The data of Release 1 and Release 2.0 are very different. The most important points are
- Better 2D-3D converter
The 2D-3D converters were changed. The program of CONCORD standalone was used in Release 2.0
(The programs of CONCORD sketch and MOE were used in Release1). Among all methods that we tested,
CONCORD standalone gave the best performance in converting structures.
- Confirmed 3D structures with respect to chirality
During 2D-3D conversion, many structures were changed with respect to chirality. We have
determined how to confirm conservation of chiralities. In Release 2.0, we used InChI and
canonical SMILES for these determinations. These two programs were the best choice for the
identification of canonicalized compound structures among all methods we tested. In Release
1, this check was performed by strings of unique SMILES by SYBYLTM
and aRSChirality by MOETM .
- Increased descriptor numbers
In Release 2.0, the descriptors were calculated by MOETM.
The package can calculate many descriptors. Some descriptors in the previous version
were substituted, such as CLogP (to logP(o/w) and SlogP) and CMR (to SMR).
- 3D-Structures are visible from Macintosh computers
In the previous version, MDL Chime was used to visualize 3D structures. In Release 2.0, we
adopted Jmol which allows the structures to be viewed from all types of computers.
- Possible differences in 3D-structures
Because of changes of the conversion methods, the resulting 3D-structures may differ, even if
the same 3DMET ID is used. In Release 2.0, 3DMET recorded only the structures that were newly