The BioGRID interaction database: 2015 update

Decoding glycosylation potential from protein structure across human glycoproteins with a multi-view recurrent neural network

Glycosylation is described as a non-templated biosynthesis. Yet, the template-free premise is antithetical to the observation that different N-glycans are consistently placed at specific sites. It has been proposed that glycosite-proximal protein structures could constrain glycosylation and explain the observed microheterogeneity. Using site-specific glycosylation data, we trained a hybrid neural network to parse glycosites (recurrent neural network) and match them to feasible N-glycosylation events (graph neural network). From glycosite-flanking sequences, the algorithm predicts most human N-glycosylation events documented in the GlyConnect database and proposed structures corresponding to observed monosaccharide composition of the glycans at these sites. The algorithm also recapitulated glycosylation in Enhanced Aromatic Sequons, SARS-CoV-2 spike, and IgG3 variants, thus demonstrating the ability of the algorithm to predict both glycan structure and abundance. Thus, protein structure constrains glycosylation, and the neural network enables predictive in silico glycosylation of uncharacterized or novel protein sequences and genetic variants.

Yeast profilin mutants inhibit classical nuclear import and alter the balance between actin and tubulin levels

Profilin is a small protein that controls actin polymerization in yeast and higher eukaryotes. In addition, profilin has emerged as a multifunctional protein that contributes to other processes in multicellular organisms. This study focuses on profilin (Pfy1) in the budding yeast Saccharomyces cerevisiae. The primary sequences of yeast Pfy1 and its metazoan orthologs diverge vastly. However, structural elements of profilin are conserved among different species. To date, the full spectrum of Pfy1 functions has yet to be defined. The current work explores the possible involvement of yeast profilin in nuclear protein import. To this end, a panel of well-characterized yeast profilin mutants was evaluated. The experiments demonstrate that yeast profilin (i) regulates nuclear protein import, (ii) determines the subcellular localization of essential nuclear transport factors, and (iii) controls the relative abundance of actin and tubulin. Together, these results define yeast profilin as a moonlighting protein that engages in multiple essential cellular activities.

Oxidative stress and signaling through EGFR and PKA pathways converge on the nuclear transport factor RanBP1

Nuclear protein trafficking requires the soluble transport factor RanBP1. The subcellular distribution of RanBP1 is dynamic, as the protein shuttles between the nucleus and cytoplasm. To date, the signaling pathways regulating RanBP1 subcellular localization are poorly understood. During interphase, RanBP1 resides mostly in the cytoplasm. We show here that oxidative stress concentrates RanBP1 in the nucleus, and our study defines the underlying mechanisms. Specifically, RanBP1's cysteine residues are not essential for its oxidant-induced relocation. Furthermore, our pharmacological approaches uncover that signaling mediated by epidermal growth factor receptor (EGFR) and protein kinase A (PKA) control RanBP1 localization during stress. In particular, pharmacological inhibitors of EGFR or PKA diminish the oxidant-dependent relocation of RanBP1. Mutant analysis identified serine 60 and tyrosine 103 as regulators of RanBP1 nuclear accumulation during oxidant exposure. Taken together, our results define RanBP1 as a target of oxidative stress and a downstream effector of EGFR and PKA signaling routes. This positions RanBP1 at the intersection of important cellular signaling circuits.

Viruses traverse the human proteome through peptide interfaces that can be biomimetically leveraged for drug discovery

We present a drug design strategy based on structural knowledge of protein-protein interfaces selected through virus-host coevolution and translated into highly potential small molecules. This approach is grounded on Vinland, the most comprehensive atlas of virus-human protein-protein interactions with annotation of interacting domains. From this inspiration, we identified small viral protein domains responsible for interaction with human proteins. These peptides form a library of new chemical entities used to screen for replication modulators of several pathogens. As a proof of concept, a peptide from a KSHV protein, identified as an inhibitor of influenza virus replication, was translated into a small molecule series with low nanomolar antiviral activity. By targeting the NEET proteins, these molecules turn out to be of therapeutic interest in a nonalcoholic steatohepatitis mouse model with kidney lesions. This study provides a biomimetic framework to design original chemistries targeting cellular proteins, with indications going far beyond infectious diseases.

