Drug-target Binding Research Articles

NerLTR-DTA: drug-target binding affinity prediction based on neighbor relationship and learning to rank.

Drug-target interaction prediction plays an important role in new drug discovery and drug repurposing. Binding affinity indicates the strength of drug-target interactions. Predicting drug-target binding affinity is expected to provide promising candidates for biologists, which can effectively reduce the workload of wet laboratory experiments and speed up the entire process of drug research. Given that, numerous new proteins are sequenced and compounds are synthesized, several improved computational methods have been proposed for such predictions, but there are still some challenges. (i) Many methods only discuss and implement one application scenario, they focus on drug repurposing and ignore the discovery of new drugs and targets. (ii) Many methods do not consider the priority order of proteins (or drugs) related to each target drug (or protein). Therefore, it is necessary to develop a comprehensive method that can be used in multiple scenarios and focuses on candidate order. In this study, we propose a method called NerLTR-DTA that uses the neighbor relationship of similarity and sharing to extract features, and applies a ranking framework with regression attributes to predict affinity values and priority order of query drug (or query target) and its related proteins (or compounds). It is worth noting that using the characteristics of learning to rank to set different queries can smartly realize the multi-scenario application of the method, including the discovery of new drugs and new targets. Experimental results on two commonly used datasets show that NerLTR-DTA outperforms some state-of-the-art competing methods. NerLTR-DTA achieves excellent performance in all application scenarios mentioned in this study, and the rm(test)2 values guarantee such excellent performance is not obtained by chance. Moreover, it can be concluded that NerLTR-DTA can provide accurate ranking lists for the relevant results of most queries through the statistics of the association relationship of each query drug (or query protein). In general, NerLTR-DTA is a powerful tool for predicting drug-target associations and can contribute to new drug discovery and drug repurposing. The proposed method is implemented in Python and Java. Source codes and datasets are available at https://github.com/RUXIAOQING964914140/NerLTR-DTA.

VCOMBAT: a novel tool to create and visualize a computational model of bacterial antibiotic target-binding

BackgroundAs antibiotic resistance creates a significant global health threat, we need not only to accelerate the development of novel antibiotics but also to develop better treatment strategies using existing drugs to improve their efficacy and prevent the selection of further resistance. We require new tools to rationally design dosing regimens from data collected in early phases of antibiotic and dosing development. Mathematical models such as mechanistic pharmacodynamic drug-target binding explain mechanistic details of how the given drug concentration affects its targeted bacteria. However, there are no available tools in the literature that allow non-quantitative scientists to develop computational models to simulate antibiotic-target binding and its effects on bacteria.ResultsIn this work, we have devised an extension of a mechanistic binding-kinetic model to incorporate clinical drug concentration data. Based on the extended model, we develop a novel and interactive web-based tool that allows non-quantitative scientists to create and visualize their own computational models of bacterial antibiotic target-binding based on their considered drugs and bacteria. We also demonstrate how Rifampicin affects bacterial populations of Tuberculosis bacteria using our vCOMBAT tool.ConclusionsThe vCOMBAT online tool is publicly available at https://combat-bacteria.org/.

MGraphDTA: deep multiscale graph neural network for explainable drug-target binding affinity prediction.

Predicting drug–target affinity (DTA) is beneficial for accelerating drug discovery. Graph neural networks (GNNs) have been widely used in DTA prediction. However, existing shallow GNNs are insufficient to capture the global structure of compounds. Besides, the interpretability of the graph-based DTA models highly relies on the graph attention mechanism, which can not reveal the global relationship between each atom of a molecule. In this study, we proposed a deep multiscale graph neural network based on chemical intuition for DTA prediction (MGraphDTA). We introduced a dense connection into the GNN and built a super-deep GNN with 27 graph convolutional layers to capture the local and global structure of the compound simultaneously. We also developed a novel visual explanation method, gradient-weighted affinity activation mapping (Grad-AAM), to analyze a deep learning model from the chemical perspective. We evaluated our approach using seven benchmark datasets and compared the proposed method to the state-of-the-art deep learning (DL) models. MGraphDTA outperforms other DL-based approaches significantly on various datasets. Moreover, we show that Grad-AAM creates explanations that are consistent with pharmacologists, which may help us gain chemical insights directly from data beyond human perception. These advantages demonstrate that the proposed method improves the generalization and interpretation capability of DTA prediction modeling.

