Identifying blood‐brain barrier peptides by using amino acids physicochemical properties and features fusion method

Abstract

Blood-brain barrier peptides (BBPs) play a promising role in current drug study of central nervous system diseases. Hence, it is an urgent need to rapidly and accurately discriminating BBPs from non-BBPs. Experimental approaches are the first choice, however, these methods are expensive and take a lot of time. Thus, more and more researchers focused their attention on computational models. In current work, we developed a support vector machine (SVM) based model to identify BBPs. First, amino acids physicochemical properties were employed to represent peptide sequences, and Pearson's correlation coefficient (PCC) and maximal information coefficient (MIC) were applied to extract useful information. Then, similarity network fusion algorithm was utilized to integrate these two different kinds of features. Next, Fisher algorithm was used to pick out the discriminative features. Finally, these selected features were input into SVM for distinguishing BBPs from non-BBPs. The proposed model achieved 100.00% and 89.47% classification accuracies on training and independent datasets, respectively. Additionally, we found that pK2 (NH3) property of amino acid plays a key role in discriminating BBPs from non-BBPs. The results showed that our proposed method is effective, and achieved a significantly improvement in identifying BBPs, as compared with the state-of-the-art approach. The Matlab codes and datasets are freely available at https://figshare.com/articles/ online_resource/iBBPs_zip/14723766.blood-brain barrier peptides, MIC, PCC, similarity network fusion algorithm, support vector machine | INTRODUCTIONCurrently, central nervous system (CNS) diseases, such as Alzheimer's disease, [1,2] Parkinson's disease, [1,2] and brain tumor [3] are a serious threat to people's health. With the increasing of incidence of CNS diseases, more and more resource was devoted to study and treat CNS diseases.It is reported that CNS diseases account for about 25% of the burden of all diseases in high-income countries and Europe. [4] At present, the blood-brain barrier (BBB) is one of the key points in treatment of CNS diseases, which prevents drugs to arrive at their targets. [3,5] Fortunately, recent study found that some peptides have the ability to cross the BBB, these peptides were called as blood-brain barrier peptides (BBPs). [6] BBPs play an important role in CNS study and often show hormesis, [7] which provide a new idea to diagnose and treat brain diseases. Compared with conventional drugs, BBPs with the following advantages that are low immunogenicity and toxicity. [7,8] Hence, BBPs may be an ideal choice for the treatment of CNS diseases.Up to now, many efforts have been made in investigating BBPs, however, it is still a challenging task to identify BBPs. Experimental approaches are the first choice in BBPs identification, but these methods are expensive and take a lot of time, so researchers turned their attention to computational approaches. In fact, various computational models

Molecular determinants for brain targeting by peptides: a meta-analysis approach with experimental validation

“…These BBB-crossing peptides do not simply engage in cell membrane interaction, via AMT or RMT, but can achieve transcytosis through other distinct pathways, such as paracellular, direct diffusion or carrier-mediated transport [ 4 , 13 , 28 – 31 ]. However, as the main physicochemical features that allow BBB transcytosis have not been investigated in sufficient detail, researchers may rely excessively—and controversially—on analogy in assuming that CPPs are likewise able to traverse cell barriers [ 4 , 32 , 33 ].…”

Biological Membrane-Penetrating Peptides: Computational Prediction and Applications

“…However, cell-penetrating properties of peptides do not necessarily imply the ability to penetrate the BBB (Stalmans et al, 2015). Studies have demonstrated that both classes of peptides comprise different structural and physicochemical properties (de Oliveira et al, 2021;Zou, 2021). The diffusing of peptides across the BBB have been better correlated with hydrogen bonding and water desolvation than logP (Chikhale et al, 1994).…”

“…Similarly, Kumar et al proposed the B3Pred, a computational tool to evaluate the prediction of the permeability of three datasets of natural and synthetic peptides that uses as molecular descriptors some structure-based properties combined with multiple ML algorithms, such as DT, RF, LR, KNN, Gaussian Naive Bayes (GNB), XGB, and SVM (Kumar et al, 2021a). Zou (2021) proposed using SVM to identify B3PPs employing physicochemical properties of the peptides. The authors used Pearson's correlation coefficient and maximal information coefficient to extract useful information to predict the permeability, which was integrated by a similarity network fusion algorithm (Zou, 2021).…”

“…Zou (2021) proposed using SVM to identify B3PPs employing physicochemical properties of the peptides. The authors used Pearson's correlation coefficient and maximal information coefficient to extract useful information to predict the permeability, which was integrated by a similarity network fusion algorithm (Zou, 2021). Recently, He et al (2021) proposed a mutual information maximization meta-learning (MIMML) algorithm, a novel deep meta-learning method that predicts bioactive peptides using sequence-based descriptors and applies mutual information maximization and convolution kernel.…”

BrainPepPass: A Framework Based on Supervised Dimensionality Reduction for Predicting Blood-Brain Barrier-Penetrating Peptides

“…A similarity network fusion algorithm was utilized to integrate these two different types of features, followed by a Fisher algorithm to select the discriminative features. Finally, these selected features were input into support vector machines (SVMs) to distinguish B3PPs from non-B3PPs . However, despite the valuable information provided by these in silico methods for predicting peptide penetration of the BBB, the lack of computational tools to predict this pharmacokinetic property for both natural and chemically modified peptides hampered the efficient exploration of biotechnological and pharmaceutical applications of these molecules against several brain diseases.…”