A seed expansion-based method to identify essential proteins by integrating protein-protein interaction sub-networks and multiple biological characteristics

BACKGROUND

The identification of essential proteins is of great significance in biology and pathology. However, protein-protein interaction (PPI) data obtained through high-throughput technology include a high number of false positives. To overcome this limitation, numerous computational algorithms based on biological characteristics and topological features have been proposed to identify essential proteins.

RESULTS

In this paper, we propose a novel method named SESN for identifying essential proteins. It is a seed expansion method based on PPI sub-networks and multiple biological characteristics. Firstly, SESN utilizes gene expression data to construct PPI sub-networks. Secondly, seed expansion is performed simultaneously in each sub-network, and the expansion process is based on the topological features of predicted essential proteins. Thirdly, the error correction mechanism is based on multiple biological characteristics and the entire PPI network. Finally, SESN analyzes the impact of each biological characteristic, including protein complex, gene expression data, GO annotations, and subcellular localization, and adopts the biological data with the best experimental results. The output of SESN is a set of predicted essential proteins.

CONCLUSIONS

The analysis of each component of SESN indicates the effectiveness of all components. We conduct comparison experiments using three datasets from two species, and the experimental results demonstrate that SESN achieves superior performance compared to other methods.

Exploring the interactions of antihistamine with retinoic acid receptor beta (RARB) by molecular dynamics simulations and genome-wide meta-analysis

Kaposi sarcoma (KS) is one of the most common AIDS-related malignant neoplasms, which can leave lesions on the skin among HIV patients. These lesions can be treated with 9-cis-retinoic acid (9-cis-RA), an endogenous ligand of retinoic acid receptors that has been FDA-approved for treatment of KS. However, topical application of 9-cis-RA can induce several unpleasant side effects, like headache, hyperlipidemia, and nausea. Hence, alternative therapeutics with less side effects are desirable. There are case reports associating over-the-counter antihistamine usage with regression of KS. Antihistamines competitively bind to H1 receptor and block the action of histamine, best known for being released in response to allergens. Furthermore, there are already dozens of antihistamines that are FDA-approved with less side effects than 9-cis-RA. This led our team to conduct a series of in-silico assays to determine whether antihistamines can activate retinoic acid receptors. First, we utilized high-throughput virtual screening and molecular dynamics simulations to model high-affinity interactions between antihistamines and retinoic acid receptor beta (RARβ). We then performed systems genetics analysis to identify a genetic association between H1 receptor itself and molecular pathways involved in KS. Together, these findings advocate for exploration of antihistamines against KS, starting with our two promising hit compounds, bepotastine and hydroxyzine, for experimental validation study in the future.

Valosin containing protein (VCP): initiator, modifier, and potential drug target for neurodegenerative diseases

The AAA+ ATPase valosin containing protein (VCP) is essential for cell and organ homeostasis, especially in cells of the nervous system. As part of a large network, VCP collaborates with many cofactors to ensure proteostasis under normal, stress, and disease conditions. A large number of mutations have revealed the importance of VCP for human health. In particular, VCP facilitates the dismantling of protein aggregates and the removal of dysfunctional organelles. These are critical events to prevent malfunction of the brain and other parts of the nervous system. In line with this idea, VCP mutants are linked to the onset and progression of neurodegeneration and other diseases. The intricate molecular mechanisms that connect VCP mutations to distinct brain pathologies continue to be uncovered. Emerging evidence supports the model that VCP controls cellular functions on multiple levels and in a cell type specific fashion. Accordingly, VCP mutants derail cellular homeostasis through several mechanisms that can instigate disease. Our review focuses on the association between VCP malfunction and neurodegeneration. We discuss the latest insights in the field, emphasize open questions, and speculate on the potential of VCP as a drug target for some of the most devastating forms of neurodegeneration.

