Developing Enhanced Blood-Brain Barrier Permeability Models: Integrating External Bio-Assay Data in QSAR Modeling

Abstract

Purpose: Experimental Blood-Brain Barrier (BBB) permeability models for drug molecules are expensive and time-consuming. As alternative methods, several traditional Quantitative Structure-Activity Relationship (QSAR) models have been developed previously. In this study, we aimed to improve the predictivity of traditional QSAR BBB permeability models by employing relevant public bio-assay data in the modeling process.

Methods: We compiled a BBB permeability database consisting of 439 unique compounds from various resources. The database was split into a modeling set of 341 compounds and a validation set of 98 compounds. Consensus QSAR modeling workflow was employed on the modeling set to develop various QSAR models. A five-fold cross-validation approach was used to validate the developed models, and the resulting models were used to predict the external validation set compounds. Furthermore, we used previously published membrane transporter models to generate relevant transporter profiles for target compounds. The transporter profiles were used as additional biological descriptors to develop hybrid QSAR BBB models.

Results: The consensus QSAR models have R(2) = 0.638 for five-fold cross-validation and R(2) = 0.504 for external validation. The consensus model developed by pooling chemical and transporter descriptors showed better predictivity (R(2) = 0.646 for five-fold cross-validation and R(2) = 0.526 for external validation). Moreover, several external bio-assays that correlate with BBB permeability were identified using our automatic profiling tool.

Conclusions: The BBB permeability models developed in this study can be useful for early evaluation of new compounds (e.g., new drug candidates). The combination of chemical and biological descriptors shows a promising direction to improve the current traditional QSAR models.

Fig. 1

Distribution of compounds by logBB values. Left (blue) are “non-permeable” compounds with logBB≤0, right (red) are “permeable” compounds with logBB>0.

Fig. 2

Modeling workflow in this study.

Fig. 3

Chemical space of logBB database (n=439) using top three principal components of MOE 2D descriptors (59% variance explained). Purple dots are “non-permeable” compounds with logBB≤0, red dots are “permeable” compounds with logBB>0.

Fig. 4

Performance of conventional QSAR (based on only MOE descriptors, represented as MOE) and hybrid models (based on both MOE descriptors and transporter assays, represented as HBD) on a five-fold cross-validation sets and b external set. Last category of each figure is the performance of the consensus model (represented as CSS). Prediction coverage was 100% in all cases.

Fig. 5

The PubChem assay response-BBB permeability correlations: (a) Heat-map for the response profiles of 275 compounds against 310 PubChem assays. The assays were sorted by predictivity to BBB permeability, and the AIDs were shown every five assays. The P_sum of each assay were calculated for the eight groups consist of 32–37 compounds with similar logBB values within each group. Outlined assays are 11 assays negatively correlated to BBB permeability (circled by green dots) and 144 assays positively correlated to BBB permeability (circled by orange dots). (b) Average P_sum values for different PubChem assays with the same compound distribution as above heat-map. (Orange line: 144 positively correlated assays, green line: 11 negatively correlated assays, yellow line: remaining 155 uncorrelated assays).