The application of Bayesʹ theorem in natural products as a guide for skeletons identification

In this work, we present an application of Bayesʹ theorem in natural products as a guide for skeletons identification. We wrote an algorithm in PROLOG language, which computes conditional probabilities for finding skeletons types of sesquiterpene lactones in any taxon, using botanic information as input. The approach was applied to the Asteraceae family, and the results are accurate enough to allow good skeletons identification, even for the most complex chemical compound. The relationship between our approach and others procedures for skeleton identification is briefly discussed.