Multiple Biological Networks Research Articles

Drug-target interaction prediction with collaborative contrastive learning and adaptive self-paced sampling strategy

BackgroundDrug-target interaction (DTI) prediction plays a pivotal role in drug discovery and drug repositioning, enabling the identification of potential drug candidates. However, most previous approaches often do not fully utilize the complementary relationships among multiple biological networks, which limits their ability to learn more consistent representations. Additionally, the selection strategy of negative samples significantly affects the performance of contrastive learning methods.ResultsIn this study, we propose CCL-ASPS, a novel deep learning model that incorporates Collaborative Contrastive Learning (CCL) and Adaptive Self-Paced Sampling strategy (ASPS) for drug-target interaction prediction. CCL-ASPS leverages multiple networks to learn the fused embeddings of drugs and targets, ensuring their consistent representations from individual networks. Furthermore, ASPS dynamically selects more informative negative sample pairs for contrastive learning. Experiment results on the established dataset demonstrate that CCL-ASPS achieves significant improvements compared to current state-of-the-art methods. Moreover, ablation experiments confirm the contributions of the proposed CCL and ASPS strategies.ConclusionsBy integrating Collaborative Contrastive Learning and Adaptive Self-Paced Sampling, the proposed CCL-ASPS effectively addresses the limitations of previous methods. This study demonstrates that CCL-ASPS achieves notable improvements in DTI predictive performance compared to current state-of-the-art approaches. The case study and cold start experiments further illustrate the capability of CCL-ASPS to effectively predict previously unknown DTI, potentially facilitating the identification of new drug-target interactions.

A genome-scale deep learning model to predict gene expression changes of genetic perturbations from multiplex biological networks.

Systematic characterization of biological effects to genetic perturbation is essential to the application of molecular biology and biomedicine. However, the experimental exhaustion of genetic perturbations on the genome-wide scale is challenging. Here, we show TranscriptionNet, a deep learning model that integrates multiple biological networks to systematically predict transcriptional profiles to three types of genetic perturbations based on transcriptional profiles induced by genetic perturbations in the L1000 project: RNA interference, clustered regularly interspaced short palindromic repeat, and overexpression. TranscriptionNet performs better than existing approaches in predicting inducible gene expression changes for all three types of genetic perturbations. TranscriptionNet can predict transcriptional profiles for all genes in existing biological networks and increases perturbational gene expression changes for each type of genetic perturbation from a few thousand to 26945 genes. TranscriptionNet demonstrates strong generalization ability when comparing predicted and true gene expression changes on different external tasks. Overall, TranscriptionNet can systemically predict transcriptional consequences induced by perturbing genes on a genome-wide scale and thus holds promise to systemically detect gene function and enhance drug development and target discovery.

SAMNA: accurate alignment of multiple biological networks based on simulated annealing.

Proteins are important parts of the biological structures and encode a lot of biological information. Protein-protein interaction network alignment is a model for analyzing proteins that helps discover conserved functions between organisms and predict unknown functions. In particular, multi-network alignment aims at finding the mapping relationship among multiple network nodes, so as to transfer the knowledge across species. However, with the increasing complexity of PPI networks, how to perform network alignment more accurately and efficiently is a new challenge. This paper proposes a new global network alignment algorithm called Simulated Annealing Multiple Network Alignment (SAMNA), using both network topology and sequence homology information. To generate the alignment, SAMNA first generates cross-network candidate clusters by a clustering algorithm on a k-partite similarity graph constructed with sequence similarity information, and then selects candidate cluster nodes as alignment results and optimizes them using an improved simulated annealing algorithm. Finally, the SAMNA algorithm was experimented on synthetic and real-world network datasets, and the results showed that SAMNA outperformed the state-of-the-art algorithm in biological performance.

Explainable Multilayer Graph Neural Network for Cancer Gene Prediction.

