User a827e3e366
10-04-2014 17:51:18
Hi,
I have checked the conversion from SDF to InChI using molconvert and I have observed some issues:
1) The molconvert tool ignores the bond stereo flags denoting that the conformation of a double bond is not specified. For example, for DB00265, the molconvert output is InChI=1S/C13H17NO/c1-4-8-13(15)14(5-2)12-10-7-6-9-11(12)3/h4,6-10H,5H2,1-3H3/b8-4+, but it should be InChI=1S/C13H17NO/c1-4-8-13(15)14(5-2)12-10-7-6-9-11(12)3/h4,6-10H,5H2,1-3H3.
2) The ChEBI datatase uses a different way to denote that the conformation of a double bond is not specified. However, there is the same issue. For example, for ChEBI:17205, the molconvert output is InChI=1S/C7H8O5/c8-5(7(11)12)3-1-2-4-6(9)10/h1,3H,2,4H2,(H,9,10)(H,11,12)/p-2/b3-1+, but it should be InChI=1S/C7H8O5/c8-5(7(11)12)3-1-2-4-6(9)10/h1,3H,2,4H2,(H,9,10)(H,11,12)/p-2.
3) The molconvert tool probably does not support axial chirality. For example, for ChEBI:38413, the molconvert output is InChI=1/C18H30O2/c1-2-3-4-5-6-7-8-9-10-11-12-13-14-15-16-17-18(19)20/h2-3,12,14H,4-11,15-17H2,1H3,(H,19,20)/b3-2+, but it should be InChI=1S/C18H30O2/c1-2-3-4-5-6-7-8-9-10-11-12-13-14-15-16-17-18(19)20/h2-3,12,14H,4-11,15-17H2,1H3,(H,19,20)/b3-2+/t13-/m1/s1.
Best regards,
Jakub Galgonek