QSAR model for blood-brain barrier permeation.

BACKGROUND AND OBJECTIVE: Predicting blood-brain barrier permeability for novel compounds is an important goal for neurotherapeutics-focused drug discovery. It is impossible to determine experimentally the blood-brain barrier partitioning of all possible candidates. Consequently, alternative evaluation methods based on computational models are desirable or even necessary. The CORAL software (https://www.insilico.eu/coral) has been checked up as a tool to build up quantitative structure - activity relationships for blood-brain barrier permeation.

METHODS: The Monte Carlo technique gives possibility to build up predictive model of an endpoint by means of selection of so-called correlation weights of various molecular features. Descriptors calculated with these weights are basis for correlations "structure-endpoint".

RESULTS: The approach gives good models for three random splits into the training and validation sets. The best model characterized by the following statistics for the external validation set: the number of compounds is 41, determination coefficient is equal to 0.896, root mean squared error is equal to 0.175.

CONCLUSIONS: The suggested approach can be applied as a tool for prediction of blood-brain barrier permeation.