Axeleratio

Evaluating relationships
Relationships derived from huge data collections can rationally compress data sets, often support hypothesis testing, and—depending on the type of data— may provide new insights, allow predictions, and facilitate decision making. In chemistry, drug design and materials science, relationships are frequently established between structures and properties. While structure parameters define a particular material, its properties are derived, for example, by quantumchemical computation or experimental determination.

By integrating SMILES and CurlySMILES for molecule and nanostructure encoding, we investigate structure-property correlations; for example, to estimate flash points of Si- and Ge-organic compounds.