In a new training material for RiskGone, a neural network provides an overview of QSAR models and how they can be used. QSAR models define the relationship between the variance in molecular structures and the variance in a modelled biological activity for a group of sufficiently similar compounds. Models can be used to obtain missing data describing the physical chemical properties or activity of compounds. They can predict the modelled activity for untested chemicals without the necessity of providing experiments. The advantages of QSAR methods include a reduction in the cost of products on the market, a reduction of time needed to conduct experiments, a reduction of the need for experimental research using animals, and a reduction in waste caused by experiments.
Watch the presentation to find out more!