The identification of cancer genes is a critical yet challenging problem in cancer genomics research. Existing computational methods, including deep graph neural networks, fail to exploit the multilayered gene-gene interactions or provide limited explanations for their predictions. These methods are restricted to a single biological network, which cannot capture the full complexity of tumorigenesis. Models trained on different biological networks often yield different and even opposite cancer gene predictions, hindering their trustworthy adaptation. Here, we introduce an Explainable Multilayer Graph Neural Network (EMGNN) approach to identify cancer genes by leveraging multiple gene-gene interaction networks and pan-cancer multi-omics data. Unlike conventional graph learning on a single biological network, EMGNN uses a multilayered graph neural network to learn from multiple biological networks for accurate cancer gene prediction. Our method consistently outperforms all existing methods, with an average 7.15% improvement in area under the precision-recall curve (AUPR) over the current state-of-the-art method. Importantly, EMGNN integrated multiple graphs to prioritize newly predicted cancer genes with conflicting predictions from single biological networks. For each prediction, EMGNN provided valuable biological insights via both model-level feature importance explanations and molecular-level gene set enrichment analysis. Overall, EMGNN offers a powerful new paradigm of graph learning through modeling the multilayered topological gene relationships and provides a valuable tool for cancer genomics research. Our code is publicly available at https://github.com/zhanglab-aim/EMGNN.

Denoising Autoencoder based Long non-coding RNA-Disease Association Prediction

Long non-coding RNAs (lncRNAs) are recent listing in RNA Bioinformatics, which is getting more popular due to their important functional roles. According to the available research, lncRNAs play an essential role in multiple complex diseases. Determining the function of lncRNAs in diseases will help to comprehend many missing links in the disease mechanism. Predicting lncRNAdisease association (LDA) is a crucial stage in this process which is getting at most research interest nowadays. The developments in machine learning and deep learning technologies influenced recent research on LDA models. Most of the methods analyse the interactions of lncRNA with other molecules such as microRNA (miRNA), messenger RNA(mRNA), and proteins. Deep learning models, specifically from autoencoder classes, used extensively in unsupervised learning of features from these associations. This research paper proposes a denoising autoencoder (DAE) based LDA prediction approach. The proposed model uses DAE to learn lncRNA-disease representations from multiple biological networks such as lncRNA-miRNA, miRNA-disease, and disease-lncRNA interactions. The experiments show that the model outperforms other state-of-the-art LDA models concerning the area under the ROC curve (AUC-ROC, 0.94) and the area under precision-recall (AUPR, 0.9592).

The roles of eicosanoids in myocardial diseases.

Myocardial disease, the abnormalities of the cardiac muscle, is the leading cause of death in humans. Eicosanoids represent a large spectrum of lipid mediators with critical roles in physiological and pathophysiological conditions. Arachidonic acid (AA) is the major resource of eicosanoids and is metabolized via cyclooxygenases (COXs), lipoxygenases (LOXs), and cytochrome P450 (CYP) enzymes producing a diverse family of lipid mediators called eicosanoids, including prostanoids, leukotrienes (LTs), epoxyeicosatrienoic acids (EETs), dihydroxyeicosatetraenoic acid (diHETEs), eicosatetraenoic acids (ETEs), and lipoxins (LXs). Beyond the well-established roles of eicosanoids in inflammation and vascular biology, a growing body of evidence showed that eicosanoids, especially CYP450 derived eicosanoids EETs, are preventive and therapeutic targets for many of the myocardial diseases. EETs not only ameliorate the cardiac injury and remodeling in different pathological models, but also attenuate subsequent hemodynamic disturbances and cardiac dysfunction. EETs have direct and indirect protective properties in the myocardium, and thus relieve dietetic cardiomyopathy and inflammatory cardiomyopathy. Moreover, EETs are capable to attenuate the ischemic cardiomyopathy, including the myocardial infarction and cardiac ischemic reperfusion injury. Multiple biological events and signaling networks are targeted during the myocardial protection of EETs, these are including mitochondria hemostasis, angiogenesis, oxidative stress, inflammatory response, metabolic regulation, endoplasmic reticulum (ER) stress and cell death. Additionally, eicosanoids from COX and LOX also have important roles in some of the myocardial diseases, such as cardiac hypertrophy and ischemic heart disease. This chapter summarizes the physiological and pathophysiological significance, and the signal mechanisms of the eicosanoids, especially the EETs, in myocardial diseases.

