Page 296 - Data Science Algorithms in a Week

In: Artificial Intelligence ISBN: 978-1-53612-677-8
Editors: L. Rabelo, S. Bhide and E. Gutierrez © 2018 Nova Science Publishers, Inc.

Chapter 12

ARTIFICIAL INTELLIGENCE FOR THE MODELING

AND PREDICTION OF THE BIOACTIVITIES OF

COMPLEX NATURAL PRODUCTS

*
Jose M. Prieto
UCL School of Pharmacy, London, UK

ABSTRACT

Complex natural products such as herbal crude extracts, herbal semi purified
fractions and Essential oils (EOs) are vastly used as active principles (APIs) of medicinal
products in both Clinical and Complementary/Alternative Medicine. In the food industry,
they are used to add ‘functionality’ to many nutraceuticals. However, the intrinsic
variability of their composition and synergisms and antagonisms between major and
minor components makes difficult to ensure consistent effects through different batches.
The use of Artificial Neural Networks (ANNs) for the modeling and/or prediction of the
bioactivity of such active principles as a substitute of laboratory tests has been actively
explored during the last two decades. Notably, the prediction of antioxidant and
antimicrobial properties of natural products have been a common target for researchers.
The accuracy of the predictions seems to be limited only by the inherent errors of the
modelled tests and the lack of international agreements in terms of experimental
protocols. However, with sufficient accumulation of suitable information, ANNs can
become reliable, fast and cheap tools for the prediction of anti-inflammatory, antioxidant,
antimicrobial and antiinflammatory activities, thus improving their use in medicine and
nutrition.

Keywords: artificial neural networks, natural products, bioactivity

* Corresponding Author Email: j.prieto@ucl.ac.uk.