Qualitative and quantitative measures of drugs’ placenta permeability - A chromatographic and computational approach

Novel Quantitative Structure-Activity Relationship (QSAR) models of compounds’ placenta (PL) permeability expressed as their: i) log FM (fetus-to-mother blood concentration) values; ii) Clearance Index (CI); iii) binary PL1/0 (crossing/non-crossing; Yes/No) score were generated based on chromatographic and computational descriptors. Chromatographic data were collected using Micellar Liquid Chromatography; computational descriptors were calculated using SwissADME and Mordred software available freely on-line.  Suitable computational descriptors were selected using Partial Least Square (PLS) technique. Predictive models of compounds’ log FM and CI were generated using a number of statistical tools: Multiple Linear Regression, Boosted Trees, and Artificial Neural Networks. Binary (qualitative) models of compounds’ placenta permeability (Yes/No) were generated using Discriminant Function Analysis and Principal Component Analysis.  All models were validated using a test set of compounds that were not used for model building.  Compounds of interest include drugs and environmental pollutants, e.g. pesticides. 

Audience Take Away Notes:

• The ability of compounds to cross the placenta may be predicted using chromatographic and computational data
• No human or animal experiments are needed
• Proposed models are applicable to drugs and environmental contaminants (e.g. pesticides) alike