Advances in Chemical Biology

0503195 - ÚMG 2019 RIV CZ eng M - Monography Chapter
Škuta, Ctibor
QSAR - Searching for a relationship between a compound’s structure and its biological aktivity. Advances in Chemical Biology.
Advances in Chemical Biology. Praha: OPTIO CZ, 2019 - (Bartůněk, P.), s. 187-196. ISBN 978-80-88011-03-3
R&D Projects: GA MŠMT LO1220
Institutional support: RVO:68378050
Keywords : Quantitative structure-activity relationship modelling * QSAR classification model * small-molecular compound * biological activity
OECD category: Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Quantitative structure-activity relationship (QSAR) modeling is one of the most popular techniques of virtual screening able to predict the biological activity of small-molecular compounds. Using QSAR classification models, a compound can be labeled as active or inactive on a target, while regression models try to determine its exact activity value. In order to reveal the structure-activity relationships, almost any combination of common machine learning methods (e.g., Support Vector Machines, Random Forest, Neural Networks etc.) with various types of structure descriptors (e.g., physicochemical properties, structural keys, binary fingerprints, etc.) can be utilized. QSAR models are generally fast and are considered as reliable, providing that a correct approach to their validation and an application domain assessment are employed. Nowadays, the techniques of QSAR modeling represent a common part of computational drug design workflows used to detect new biologically active compounds, elucidate their side effects, or assess their potential toxicity risks.
Permanent Link: http://hdl.handle.net/11104/0295020