Drug-target binding affinity prediction method based on a deep graph neural network.

The development of new drugs is a long and costly process, Computer-aided drug design reduces development costs while computationally shortening the new drug development cycle, in which DTA (Drug-Target binding Affinity) prediction is a key step to screen out potential drugs. With the development of deep learning, various types of deep learning models have achieved notable performance in a wide range of fields. Most current related studies focus on extracting the sequence features of molecules while ignoring the valuable structural information; they employ sequence data that represent only the elemental composition of molecules without considering the molecular structure maps that contain structural information. In this paper, we use graph neural networks to predict DTA based on corresponding graph data of drugs and proteins, and we achieve competitive performance on two benchmark datasets, Davis and KIBA. In particular, an MSE of 0.227 and CI of 0.895 were obtained on Davis, and an MSE of 0.127 and CI of 0.903 were obtained on KIBA.

Mechanisms of antibiotic action shape the fitness landscapes of resistance mutations

Antibiotic-resistant pathogens are a major public health threat. A deeper understanding of how an antibiotic’s mechanism of action influences the emergence of resistance would aid in the design of new drugs and help to preserve the effectiveness of existing ones. To this end, we developed a model that links bacterial population dynamics with antibiotic-target binding kinetics. Our approach allows us to derive mechanistic insights on drug activity from population-scale experimental data and to quantify the interplay between drug mechanism and resistance selection. We find that both bacteriostatic and bactericidal agents can be equally effective at suppressing the selection of resistant mutants, but that key determinants of resistance selection are the relationships between the number of drug-inactivated targets within a cell and the rates of cellular growth and death. We also show that heterogeneous drug-target binding within a population enables resistant bacteria to evolve fitness-improving secondary mutations even when drug doses remain above the resistant strain’s minimum inhibitory concentration. Our work suggests that antibiotic doses beyond this “secondary mutation selection window” could safeguard against the emergence of high-fitness resistant strains during treatment.

FusionDTA: attention-based feature polymerizer and knowledge distillation for drug-target binding affinity prediction.

The prediction of drug-target affinity (DTA) plays an increasingly important role in drug discovery. Nowadays, lots of prediction methods focus on feature encoding of drugs and proteins, but ignore the importance of feature aggregation. However, the increasingly complex encoder networks lead to the loss of implicit information and excessive model size. To this end, we propose a deep-learning-based approach namely FusionDTA. For the loss of implicit information, a novel muti-head linear attention mechanism was utilized to replace the rough pooling method. This allows FusionDTA aggregates global information based on attention weights, instead of selecting the largest one as max-pooling does. To solve the redundancy issue of parameters, we applied knowledge distillation in FusionDTA by transfering learnable information from teacher model to student. Results show that FusionDTA performs better than existing models for the test domain on all evaluation metrics. We obtained concordance index (CI) index of 0.913 and 0.906 in Davis and KIBA dataset respectively, compared with 0.893 and 0.891 of previous state-of-art model. Under the cold-start constrain, our model proved to be more robust and more effective with unseen inputs than baseline methods. In addition, the knowledge distillation did save half of the parameters of the model, with only 0.006 reduction in CI index. Even FusionDTA with half the parameters could easily exceed the baseline on all metrics. In general, our model has superior performance and improves the effect of drug-target interaction (DTI) prediction. The visualization of DTI can effectively help predict the binding region of proteins during structure-based drug design.

Current Approaches of Building Mechanistic Pharmacodynamic Drug-Target Binding Models.

