Title:CORAL: Monte Carlo Method to Predict Endpoints for Medical Chemistry
Volume: 18
Issue: 5
Author(s): Alla P. Toropova*Andrey A. Toropov*
Affiliation:
- Department of Environmental Health Science, Laboratory of Environmental Chemistry and Toxicology, IRCCS-Istituto di Ricerche Farmacologiche Mario Negri, Via La Masa 19, 20156, Milano,Italy
- Department of Environmental Health Science, Laboratory of Environmental Chemistry and Toxicology, IRCCS-Istituto di Ricerche Farmacologiche Mario Negri, Via La Masa 19, 20156, Milano,Italy
Keywords:
CORAL software, medicinal chemistry, monte carlo method, optimal descriptor, QSAR, SMILES.
Abstract: The applications of optimal molecular descriptors as a tool to predict endpoints related to
medicinal chemistry are listed. The general scheme of building up of the optimal descriptors is represented
in detail. Simplified molecular input-line entry system (SMILES) is being used to represent the
molecular architecture. The optimal descriptor is the sum of correlation weights of molecular fragments
extracted from SMILES. The numerical data on the correlation weights are calculated by the
Monte Carlo method. The data should provide maximal correlation coefficient between experimental
values of endpoint and corresponding values of the optimal descriptor. The scheme contains two phases:
(i) selection of reliable parameters of the Monte Carlo optimization; and (ii) building up a model.
The mechanistic interpretation for models based on the optimal descriptors is suggested. The interpretation
is calculated on results of several runs of the Monte Carlo optimization. The domain of applicability
for these models is defined according to the prevalence of molecular fragments in the training
and calibration sets.
