fsentin / diterpene Goto Github PK

Technische Universität Wien, Theoretical Foundations and Research Topics in Machine Learning

Diterpenes belong to the class of terpenes, which are molecules with a carbon skeleton and the formula (C5H8)n. They constitute a major component of essential oils in certain plants and often exhibit interesting medical properties. NMR spectroscopy is a commonly employed method to determine the chemical structure of a compound, something that is referred to as structure elucidation. In NMR experiments, molecules are placed in a strong magnetic field, resulting in the molecules resonating at a specific frequency. These frequencies can then be used to infer information about the molecules' chemical structures.

The full dataset used comprises of 1503 spectra of diterpenes, classified into 23 different classes according to their skeleton structure. For a more detailed description of the classification task and the data, please refer to the original paper:

Saso Dzeroski, Steffen Schulze-Kremer , Karsten R. Heidtke, Karsten Siems, Dietrich Wettschereck & Hendrik Blockeel. (1998). Diterpene structure elucidation from 13cnmr spectra with inductive logic programming. Applied Artificial Intelligence, 12(5). 363-383. DOI: 10.1080/088395198117686