Mechanistic pharmacodynamic models that incorporate the binding kinetics of drug-target interactions have several advantages in understanding target engagement and the efficacy of a drug dose. However, guidelines on how to build and interpret mechanistic pharmacodynamic drug-target binding models considering both biological and computational factors are still missing in the literature. In this chapter, current approaches of building mechanistic PD models and their advantages are discussed. We also present a methodology on how to select a suitable model considering both biological and computational perspectives, as well as summarize the challenges of current mechanistic PD models.

Prediction of Drug-target Binding Affinity by An Ensemble Lear ning System with Network Fusion Information

Background: Verifying interactions between drugs and targets is key to discover new drugs. Many computational methods have been developed to predict drug-target interactions and performed successfully, but challenges still exist in the field. Objective: We have made an attempt to develop a machine learning method to predict drug-target affinity, which can determine the strength of the binding relationship between drug and target. Methods: This paper proposes an integrated machine learning system for drug-target binding affinity prediction based on network fusion. First, multiple similarity networks representing drugs or targets are calculated. Second, multiple networks representing drugs (targets) are fused separately. Finally, the characteristic information of splicing drugs and targets was used for model construction and training. By integrating multiple similarity networks, the model fully embodies the complementarity of network information, and the most complete features of information can be obtained after the redundancy is removed. Results: Experimental results showed that our model obtained good results for DTI binding affinity. Conclusion: It is still challenging to predict drug-target affinity. This paper proposes to use an integrated system of fusion network information for addressing the issue, and the proposed method performs well, which can provide a certain data basis for the subsequent work.

Modelling Drug-Target Binding Affinity using a BERT based Graph Neural network.

Understanding the interactions between novel drugs and target proteins is fundamentally important in disease research as discovering drug-protein interactions can be an exceptionally time-consuming and expensive process. Alternatively, this process can be simulated using modern deep learning methods that have the potential of utilising vast quantities of data to reduce the cost and time required to provide accurate predictions. We seek to leverage a set of BERT-style models that have been pre-trained on vast quantities of both protein and drug data. The encodings produced by each model are then utilised as node representations for a graph convolutional neural network, which in turn are used to model the interactions without the need to simultaneously fine-tune both protein and drug BERT models to the task. We evaluate the performance of our approach on two drug-target interaction datasets that were previously used as benchmarks in recent work.Our results significantly improve upon a vanilla BERT baseline approach as well as the former state-of-the-art methods for each task dataset. Our approach builds upon past work in two key areas; firstly, we take full advantage of two large pre-trained BERT models that provide improved representations of task-relevant properties of both drugs and proteins. Secondly, inspired by work in natural language processing that investigates how linguistic structure is represented in such models, we perform interpretability analyses that allow us to locate functionally-relevant areas of interest within each drug and protein. By modelling the drug-target interactions as a graph as opposed to a set of isolated interactions, we demonstrate the benefits of combining large pre-trained models and a graph neural network to make state-of-the-art predictions on drug-target binding affinity.

G Protein-Coupled Receptor-Ligand Dissociation Rates and Mechanisms from τRAMD Simulations.

There is a growing appreciation of the importance of drug-target binding kinetics for lead optimization. For G protein-coupled receptors (GPCRs), which mediate signaling over a wide range of time scales, the drug dissociation rate is often a better predictor of in vivo efficacy than binding affinity, although it is more challenging to compute. Here, we assess the ability of the τ-Random Acceleration Molecular Dynamics (τRAMD) approach to reproduce relative residence times and reveal dissociation mechanisms and the effects of allosteric modulation for two important membrane-embedded drug targets: the β2-adrenergic receptor and the muscarinic acetylcholine receptor M2. The dissociation mechanisms observed in the relatively short RAMD simulations (in which molecular dynamics (MD) simulations are performed using an additional force with an adaptively assigned random orientation applied to the ligand) are in general agreement with much more computationally intensive conventional MD and metadynamics simulations. Remarkably, although decreasing the magnitude of the random force generally reduces the number of egress routes observed, the ranking of ligands by dissociation rate is hardly affected and agrees well with experiment. The simulations also reproduce changes in residence time due to allosteric modulation and reveal associated changes in ligand dissociation pathways.

