In silico prediction of substrate properties for ABC-multidrug transporters

Abstract

Overexpression of ABC (ATP-binding cassette)-type drug efflux pumps, such as ABCB1, ABCC1 and ABCG2 in cancer cells confers multi-drug resistance (MDR) and represents a major cause of treatment failures in cancer therapy. Furthermore, there is increasing evidence for the important contribution of ABC-transporters to bioavailability, distribution, elimination and blood–brain barrier permeation of drug candidates. This review presents an overview on the different computational methods and models pursued to predict ABC-transporter substrate properties of drug-like compounds. They range from linear discriminant analysis to pharmacophore modelling and machine learning algorithms. Many of these models show a satisfying performance within the study-specific, defined chemical space but general applicability for the whole drug-like chemical space still needs to be proven. First attempts aiming towards selectivity profiling for ligands of the two polyspecific transporters ABCB1 and ABCG2 is also discussed. This might pave the way for a pharmacological profiling of compound series with special focus on their ADMET (absorption, distribution, metabolism, excretion and toxicity) properties.

Keywords::

• ABCB1
• ABC-transporter
• machine learning
• P-glycoprotein
• pharmacophore modelling
• substrate
• support vector machine

Notes