Analysis of multiple compound–protein interactions reveals novel bioactive molecules

The discovery of novel bioactive molecules advances our systems‐level understanding of biological processes and is crucial for innovation in drug development. For this purpose, the emerging field of chemical genomics is currently focused on accumulating large assay data sets describing compound–protein interactions (CPIs). Although new target proteins for known drugs have recently been identified through mining of CPI databases, using these resources to identify novel ligands remains unexplored. Herein, we demonstrate that machine learning of multiple CPIs can not only assess drug polypharmacology but can also efficiently identify novel bioactive scaffold‐hopping compounds. Through a machine‐learning technique that uses multiple CPIs, we have successfully identified novel lead compounds for two pharmaceutically important protein families, G‐protein‐coupled receptors and protein kinases. These novel compounds were not identified by existing computational ligand‐screening methods in comparative studies. The results of this study indicate that data derived from chemical genomics can be highly useful for exploring chemical space, and this systems biology perspective could accelerate drug discovery processes.     

Recent advances in features generation for membrane protein sequences: From multiple sequence alignment to pre-trained language models

Membrane proteins play a crucial role in various cellular processes and are essential components of cell membranes. Computational methods have emerged as a powerful tool for studying membrane proteins due to their complex structures and properties that make them difficult to analyze experimentally. Traditional features for protein sequence analysis based on amino acid types, composition, and pair composition have limitations in capturing higher-order sequence patterns. Recently, multiple sequence alignment (MSA) and pre-trained language models (PLMs) have been used to generate features from protein sequences. However, the significant computational resources required for MSA-based features generation can be a major bottleneck for many applications. Several methods and tools have been developed to accelerate the generation of MSAs and reduce their computational cost, including heuristics and approximate algorithms. Additionally, the use of PLMs such as BERT has shown great potential in generating informative embeddings for protein sequence analysis. In this review, we provide an overview of traditional and more recent methods for generating features from protein sequences, with a particular focus on MSAs and PLMs. We highlight the advantages and limitations of these approaches and discuss the methods and tools developed to address the computational challenges associated with features generation. Overall, the advancements in computational methods and tools provide a promising avenue for gaining deeper insights into the function and properties of membrane proteins, which can have significant implications in drug discovery and personalized medicine.     

Discovery of Novel Liver-Stage Antimalarials through Quantum Similarity

Without quantum theory any understanding of molecular interactions is incomplete. In principal, chemistry, and even biology, can be fully derived from non-relativistic quantum mechanics. In practice, conventional quantum chemical calculations are computationally too intensive and time consuming to be useful for drug discovery on more than a limited basis. A previously described, original, quantum-based computational process for drug discovery and design bridges this gap between theory and practice, and allows the application of quantum methods to large-scale in silico identification of active compounds. Here, we show the results of this quantum-similarity approach applied to the discovery of novel liver-stage antimalarials. Testing of only five of the model-predicted compounds in vitro and in vivo hepatic stage drug inhibition assays with P. berghei identified four novel chemical structures representing three separate quantum classes of liver-stage antimalarials. All four compounds inhibited liver-stage Plasmodium as a single oral dose in the quantitative PCR mouse liver-stage sporozoites-challenge model. One of the newly identified compounds, cethromycin [ABT-773], a macrolide-quinoline hybrid, is a drug with an extensive (over 5,000 people) safety profile warranting its exploitation as a new weapon for the current effort of malaria eradication. The results of our molecular modeling exceed current state-of-the-art computational methods. Drug discovery through quantum similarity is data-driven, agnostic to any particular target or disease process that can evaluate multiple phenotypic, target-specific, or co-crystal structural data. This allows the incorporation of additional pharmacological requirements, as well as rapid exploration of novel chemical spaces for therapeutic applications.     

Emerging frontiers in virtual drug discovery: From quantum mechanical methods to deep learning approaches

Virtual screening-based approaches to discover initial hit and lead compounds have the potential to reduce both the cost and time of early drug discovery stages, as well as to find inhibitors for even challenging target sites such as protein–protein interfaces. Here in this review, we provide an overview of the progress that has been made in virtual screening methodology and technology on multiple fronts in recent years. The advent of ultra-large virtual screens, in which hundreds of millions to billions of compounds are screened, has proven to be a powerful approach to discover highly potent hit compounds. However, these developments are just the tip of the iceberg, with new technologies and methods emerging to propel the field forward. Examples include novel machine-learning approaches, which can reduce the computational costs of virtual screening dramatically, while progress in quantum-mechanical approaches can increase the accuracy of predictions of various small molecule properties.     