The E3 ligase subunit FBXO45 binds the interferon-λ receptor and promotes its degradation during influenza virus infection

Influenza remains a major public health challenge, as the viral infection activates multiple biological networks linked to altered host innate immunity. Following infection, IFN-λ, a ligand crucial for the resolution of viral infections, is known to bind to its cognate receptor, IFNLR1, in lung epithelia. However, little is known regarding the molecular expression and regulation of IFNLR1. Here, we show that IFNLR1 is a labile protein in human airway epithelia that is rapidly degraded after influenza infection. Using an unbiased proximal ligation biotin screen, we first identified that the Skp-Cullin-F box E3 ligase subunit, FBXO45, binds to IFNLR1. We demonstrate that FBXO45, induced in response to influenza infection, mediates IFNLR1 protein polyubiquitination and degradation through the ubiquitin-proteasome system by docking with its intracellular receptor domain. Furthermore, we found ectopically expressed FBXO45 and its silencing in cells differentially regulated both IFNLR1 protein stability and interferon-stimulated gene expression. Mutagenesis studies also indicated that expression of a K319R/K320R IFNLR1 variant in cells exhibited reduced polyubiquitination, yet greater stability and proteolytic resistance to FBXO45 and influenza-mediated receptor degradation. These results indicate that the IFN-λ-IFNLR1 receptor axis is tightly regulated by the Skp-Cullin-F box ubiquitin machinery, a pathway that may be exploited by influenza infection as a means to limit antiviral responses.

Structural measures of similarity and complementarity in complex networks

The principle of similarity, or homophily, is often used to explain patterns observed in complex networks such as transitivity and the abundance of triangles (3-cycles). However, many phenomena from division of labor to protein-protein interactions (PPI) are driven by complementarity (differences and synergy). Here we show that the principle of complementarity is linked to the abundance of quadrangles (4-cycles) and dense bipartite-like subgraphs. We link both principles to their characteristic motifs and introduce two families of coefficients of: (1) structural similarity, which generalize local clustering and closure coefficients and capture the full spectrum of similarity-driven structures; (2) structural complementarity, defined analogously but based on quadrangles instead of triangles. Using multiple social and biological networks, we demonstrate that the coefficients capture structural properties related to meaningful domain-specific phenomena. We show that they allow distinguishing between different kinds of social relations as well as measuring an increasing structural diversity of PPI networks across the tree of life. Our results indicate that some types of relations are better explained by complementarity than homophily, and may be useful for improving existing link prediction methods. We also introduce a Python package implementing efficient algorithms for calculating the proposed coefficients.

MGEGFP: a multi-view graph embedding method for gene function prediction based on adaptive estimation with GCN.

In recent years, a number of computational approaches have been proposed to effectively integrate multiple heterogeneous biological networks, and have shown impressive performance for inferring gene function. However, the previous methods do not fully represent the critical neighborhood relationship between genes during the feature learning process. Furthermore, it is difficult to accurately estimate the contributions of different views for multi-view integration. In this paper, we propose MGEGFP, a multi-view graph embedding method based on adaptive estimation with Graph Convolutional Network (GCN), to learn high-quality gene representations among multiple interaction networks for function prediction. First, we design a dual-channel GCN encoder to disentangle the view-specific information and the consensus pattern across diverse networks. By the aid of disentangled representations, we develop a multi-gate module to adaptively estimate the contributions of different views during each reconstruction process and make full use of the multiplexity advantages, where a diversity preservation constraint is designed to prevent the over-fitting problem. To validate the effectiveness of our model, we conduct experiments on networks from the STRING database for both yeast and human datasets, and compare the performance with seven state-of-the-art methods in five evaluation metrics. Moreover, the ablation study manifests the important contribution of the designed dual-channel encoder, multi-gate module and the diversity preservation constraint in MGEGFP. The experimental results confirm the superiority of our proposed method and suggest that MGEGFP can be a useful tool for gene function prediction.

