This method provides a possibility to correlate molecular properties of substances with experimental data. If for example a test series of experimental data is available for different substances a model can be built from the molecular properties of all substances and the corresponding experimental value.

With this model the macroscopic properties of unknown compounds can be predicted effectively with computer calculations. Quantitative-Structure-Activity-Relationships lead to a systematic optimization of desired properties in dependence of the deployed substances or combination of these.

These methods became more and more popular in industrial research in the recent years due to their benefit when decreasing development times within the R&D process. The experimentalist can avoid dead-ends using these computer simulations. These methods are used in many areas of Materials and Life Science.