Modelling and forecasting of chemical properties by neural networks
The influence of the molecular structure of different substances on
chemical properties are modelled by neural networks. The objectives here are twofold:
1) to use the neural network as a predictor of the property and 2) to extract the
knowledge in the network to understand the relations between molecular structure
and property.
Chiral Separations Using Sensors and Gas Chromatography
Under construction
Switchable Molecules - Molecular Robots
Under construction
Chirbase Project
Chirbase is a database project on the separation of enantiomers (mirror image isomers) by LC,
SFC, GC, CE and TLC, respectively. First, it is a valuable tool to select a proper
chromatographic system for a given enantiomeric pair. Second, it allows access to molecular
recognition data. The flavor section deals with the different odor properties of enantiomers.
We attempt to gather all information in a systematic way from more than 90 journals. Books,
conference proceedings, personal communications etc. are also covered.
Chirbase has been developed in a joint project between the groups of Roussel at the
Universitiy Aix-Marseille III (France) and Koppenhoefer at the Universitiy of Tⁿbingen
(Germany). It is composed of the following sections:
For contributions, suggestions, complaints, error reports, ... contact the
administrator of these pages at cowinfo@uni-tuebingen.de(cowinfo@uni-tuebingen.de)
Last change: may 1997