Tackling the conformational sampling of larger flexible compounds and macrocycles in pharmacology and drug discovery

Computational conformational sampling underpins much of molecular modeling and design in pharmaceutical work. The sampling of smaller drug-like compounds has been an active area of research. However, few studies have tested in details the sampling of larger more flexible compounds, which are also relevant to drug discovery, including therapeutic peptides, macrocycles, and inhibitors of protein–protein interactions. Here, we investigate extensively mainstream conformational sampling methods on three carefully curated compound sets, namely the ‘Drug-like’, larger ‘Flexible’, and ‘Macrocycle’ compounds. These test molecules are chemically diverse with reliable X-ray protein-bound bioactive structures. The compared sampling methods include Stochastic Search and the recent LowModeMD from MOE, all the low-mode based approaches from MacroModel, and MD/LLMOD recently developed for macrocycles. In addition to default settings, key parameters of the sampling protocols were explored. The performance of the computational protocols was assessed via (i) the reproduction of the X-ray bioactive structures, (ii) the size, coverage and diversity of the output conformational ensembles, (iii) the compactness/extendedness of the conformers, and (iv) the ability to locate the global energy minimum. The influence of the stochastic nature of the searches on the results was also examined. Much better results were obtained by adopting search parameters enhanced over the default settings, while maintaining computational tractability. In MOE, the recent LowModeMD emerged as the method of choice. Mixed torsional/low-mode from MacroModel performed as well as LowModeMD, and MD/LLMOD performed well for macrocycles. The low-mode based approaches yielded very encouraging results with the flexible and macrocycle sets. Thus, one can productively tackle the computational conformational search of larger flexible compounds for drug discovery, including macrocycles.     

Mass spectrometry-based Cellular Thermal Shift Assay (CETSA®) for target deconvolution in phenotypic drug discovery

The recent renewed interest in phenotypic drug discovery has concomitantly put a focus on target deconvolution in order to achieve drug-target identification. Even though there are prescribed therapies whose mode of action is not fully understood, knowledge of the primary target will inevitably facilitate the discovery and translation of efficacy from bench to bedside. Elucidating targets and subsequent pathways engaged will also facilitate safety studies and overall development of novel drug candidates. Today, there are several techniques available for identifying the primary target, many of which rely on mass spectrometry (MS) to identify compound – target protein interactions. The Cellular Thermal Shift Assay (CETSA®) is well suited for identifying target engagement between ligands and their protein targets. Several studies have shown that CETSA combined with MS is a powerful technique that allows unlabeled target deconvolution in complex samples such as intact cells and tissues in addition to cell lysates and other protein suspensions. The applicability of CETSA MS for target deconvolution purposes will be discussed and exemplified in this mini review.     

Knowledge-based Fragment Binding Prediction

Target-based drug discovery must assess many drug-like compounds for potential activity. Focusing on low-molecular-weight compounds (fragments) can dramatically reduce the chemical search space. However, approaches for determining protein-fragment interactions have limitations. Experimental assays are time-consuming, expensive, and not always applicable. At the same time, computational approaches using physics-based methods have limited accuracy. With increasing high-resolution structural data for protein-ligand complexes, there is now an opportunity for data-driven approaches to fragment binding prediction. We present FragFEATURE, a machine learning approach to predict small molecule fragments preferred by a target protein structure. We first create a knowledge base of protein structural environments annotated with the small molecule substructures they bind. These substructures have low-molecular weight and serve as a proxy for fragments. FragFEATURE then compares the structural environments within a target protein to those in the knowledge base to retrieve statistically preferred fragments. It merges information across diverse ligands with shared substructures to generate predictions. Our results demonstrate FragFEATURE''s ability to rediscover fragments corresponding to the ligand bound with 74% precision and 82% recall on average. For many protein targets, it identifies high scoring fragments that are substructures of known inhibitors. FragFEATURE thus predicts fragments that can serve as inputs to fragment-based drug design or serve as refinement criteria for creating target-specific compound libraries for experimental or computational screening.     