Multilevel Attention Models for Drug Target Binding Affinity Prediction

Multilevel Attention Models for Drug Target Binding Affinity Prediction

Drug-target continuous binding affinity prediction using multiple sources of information

Drug-target binding affinity prediction has a significant role in the search for new drugs or novel targets for existing drugs. The vast majority of recent computational approaches, presented for the task of drug-target binding affinity prediction, make use of a single source to measure drug-drug or protein-protein similarities. Incorporating various information sources is of the essence for improving the accuracy of drug-target prediction. The main objective of this research is to propose a method for combining the information provided from various similarity measures for drug-drug and protein-protein similarities and to show that this leads to better prediction performance. For this purpose, we propose a method that makes use of five drug-drug and five protein-protein similarity measures simultaneously to predict the binding affinity value of an input query drug-protein interaction. In the proposed method, using each pair of drug-drug and protein-protein similarity measures, k-nearest neighbor algorithm is used to find k drug-protein pairs most similar to the input interaction. The information regarding the binding affinity values of neighbors and their similarities are fed as features to a gradient boosting machine to construct the regression model. To assess the performance of our method in comparison with state-of-the-art methods in the literature, three related benchmark datasets were used. The experimental results in various settings (pairwise, new drug, and new target scenarios) indicate the superiority of the proposed method in comparison with other methods proposed in the literature.

Using drug-drug and protein-protein similarities as feature vector for drug-target binding prediction

The success of the drug discovery process heavily depends on the identification of drug-target interaction. Binding affinity indicates the strength of the interaction between a drug and its target. The drug target identification with experimental approaches is costly and time-consuming, so computational approaches have been devised to facilitate and accelerate the drug discovery process. Recent computational methods address drug target identification as a regression-based problem and try to predict binding affinity. This work aims to demonstrate that using important similarity values can improve the accuracy of prediction without the need for difficult feature extraction processes. Also, we predict binding affinity by focusing on the realist situations, where drugs or targets or both of them are not observed in the training set. While the significant role of similarity and realistic situations don't get attention in the majority of previous approaches. To this end, a simple and efficient algorithm is proposed to predict binding affinity based on similarity information. First, we compute drug-drug similarity and target-target similarity matrices. Then, our regression model for predicting binding affinity is constructed by minimizing the root mean square error, and the gradient boosting (GBM) machine is used for optimization. The similarity values are used to create similarity-based chemogenomic features and as the inputs of the gradient boosting machine. To compare the performance of our suggested algorithm with the state-of-the-art methods three related benchmark datasets were used. The superiority of experimental results indicates that using GBM as our regression model with similarity-based features creates an effective model for predicting binding affinity.

EA-based hyperparameter optimization of hybrid deep learning models for effective drug-target interactions prediction

The identification of drug-target interactions (DTIs) is an important process in drug repositioning and drug discovery. However, it is very expensive and time-consuming to determine all possible DTIs with experimental approaches. Most existing machine learning-based methods formulate the DTIs prediction problem as a binary classification problem. Nevertheless, the lack of experimentally validated negative samples results in imbalanced class distribution within the datasets, which may have a negative influence on the DTI prediction performance. Casting DTI prediction task as a regression problem seems an interesting alternative to avoid this issue especially with the recent increase in protein structural data and DTI datasets. Within this context, a twofold contribution is described in this paper. First, we propose a novel deep learning model for predicting drug-target binding affinities called “Convolution Neural Network with Attention-based bidirectional Long Short-Term Memory network” (CNN-AbiLSTM), which combines a CNN with an attention-based biLSTM. Second, building a powerful hybrid CNN-AbiLSTM model can be highly complicated and requires a suitable setting of the model’s hyperparameters. To handle this problem, we propose an evolutionary algorithm-based framework more specifically a Differential Evolution (DE) algorithm to find the optimal configuration of the proposed model. Experimental results show that the proposed DE-based CNN-AbiLSTM model achieves better performance compared with baseline methods.

