- Home
- Industry Solutions
- Services
- Support
- About Us
The blood-brain barrier (BBB) protects the central nervous system by separating brain tissue from blood as part of absorption. The permeability of the BBB is the most important factor hindering the development of neurotherapy and has become a key issue in the prediction of chemical ADMET in recent years. The blood-brain barrier penetration is commonly expressed by log BB (log BB = log (Cbrain/Cblood)), where Cbrain and Cblood are the equilibrium concentrations of the drug in the brain and the blood, respectively. However, it is time-consuming and expensive to obtain experimental data on the BBB ratio of compounds. Thus, in silico methods for predicting BBB permeability have recently gained momentum. Such as support vector machine (SVM), multilinear regression (MLR), and artificial neural network (ANN) analysis were used to establish the quantitative prediction model of log BB, and the models show good prediction performance on the test set compounds. Based on the advanced AI-assisted platform, Creative Biolabs can accurately predict the ADMET properties and help customers accelerate the drug screening process and reduce R&D costs.
Fig.1 Anatomical structure of the blood-brain barrier (BBB). (Neumaier, et al., 2021)
Neuro-pharmaceuticals are one of the most demanding areas in the global pharmaceutical market, but the success rate of neuro-pharmaceuticals is very low compared to other therapeutic areas. One reason is that the blood-brain barrier prevents drugs from entering the brain, leaving the central nervous system underexposed. The depth and breadth of understanding of the BBB and drug interactions in the brain have accelerated in recent years. For a CNS agent to be an effective therapeutic agent, it must cross the blood-brain barrier. In contrast, peripheral acting agents must exhibit limited brain accessibility in order to avoid unnecessary central nervous system effects. The examination and prediction of brain penetration remain a key challenge in CNS and non-CNS drug discovery and development. Traditional experimental methods like in vitro or ex vivo models, including Madin-Darby Canine Kidney (MDCK) permeability assay, Caco-2 permeability assay, locus ex vivo insect-based models, Parallel Artificial Membrane Permeability Assay (PAMPA), and trans-epithelial or trans-endothelial electrical resistance (TEER) assay.
Traditional methods of testing drug BBB permeability are often expensive and time-consuming. This inefficiency has driven the development of in silico methods that use computational methods and machine learning (ML) algorithms to find and validate new compounds from molecular databases. Over the years, various deep learning (DL) models have been developed to address the BBB permeability problem and these methods generate more data and enable more automated modeling to support drug design studies. Three different machine learning algorithms were used to build classification models, including multilayer perceptron (MLP), random forest (RF), and sequential minimal optimization (SMO). MLP consists of multiple layers of neurons including an input layer, an output layer, and hidden layers between the two, that interact with weighted connections. RF predictor is made up of many decision trees and combines the output of an individual decision tree to produce a prediction. Finally, SMO is a fast algorithm for training support vector machine (SVM).
Fig.2 The four phases of developing the BBB permeability model. (Alsenan, et al., 2021)
As a reliable partner to the world's leading pharmaceutical companies and research institutions, Creative Biolabs brings together the market-leading in silico expertise to build our AI-assisted platform that can provide custom professional BBB penetration prediction.
References