A comprehensive review of global alignment of multiple biological networks: background, applications and open issues

A comprehensive review of global alignment of multiple biological networks: background, applications and open issues

Learning representation for multiple biological networks via a robust graph regularized integration approach.

Learning node representation is a fundamental problem in biological network analysis, as compact representation features reveal complicated network structures and carry useful information for downstream tasks such as link prediction and node classification. Recently, multiple networks that profile objects from different aspects are increasingly accumulated, providing the opportunity to learn objects from multiple perspectives. However, the complex common and specific information across different networks pose challenges to node representation methods. Moreover, ubiquitous noise in networks calls for more robust representation. To deal with these problems, we present a representation learning method for multiple biological networks. First, we accommodate the noise and spurious edges in networks using denoised diffusion, providing robust connectivity structures for the subsequent representation learning. Then, we introduce a graph regularized integration model to combine refined networks and compute common representation features. By using the regularized decomposition technique, the proposed model can effectively preserve the common structural property of different networks and simultaneously accommodate their specific information, leading to a consistent representation. A simulation study shows the superiority of the proposed method on different levels of noisy networks. Three network-based inference tasks, including drug-target interaction prediction, gene function identification and fine-grained species categorization, are conducted using representation features learned from our method. Biological networks at different scales and levels of sparsity are involved. Experimental results on real-world data show that the proposed method has robust performance compared with alternatives. Overall, by eliminating noise and integrating effectively, the proposed method is able to learn useful representations from multiple biological networks.

A genome-scale metabolic model of Saccharomyces cerevisiae that integrates expression constraints and reaction thermodynamics

Eukaryotic organisms play an important role in industrial biotechnology, from the production of fuels and commodity chemicals to therapeutic proteins. To optimize these industrial systems, a mathematical approach can be used to integrate the description of multiple biological networks into a single model for cell analysis and engineering. One of the most accurate models of biological systems include Expression and Thermodynamics FLux (ETFL), which efficiently integrates RNA and protein synthesis with traditional genome-scale metabolic models. However, ETFL is so far only applicable for E. coli. To adapt this model for Saccharomyces cerevisiae, we developed yETFL, in which we augmented the original formulation with additional considerations for biomass composition, the compartmentalized cellular expression system, and the energetic costs of biological processes. We demonstrated the ability of yETFL to predict maximum growth rate, essential genes, and the phenotype of overflow metabolism. We envision that the presented formulation can be extended to a wide range of eukaryotic organisms to the benefit of academic and industrial research.

Comprehensive transcriptome analysis of erythroid differentiation potential of olive leaf in haematopoietic stem cells.

Anaemia is one of the leading causes of disability in young adults and is associated with increased morbidity and mortality in elderly. With a global target to reduce the disease burden of anaemia, recent researches focus on novel compounds with the ability to induce erythropoiesis and regulate iron homeostasis. We aimed to explore the biological events and potential polypharmacological effects of water‐extracted olive leaf (WOL) on human bone marrow–derived haematopoietic stem cells (hHSCs) using a comprehensive gene expression analysis. HPLC analysis identifies six bioactive polyphenols in the WOL. Treatment with WOL for 12 days regulated gene expressions related to erythroid differentiation, oxygen homeostasis, iron homeostasis, haem metabolism and Hb biosynthesis in hHSCs. Functional clustering analysis reveals several major functions of WOL such as ribosomal biogenesis and mitochondrial translation machinery, glycolytic process, ATP biosynthesis and immune response. Additionally, the colonies of both primitive and mature erythroid progenitors, CFU‐E and BFU‐E, were significantly increased in WOL‐treated hHSCs. The expressions of erythroid markers, CD47, glycophorin A (GYPA), and transferrin receptor (TFRC) and adult Hb subunits‐HBA and HBB were also confirmed in immunofluorescent staining and flow cytometer analysis in WOL‐treated hHSCs. It is well known that induction of lineage‐specific differentiation, as well as the maturation of early haematopoietic precursors into fully mature erythrocytes, involves multiple simultaneous biological events and complex signalling networks. In this regard, our genome‐wide transcriptome profiling with microarray study on WOL‐treated hHSCs provides general insights into the multitarget prophylactic and/or therapeutic potential of WOL in anaemia and other haematological disorders.