Discovery of potential imaging and therapeutic targets for severe inflammation in COVID-19 patients

The Coronavirus disease 2019 (COVID-19) has been spreading worldwide with rapidly increased number of deaths. Hyperinflammation mediated by dysregulated monocyte/macrophage function is considered to be the key factor that triggers severe illness in COVID-19. However, no specific targeting molecule has been identified for detecting or treating hyperinflammation related to dysregulated macrophages in severe COVID-19. In this study, previously published single-cell RNA-sequencing data of bronchoalveolar lavage fluid cells from thirteen COVID-19 patients were analyzed with publicly available databases for surface and imageable targets. Immune cell composition according to the severity was estimated with the clustering of gene expression data. Expression levels of imaging target molecules for inflammation were evaluated in macrophage clusters from single-cell RNA-sequencing data. In addition, candidate targetable molecules enriched in severe COVID-19 associated with hyperinflammation were filtered. We found that expression of SLC2A3, which can be imaged by [18F]fluorodeoxyglucose, was higher in macrophages from severe COVID-19 patients. Furthermore, by integrating the surface target and drug-target binding databases with RNA-sequencing data of severe COVID-19, we identified candidate surface and druggable targets including CCR1 and FPR1 for drug delivery as well as molecular imaging. Our results provide a resource in the development of specific imaging and therapy for COVID-19-related hyperinflammation.

Extracellular Vesicle Nanoarchitectonics for Novel Drug Delivery Applications.

Extracellular vesicles (EVs) can transfer intercellular messages in various (patho)physiological processes and transport biomolecules to recipient cells. EVs possess the capacity to evade the immune system and remain stable over long periods, identifying them as natural carriers for drugs and biologics. However, the challenges associated with EVs isolation, heterogeneity, coexistence with homologous biomolecules, and lack of site-specific delivery, have impeded their potential. In recent years, the amalgamation of EVs with rationally engineered nanostructures has been proposed for achieving effective drug loading and site-specific delivery. With the advancement of nanotechnology and nanoarchitectonics, different nanostructures with tunable size, shapes, and surface properties can be integrated with EVs for drug loading, target binding, efficient delivery, and therapeutics. Such integration may enable improved cellular targeting and the protection of encapsulated drugs for enhanced and specific delivery to target cells. This review summarizes the recent development of nanostructure amalgamated EVs for drug delivery, therapeutics, and real-time monitoring of disease progression. With a specific focus on the exosomal cargo, diverse drug delivery system, and biomimetic nanostructures based on EVs for selective drug delivery, this review also chronicles the needs and challenges of EV-based biomimetic nanostructures and provides a future outlook on the strategies posed.

Prediction of the Drug–Target Binding Kinetics for Flexible Proteins by Comparative Binding Energy Analysis

There is growing consensus that the optimization of the kinetic parameters for drug-protein binding leads to improved drug efficacy. Therefore, computational methods have been developed to predict kinetic rates and to derive quantitative structure-kinetic relationships (QSKRs). Many of these methods are based on crystal structures of ligand-protein complexes. However, a drawback is that each ligand-protein complex is usually treated as having a single structure. Here, we present a modification of COMparative BINding Energy (COMBINE) analysis, which uses the structures of ligand-protein complexes to predict binding parameters. We introduce the option of using multiple structures to describe each ligand-protein complex in COMBINE analysis and apply this to study the effects of protein flexibility on the derivation of dissociation rate constants (koff) for inhibitors of p38 mitogen-activated protein (MAP) kinase, which has a flexible binding site. Multiple structures were obtained for each ligand-protein complex by performing docking to an ensemble of protein configurations obtained from molecular dynamics simulations. Coefficients to scale ligand-protein interaction energies determined from energy-minimized structures of ligand-protein complexes were obtained by partial least squares regression, and they allowed for the computation of koff values. The QSKR model obtained using single, energy-minimized crystal structures for each ligand-protein complex had higher predictive power than the QSKR model obtained with multiple structures from ensemble docking. However, incorporation of ligand-protein flexibility helped to highlight additional ligand-protein interactions that lead to longer residence times, such as interactions with residues Arg67 and Asp168, which are close to the ligand in many crystal structures. These results show that COMBINE analysis is a promising method to guide the design of compounds that bind to flexible proteins with improved binding kinetics.