Identifying Novel Drug Targets by iDTPnd: A Case Study of Kinase Inhibitors

Current FDA-approved kinase inhibitors cause diverse adverse effects, some of which are due to the mechanism-independent effects of these drugs. Identifying these mechanism-independent interactions could improve drug safety and support drug repurposing. Here, we develop iDTPnd (integrated Drug Target Predictor with negative dataset), a computational approach for large-scale discovery of novel targets for known drugs. For a given drug, we construct a positive structural signature as well as a negative structural signature that captures the weakly conserved structural features of drug-binding sites. To facilitate assessment of unintended targets, iDTPnd also provides a docking-based interaction score and its statistical significance. We confirm the interactions of sorafenib, imatinib, dasatinib, sunitinib, and pazopanib with their known targets at a sensitivity of 52% and a specificity of 55%. We also validate 10 predicted novel targets by using in vitro experiments. Our results suggest that proteins other than kinases, such as nuclear receptors, cytochrome P450, and MHC class I molecules, can also be physiologically relevant targets of kinase inhibitors. Our method is general and broadly applicable for the identification of protein–small molecule interactions, when sufficient drug–target 3D data are available. The code for constructing the structural signatures is available at https://sfb.kaust.edu.sa/Documents/iDTP.zip.     

Artificial intelligence for compound pharmacokinetics prediction

Optimisation of compound pharmacokinetics (PK) is an integral part of drug discovery and development. Animal in vivo PK data as well as human and animal in vitro systems are routinely utilised to evaluate PK in humans. In recent years machine learning and artificial intelligence (AI) emerged as a major tool for modelling of in vivo animal and human PK, enabling prediction from chemical structure early in drug discovery, and therefore offering opportunities to guide the design and prioritisation of molecules based on relevant in vivo properties and, ultimately, predicting human PK at the point of design. This review presents recent advances in machine learning and AI models for in vivo animal and human PK for small-molecule compounds as well as some examples for antibody therapeutics.     

Exploring c-Met kinase flexibility by sampling and clustering its conformational space

It is now widely recognized that the flexibility of both partners has to be considered in molecular docking studies. However, the question how to handle the best the huge computational complexity of exploring the protein binding site landscape is still a matter of debate. Here we investigate the flexibility of c-Met kinase as a test case for comparing several simulation methods. The c-Met kinase catalytic site is an interesting target for anticancer drug design. In particular, it harbors an unusual plasticity compared with other kinases ATP binding sites. Exploiting this feature may eventually lead to the discovery of new anticancer agents with exquisite specificity. We present in this article an extensive investigation of c-Met kinase conformational space using large-scale computational simulations in order to extend the knowledge already gathered from available X-ray structures. In the process, we compare the relevance of different strategies for modeling and injecting receptor flexibility information into early stage in silico structure-based drug discovery pipeline. The results presented here are currently being exploited in on-going virtual screening investigations on c-Met.     

Multiplexed site-specific genome engineering for overproducing bioactive secondary metabolites in actinomycetes

Actinomycetes produce a large variety of pharmaceutically active compounds, yet production titers often require to be improved for discovery, development and large-scale manufacturing. Here, we describe a new technique, multiplexed site-specific genome engineering (MSGE) via the ‘one integrase-multiple attB sites’ concept, for the stable integration of secondary metabolite biosynthetic gene clusters (BGCs). Using MSGE, we achieved five-copy chromosomal integration of the pristinamycin II (PII) BGC in Streptomyces pristinaespiralis, resulting in the highest reported PII titers in flask and batch fermentations (2.2 and 2 g/L, respectively). Furthermore, MSGE was successfully extended to develop a panel of powerful Streptomyces coelicolor heterologous hosts, in which up to four copies of the BGCs for chloramphenicol or anti-tumour compound YM-216391 were efficiently integrated in a single step, leading to significantly elevated productivity (2–23 times). Our multiplexed approach holds great potential for robust genome engineering of industrial actinomycetes and novel drug discovery by genome mining.     