Fusing Multiple Biological Networks to Effectively Predict miRNA-disease Associations

Background: MicroRNAs (miRNAs) are a class of endogenous non-coding RNAs with about 22 nucleotides, and they play a significant role in a variety of complex biological processes. Many researches have shown that miRNAs are closely related to human diseases. Although the biological experiments are reliable in identifying miRNA-disease associations, they are timeconsuming and costly. Objective: Thus, computational methods are urgently needed to effectively predict miRNA-disease associations. Methods: In this paper, we proposed a novel method, BIRWMDA, based on a bi-random walk model to predict miRNA-disease associations. Specifically, in BIRWMDA, the similarity network fusion algorithm is used to combine the multiple similarity matrices to obtain a miRNA-miRNA similarity matrix and a disease-disease similarity matrix, then the miRNA-disease associations were predicted by the bi-random walk model. Results: To evaluate the performance of BIRWMDA, we ran the leave-one-out cross-validation and 5-fold cross-validation, and their corresponding AUCs were 0.9303 and 0.9223 ± 0.00067, respectively. To further demonstrate the effectiveness of the BIRWMDA, from the perspective of exploring disease-related miRNAs, we conducted three case studies of breast neoplasms, prostate neoplasms and gastric neoplasms, where 48, 50 and 50 out of the top 50 predicted miRNAs were confirmed by literature, respectively. From the perspective of exploring miRNA-related diseases, we conducted two case studies of hsa-mir-21 and hsa-mir-155, where 7 and 5 out of the top 10 predicted diseases were confirmed by literatures, respectively. Conclusion: The fusion of multiple biological networks could effectively predict miRNA-diseases associations. We expected BIRWMDA to serve as a biological tool for mining potential miRNAdisease associations.

Fusing Multiple Biological Networks to Effectively Predict miRNA-disease Associations

Bentham Science - STM publisher of online and print journals, and related print/online book series. Bentham Science answers the information needs of scientists in the fields of pharmaceutical, biomedical, medical, engineering, technology, computer and social sciences.

Beyond Lipid Signaling: Pleiotropic Effects of Diacylglycerol Kinases in Cellular Signaling.

The diacylglycerol kinase family, which can attenuate diacylglycerol signaling and activate phosphatidic acid signaling, regulates various signaling transductions in the mammalian cells. Studies on the regulation of diacylglycerol and phosphatidic acid levels by various enzymes, the identification and characterization of various diacylglycerol and phosphatidic acid-regulated proteins, and the overlap of different diacylglycerol and phosphatidic acid metabolic and signaling processes have revealed the complex and non-redundant roles of diacylglycerol kinases in regulating multiple biochemical and biological networks. In this review article, we summarized recent progress in the complex and non-redundant roles of diacylglycerol kinases, which is expected to aid in restoring dysregulated biochemical and biological networks in various pathological conditions at the bed side.