Greater Efficacy and Potency of Obinutuzumab Compared with Rituximab in Chronic Lymphocytic Leukemia Patients Confirmed By a Semi-Mechanistic Pharmacokinetic/Pharmacodynamic Model

▪Introduction: In the pivotal CLL11 study (NCT02053610) that compared two CD20-targeting monoclonal antibodies in combination with chlorambucil (Clb) for the treatment of chronic lymphocytic leukemia (CLL), obinutuzumab (GA101; G) was shown to be superior to rituximab (R), with higher rates of complete response and prolongation of progression-free survival (PFS) in elderly CLL patients (pts) with comorbidities. However, the relationship between dose, exposure, and outcome is often complicated by the extent of tumor burden for antibody treatments for which target-mediated elimination is substantial, a case in which drug-binding to the pharmacological target influences drug exposure. Addressing this complexity necessitates the use of mechanistic mathematical models incorporating the extent of changes in pt target burden and its effect on drug disposition. We conducted a retrospective analysis to develop and evaluate a semi-mechanistic pharmacokinetic/pharmacodynamic (PK/PD) model for comparing drug exposure and clinical outcomes in CLL11.Methods: A semi-mechanistic PK/PD model (Figure 1) was developed for temporal changes in drug and lymphocyte counts in 632 pts who participated in the CLL11 study (325 pts treated with R-Clb and 307 pts with G-Clb). The model features both drug- and system-specific parameters allowing for the direct comparison of key factors that govern the critical responses observed with each antibody. Measurements of serum R concentrations were not available in CLL11. Therefore, R PK data were simulated using a prior population model describing R PK in CLL pts (Li et al. J Clin Pharmacol 2012) and actual dosing information from CLL11 pts. Drug concentrations and lymphocyte counts were analyzed with a nonlinear mixed effects approach using the Stochastic Approximation to Expectation Maximization (SAEM) algorithm in Monolix (version 4.3.3; Lixoft, Orsay, France). Covariate analysis was performed to identify major factors contributing to the inter-individual variability (IIV) in model parameters. The final PK/PD model was used in simulations of alternative dosing amounts and schedules (e.g. R given at the dose of G, G at the dose of R), and time-to-event modeling was used to assess the potential of various model-derived or pt-specific factors to predict PFS.Results: A semi-mechanistic PK/PD model structure was developed for joint modeling of both CD20-targeting agents, and all parameters were estimated with good precision (Table 1), albeit with high IIV, especially for PD parameters. Drug removal due to drug-target binding (kint) was 4.3-fold lower for G than that for R (0.0642 vs. 0.274 nM-1×day-1, p<0.01). The antitumor effect of R did not correlate with this greater clearance, suggesting this antibody loss is likely due to mechanisms not associated with its cytotoxic effects. The coefficient of cell kill (kkill) for G was estimated to be 4.7-fold greater than that for R (0.851 vs. 0.182 day-1, p<0.01), whereas the half-maximal effective concentration (EC50) of G was 15-fold lower (64.5 vs. 962 nM, p<0.01), indicating a significantly greater efficacy and potency compared with R based on the reduction of circulating lymphocytes. Body surface area was identified as a covariate for the IIV of the central drug volume of distribution. Model-based simulations suggest that G exhibits superior efficacy, regardless of alternative dosing schedules, as compared with R. Finally, time-to-progression analysis revealed that model-predicted antibody concentration and the percent reduction from baseline in minimal residual disease (MRD) were predictors of PFS in CLL pts.Conclusion: A comparative population-based semi-mechanistic PK/PD model was developed for analyzing exposure-outcome relationships for R and G in CLL11. Parameters associated with antibody elimination from the system indicated a faster rate of target-mediated clearance for R, whereas greater efficacy and potency of G were identified from differences in drug-specific PD parameters. A time-to-event model suggested that serum drug concentrations and percent reduction from baseline in MRD are predictors of PFS in CLL pts. [Display omitted] DisclosuresKamisoglu:Roche: Other: CLL11 is sponsored by F. Hoffmann-La Roche Ltd in collaboration with the GCLLSG and Genentech, Inc. Third-party editorial support, under the direction of Dr Kamisoglu, was provided by Scott Malkin of Gardiner-Caldwell Communications, and funded by Roche. Phipps:Roche: Equity Ownership. Jamois:Roche: Employment, Other: GALLIUM was sponsored by F. Hoffmann-La Roche Ltd.. Buchheit:Roche: Employment. Meneses-Lorente:Roche Products Ltd: Employment. Fingerle-Rowson:F. Hoffmann-La Roche Ltd: Employment, Equity Ownership. Wagg:F. Hoffmann-La Roche Ltd: Employment. Mager:Enhanced Pharmacodynamics, LLC: Employment, Equity Ownership; Roche: Research Funding.

