-discuss mailing list
OpenBabel-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