NetConfer: a web application for comparative analysis of multiple biological networks

BackgroundMost biological experiments are inherently designed to compare changes or transitions of state between conditions of interest. The advancements in data intensive research have in particular elevated the need for resources and tools enabling comparative analysis of biological data. The complexity of biological systems and the interactions of their various components, such as genes, proteins, taxa, and metabolites, have been inferred, represented, and visualized via graph theory-based networks. Comparisons of multiple networks can help in identifying variations across different biological systems, thereby providing additional insights. However, while a number of online and stand-alone tools exist for generating, analyzing, and visualizing individual biological networks, the utility to batch process and comprehensively compare multiple networks is limited.ResultsHere, we present a graphical user interface (GUI)-based web application which implements multiple network comparison methodologies and presents them in the form of organized analysis workflows. Dedicated comparative visualization modules are provided to the end-users for obtaining easy to comprehend, insightful, and meaningful comparisons of various biological networks. We demonstrate the utility and power of our tool using publicly available microbial and gene expression data.ConclusionNetConfer tool is developed keeping in mind the requirements of researchers working in the field of biological data analysis with limited programming expertise. It is also expected to be useful for advanced users from biological as well as other domains (working with association networks), benefiting from provided ready-made workflows, as they allow to focus directly on the results without worrying about the implementation. While the web version allows using this application without installation and dependency requirements, a stand-alone version has also been supplemented to accommodate the offline requirement of processing large networks.

CLING: Candidate Cancer-Related lncRNA Prioritization via Integrating Multiple Biological Networks.

Identification and characterization of lncRNAs in cancer with a view to their application in improving diagnosis and therapy remains a major challenge that requires new and innovative approaches. We have developed an integrative framework termed “CLING”, aimed to prioritize candidate cancer-related lncRNAs based on their associations with known cancer lncRNAs. CLING focuses on joint optimization and prioritization of all candidates for each cancer type by integrating lncRNA topological properties and multiple lncRNA-centric networks. Validation analyses revealed that CLING is more effective than prioritization based on a single lncRNA network. Reliable AUC (Area Under Curve) scores were obtained across 10 cancer types, ranging from 0.85 to 0.94. Several novel lncRNAs predicted in the top 10 candidates for various cancer types have been confirmed by recent biological experiments. Furthermore, using a case study on liver hepatocellular carcinoma as an example, CLING facilitated the successful identification of novel cancer lncRNAs overlooked by differential expression analyses (DEA). This time- and cost-effective computational model may provide a valuable complement to experimental studies and assist in future investigations on lncRNA involvement in the pathogenesis of cancers. We have developed a web-based server for users to rapidly implement CLING and visualize data, which is freely accessible at http://bio-bigdata.hrbmu.edu.cn/cling/. CLING has been successfully applied to predict a few potential lncRNAs from thousands of candidates for many cancer types.

A systems approach to clinical oncology uses deep phenotyping to deliver personalized care.

Cancer encompasses a complex, heterogeneous and dynamic group of diseases that arise from perturbations to multiple biological networks within the body. A systems biology-based approach would help to decipher this complexity, to deeply characterize the pathophysiology of the disease and to stratify cancers into appropriate molecular subtypes to facilitate the development of personalized therapies. Technological advances made over the past decade have enabled multiscale, longitudinal measurements ('snapshots') of human biology, from single-cell analyses to whole-body monitoring. In this Perspective, we discuss some of these technologies and how they have (and will) contributed to our understanding of cancer biology as well as to the development of early diagnostics and personalized therapies. We argue that the integration of molecular profiling of cancerous tissues with deep, longitudinal profiling of the physiological state of an individual ('deep phenotyping') is key to understanding the prevention, initiation, progression and response to treatment of cancers. Systems biology-based approaches can provide an unprecedented trove of data for early detection of disease transitions, prediction of therapeutic responses and clinical outcomes, and for the design of personalized treatments.

Corrections to “A Novel Computational Approach for Global Alignment for Multiple Biological Networks”

Presents corrections to the paper, A novel computational approach for global alignment for multiple biological networks,” (Djeddi, W.E., et al), Trans. Comput. Biol. Bioinf., vol. 15, no. 6, pp. 2060–2066, Nov./Dec. 2018.