Chemogenomic Approaches for Revealing Drug Target Interactions in Drug Discovery

The drug discovery process has been a crucial and cost-intensive process. This cost is not only monetary but also involves risks, time, and labour that are incurred while introducing a drug in the market. In order to reduce this cost and the risks associated with the drugs that may result in severe side effects, the in silico methods have gained popularity in recent years. These methods have had a significant impact on not only drug discovery but also the related areas such as drug repositioning, drug-target interaction prediction, drug side effect prediction, personalised medicine, etc. Amongst these research areas predicting interactions between drugs and targets forms the basis for drug discovery. The availability of big data in the form of bioinformatics, genetic databases, along with computational methods, have further supported data-driven decision-making. The results obtained through these methods may be further validated using in vitro or in vivo experiments. This validation step can further justify the predictions resulting from in silico approaches, further increasing the accuracy of the overall result in subsequent stages. A variety of approaches are used in predicting drug-target interactions, including ligand-based, molecular docking based and chemogenomic-based approaches. This paper discusses the chemogenomic methods, considering drug target interaction as a classification problem on whether or not an interaction between a particular drug and target would serve as a basis for understanding drug discovery/drug repositioning. We present the advantages and disadvantages associated with their application.     

Feature selection and classification of protein–protein complexes based on their binding affinities using machine learning approaches

Protein–protein interactions are intrinsic to virtually every cellular process. Predicting the binding affinity of protein–protein complexes is one of the challenging problems in computational and molecular biology. In this work, we related sequence features of protein–protein complexes with their binding affinities using machine learning approaches. We set up a database of 185 protein–protein complexes for which the interacting pairs are heterodimers and their experimental binding affinities are available. On the other hand, we have developed a set of 610 features from the sequences of protein complexes and utilized Ranker search method, which is the combination of Attribute evaluator and Ranker method for selecting specific features. We have analyzed several machine learning algorithms to discriminate protein‐protein complexes into high and low affinity groups based on their K_d values. Our results showed a 10‐fold cross‐validation accuracy of 76.1% with the combination of nine features using support vector machines. Further, we observed accuracy of 83.3% on an independent test set of 30 complexes. We suggest that our method would serve as an effective tool for identifying the interacting partners in protein–protein interaction networks and human–pathogen interactions based on the strength of interactions. Proteins 2014; 82:2088–2096. © 2014 Wiley Periodicals, Inc.     

Protein–DNA/RNA interactions: Machine intelligence tools and approaches in the era of artificial intelligence and big data

With the development of artificial intelligence (AI) technologies and the availability of large amounts of biological data, computational methods for proteomics have undergone a developmental process from traditional machine learning to deep learning. This review focuses on computational approaches and tools for the prediction of protein – DNA/RNA interactions using machine intelligence techniques. We provide an overview of the development progress of computational methods and summarize the advantages and shortcomings of these methods. We further compiled applications in tasks related to the protein – DNA/RNA interactions, and pointed out possible future application trends. Moreover, biological sequence-digitizing representation strategies used in different types of computational methods are also summarized and discussed.     

Prediction of protein–protein interaction sites from weakly homologous template structures using meta‐threading and machine learning

The identification of protein–protein interactions is vital for understanding protein function, elucidating interaction mechanisms, and for practical applications in drug discovery. With the exponentially growing protein sequence data, fully automated computational methods that predict interactions between proteins are becoming essential components of system‐level function inference. A thorough analysis of protein complex structures demonstrated that binding site locations as well as the interfacial geometry are highly conserved across evolutionarily related proteins. Because the conformational space of protein–protein interactions is highly covered by experimental structures, sensitive protein threading techniques can be used to identify suitable templates for the accurate prediction of interfacial residues. Toward this goal, we developed eFindSite^PPI, an algorithm that uses the three‐dimensional structure of a target protein, evolutionarily remotely related templates and machine learning techniques to predict binding residues. Using crystal structures, the average sensitivity (specificity) of eFindSite^PPI in interfacial residue prediction is 0.46 (0.92). For weakly homologous protein models, these values only slightly decrease to 0.40–0.43 (0.91–0.92) demonstrating that eFindSite^PPI performs well not only using experimental data but also tolerates structural imperfections in computer‐generated structures. In addition, eFindSite^PPI detects specific molecular interactions at the interface; for instance, it correctly predicts approximately one half of hydrogen bonds and aromatic interactions, as well as one third of salt bridges and hydrophobic contacts. Comparative benchmarks against several dimer datasets show that eFindSite^PPI outperforms other methods for protein‐binding residue prediction. It also features a carefully tuned confidence estimation system, which is particularly useful in large‐scale applications using raw genomic data. eFindSite^PPI is freely available to the academic community at http://www.brylinski.org/efindsiteppi . Copyright © 2014 John Wiley & Sons, Ltd.     

