Quantitative Structure-Property Relationship Model for Prediction of Blood-Brain Barrier Permeability

Homeostasis of the brain is maintained by the blood-brain barrier (BBB), which is a physiological barrier that prevents substances from entering the brain from the systemic circulation. However, this same protective capability can be detrimental in the effective delivery of important therapeutics to the brain. Thus, the permeation of a therapeutic across the BBB to the brain is a significant part of successful drug design. Current in vitro and in vivo techniques to determine permeation are both costly and time consuming, a computational prediction method would offer an alternative, which is more economically feasible and may provide insight into the permeation process.

Most early modeling efforts utilized traditional and linear quantitative structure-property relationship (QSPR) models; however, since most thermo-physical properties have non-linear relationships with chemical structure, traditional linear algorithms often result in inferior QSPR models. New models accounting for all the structural features of importance to BBB permeation, which avoid the pitfall of over-reliance on linear models, are required. We improve on other literature models in several aspects, including (a) use of multiple QSPR software to provide descriptors assuring both model superiority and stability, (b) automated optimization of molecular structure conformations, (c) a wholly nonlinear descriptor reduction strategy, and (d) use of robust non-linear neural networks with multiple initializations to ensure network stability.

Employing these state-of-the-art nonlinear algorithms, we have modeled accurately BBB permeation. The results obtained indicate that in silico predictions may be a viable substitute for experimental determination of BBB permeation in the initial stages of therapeutic design.