19th International CODATA Conference
Category: Keynote

Design and Data in Chemistry: The Interplay of Ideas and Facts

Johann Gasteiger (Gasteiger@chemie.uni-erlangen.de)
Computer-Chemie-Centrum, Universitaet Erlangen-Nuernberg, Germany
http://www2.chemie.uni-erlangen.de

The essential task of chemistry is not so much to produce compounds but to produce compounds with a certain property, be it a physical, chemical, or biological property. In order to do that, chemists have to answer three fundamental questions that correspond to major fields of applying computers in chemistry:

- which structure do I need for the desired property?
             structure-property/activity relationships
- how can I synthesize this structure?
             synthesis design
- what is the product of my reaction?
             reaction simulation
             structure elucidation

All these problems are highly complex and cannot be solved by calculations based on first principles. Thus, inductive learning methods have to be used to analyze the results of experiments, to analyze data.

Data have to be integrated into information and information has to be condensed into knowledge. Clearly, the amount and the quality of the available data are of high influence on the accuracy of the predictions that can be made.

The progress in the above three problem areas will be presented.