On a benchmark of 250251 compounds in the NCI00 database, mol2nam
is able to convert 233010 structures (93.11%) to names without BLAH.
Of these 233010 names, nam2mol is able to convert 221331 (94.99%)
back into structures.
This release includes a significant number of improvements to both
name generation and name parsing. For example, both name generation and
parsing now do a much better job on ring fusion nomenclature, for names
like ‘5,6,7,8-tetrahydro[1,2,4]triazolo[4,3-a]pyridine‘. There’s
also much improved handling of charged ring systems. The name parsing
conversion rate for the 71367 compound names in the 2003 Maybridge
catalog is now 93.25% in v1.6, up from 80.80% in v1.5.
In name generation, new naming styles have been added for MDL/Beilstein
AutoNom style names, for CAS permuted index style names (and there are
new placeholder styles for IUPAC79 and IUPAC93 naming). A large
number of improvements have been made to names generated using the
‘traditional’ naming style. A new OECapitalizeName
API function is available to capitalizing the appropriate first letter
of a generated name, such as ‘p-tert-Butylbenzoicacid‘.
Several bug fixes have been made to the Cahn-Ingold-Prelog (CIP)
chirality perception implementation.
The OEParseIUPACName function is now able return supplementary
locant annotations for each atom. This function now stores an integer
locant code/identifier in the integer atom type field of each atom, which
may be retrieved using the OEAtomBase.GetIntType method
and converted into a readable/displayable string using the recently
exposed OENameLocant function. This functionality is
a recent addition (obviously), and most but not all supported ring systems
and parents have locant annotations in this initial release.
Finally, for the adventurous, new APIs for translating compound names
from foreign languages into English are available as the experimental
OEFromJapanese, OEFromSwedish and
Additionally, a OEFromUTF8 function is available for
converting UTF-8 encoded strings into the escaped sequences expected by
these functions (effectively the inverse of OEToUTF8).