基于深度学习与多层次信息融合的药物靶标亲和力预测*

药物研发是非常重要但也十分耗费人力物力的过程。利用计算机辅助预测药物与蛋白质亲和力的方法可以极大地加快药物研发过程。药物靶标亲和力预测的关键在于对药物和蛋白质进行准确详细地信息表征。提出一种基于深度学习与多层次信息融合的药物靶标亲和力的预测模型,试图通过综合药物与蛋白质的多层次信息,来获得更好的预测表现。首先将药物表述成分子图和扩展连接指纹两种形式,分别利用图卷积神经网络模块和全连接层进行学习;其次将蛋白质序列和蛋白质K-mer特征分别输入卷积神经网络模块和全连接层来学习蛋白质潜在特征;随后将4个通道学习到的特征进行融合,再利用全连接层进行预测。在两个基准药物靶标亲和力数据集上验证了所提方法的有效性,并与其他已有模型作对比研究。结果说明提出的模型相比基准模型能得到更好的预测性能,表明提出的综合药物与蛋白质多层次信息的药物靶标亲和力预测策略是有效的。     

AI in 3D compound design

The success of Artificial Intelligence (AI) across a wide range of domains has fuelled significant interest in its application to designing novel compounds and screening compounds against a specific target. However, many existing AI methods either do not account for the 3D structure of the target at all or struggle to capture meaningful spatial information from the target. In this Opinion, we highlight a range of recent structure-aware approaches which utilise deep learning for compound design and virtual screening. We discuss how such methods can be better integrated into existing drug discovery pipelines by facilitating the design of compounds which conform to a specified design hypothesis and by uncovering key protein-ligand interactions which can be used to aid molecule design.     

Structural insights into pharmacophore-assisted in silico identification of protein–protein interaction inhibitors for inhibition of human toll-like receptor 4 – myeloid differentiation factor-2 (hTLR4−MD-2) complex

Toll-like receptor 4 (TLR4) is a member of Toll-Like Receptors (TLRs) family that serves as a receptor for bacterial lipopolysaccharide (LPS). TLR4 alone cannot recognize LPS without aid of co-receptor myeloid differentiation factor-2 (MD-2). Binding of LPS with TLR4 forms a LPS?TLR4?MD-2 complex and directs downstream signaling for activation of immune response, inflammation and NF-κB activation. Activation of TLR4 signaling is associated with various pathophysiological consequences. Therefore, targeting protein–protein interaction (PPI) in TLR4?MD-2 complex formation could be an attractive therapeutic approach for targeting inflammatory disorders. The aim of present study was directed to identify small molecule PPI inhibitors (SMPPIIs) using pharmacophore mapping-based approach of computational drug discovery. Here, we had retrieved the information about the hot spot residues and their pharmacophoric features at both primary (TLR4?MD-2) and dimerization (MD-2?TLR4*) protein–protein interaction interfaces in TLR4?MD-2 homo-dimer complex using in silico methods. Promising candidates were identified after virtual screening, which may restrict TLR4?MD-2 protein–protein interaction. In silico off-target profiling over the virtually screened compounds revealed other possible molecular targets. Two of the virtually screened compounds (C11 and C15) were predicted to have an inhibitory concentration in μM range after HYDE assessment. Molecular dynamics simulation study performed for these two compounds in complex with target protein confirms the stability of the complex. After virtual high throughput screening we found selective hTLR4?MD-2 inhibitors, which may have therapeutic potential to target chronic inflammatory diseases.