Deep drug-target binding affinity prediction with multiple attention blocks

Drug-target interaction (DTI) prediction has drawn increasing interest due to its substantial position in the drug discovery process. Many studies have introduced computational models to treat DTI prediction as a regression task, which directly predict the binding affinity of drug-target pairs. However, existing studies (i) ignore the essential correlations between atoms when encoding drug compounds and (ii) model the interaction of drug-target pairs simply by concatenation. Based on those observations, in this study, we propose an end-to-end model with multiple attention blocks to predict the binding affinity scores of drug-target pairs. Our proposed model offers the abilities to (i) encode the correlations between atoms by a relation-aware self-attention block and (ii) model the interaction of drug representations and target representations by the multi-head attention block. Experimental results of DTI prediction on two benchmark datasets show our approach outperforms existing methods, which are benefit from the correlation information encoded by the relation-aware self-attention block and the interaction information extracted by the multi-head attention block. Moreover, we conduct the experiments on the effects of max relative position length and find out the best max relative position length value n}{}k in {3, 5}. Furthermore, we apply our model to predict the binding affinity of Corona Virus Disease 2019 (COVID-19)-related genome sequences and n}{}3137 FDA-approved drugs.

Study on drug-target binding kinetics profiles of flavonoids in Chrysanthemum morifolium and xanthine oxidase

Based on the target occupancy mathematical model, the binding kinetic process of potential active ingredients of lowering uric acid in Chrysanthemum morifolium with xanthine oxidase(XOD) was evaluated. The potential active ingredients of lowering uric acid in Ch. morifolium were screened by UPLC-Q-Exactivems MS technology, reference substance identification and in vitro enzymatic kinetics experiments. The binding kinetic parameters of xanthine oxidase and potential inhibitor in Ch. morifolium were determined by surface plasma resonance(SPR). The verified mathematical model of the XOD target occupancy evaluated the kinetic binding process of inhibitors and xanthine oxidase in vivo. According to UPLC-Q-Exactive MS and reference substance identification, 39 potential uric acid-lowering active ingredients in Ch. morifolium extracts were identified and the inhibitory activities of 23 compounds were determined. Three potential xanthine oxidase inhibitors were screened, namely genistein, luteolin, and apigenin. whose IC_(50 )were 1.23, 1.47 and 1.59 μmol·L~(-1), respectively. And the binding rate constants(K_(on)) were 1.26×10~6, 5.23×10~5 and 6.36×10~5 mol·L~(-1)·s~(-1), respectively. The dissociation rate constants(K_(off)) were 10.93×10~(-2), 1.59×10~(-2), and 5.3×10~(-2 )s~(-1), respectively. After evaluation by different administration methods, the three selected compounds can perform rapid and sustained inhibition of xanthine oxidase in vivo under combined administration. This study comprehensively evaluated the target occupancy process of three effective components in different ways of administration in vivo by UPLC-MS, concentration-response method, SPR technology and xanthine oxidase target occupancy model, which would provide a new research idea and method for screening active ingredients in traditional Chinese medicine.