A Structural Network Analysis of Neuronal ArhGAP21/23 Interactors by Computational Modeling

RhoGTPase-activating proteins (RhoGAPs) play multiple roles in neuronal development; however, details of their substrate recognition system remain elusive. ArhGAP21 and ArhGAP23 are RhoGAPs that contain N-terminal PDZ and pleckstrin homology domains. In the present study, the RhoGAP domain of these ArhGAPs was computationally modeled by template-based methods and the AlphaFold2 software program, and their intrinsic RhoGTPase recognition mechanism was analyzed from the domain structures using the protein docking programs HADDOCK and HDOCK. ArhGAP21 was predicted to preferentially catalyze Cdc42, RhoA, RhoB, RhoC, and RhoG and to downregulate RhoD and Tc10 activities. Regarding ArhGAP23, RhoA and Cdc42 were deduced to be its substrates, whereas RhoD downregulation was predicted to be less efficient. The PDZ domains of ArhGAP21/23 possess the FTLRXXXVY sequence, and similar globular folding consists of antiparalleled β-sheets and two α-helices that are conserved with PDZ domains of MAST-family proteins. A peptide docking analysis revealed the specific interaction of the ArhGAP23 PDZ domain with the PTEN C-terminus. The pleckstrin homology domain structure of ArhGAP23 was also predicted, and the functional selectivity for the interactors regulated by the folding and disordered domains in ArhGAP21 and ArhGAP23 was examined by an in silico analysis. An interaction analysis of these RhoGAPs revealed the existence of mammalian ArhGAP21/23-specific type I and type III Arf- and RhoGTPase-regulated signaling. Multiple recognition systems of RhoGTPase substrates and selective Arf-dependent localization of ArhGAP21/23 may form the basis of the functional core signaling necessary for synaptic homeostasis and axon/dendritic transport regulated by RhoGAP localization and activities.

Explainable drug side effect prediction via biologically informed graph neural network

Early detection of potential side effects (SE) is a critical and challenging task for drug discovery and patient care. In-vitro or in-vivo approach to detect potential SEs is not scalable for many drug candidates during the preclinical stage. Recent advances in explainable machine learning may facilitate detecting potential SEs of new drugs before market release and elucidating the critical mechanism of biological actions. Here, we leverage multi-modal interactions among molecules to develop a biologically informed graph-based SE prediction model, called HHAN-DSI. HHAN-DSI predicted frequent and even uncommon SEs of the unseen drug with higher or comparable accuracy against benchmark methods. When applying HHAN-DSI to the central nervous system, the organs with the largest number of SEs, the model revealed diverse psychiatric medications' previously unknown but probable SEs, together with the potential mechanisms of actions through a network of genes, biological functions, drugs, and SEs.

Machine Learning Advances in Predicting Peptide/Protein-Protein Interactions Based on Sequence Information for Lead Peptides Discovery

Peptides have shown increasing advantages and significant clinical value in drug discovery and development. With the development of high-throughput technologies and artificial intelligence (AI), machine learning (ML) methods for discovering new lead peptides have been expanded and incorporated into rational drug design. Predictions of peptide-protein interactions (PepPIs) and protein-protein interactions (PPIs) are both opportunities and challenges in computational biology, which will help to better understand the mechanisms of disease and provide the impetus for the discovery of lead peptides. This paper comprehensively reviews computational models for PepPI and PPI predictions. It begins with an introduction of various databases of peptide ligands and target proteins. Then it discusses data formats and feature representations for proteins and peptides. Furthermore, classical ML methods and emerging deep learning (DL) methods that can be used to train prediction models of PepPI and PPI are classified into four categories, and their advantages and disadvantages are analyzed. To assess the relative performance of different models, different validation protocols and evaluation indexes are discussed. The goal of this review is to help researchers quickly get started to develop computational frameworks using these integrated resources and eventually promote the discovery of lead peptides.

Drosophila DAxud1 Has a Repressive Transcription Activity on Hsp70 and Other Heat Shock Genes

Drosophila melanogaster DAxud1 is a transcription factor that belongs to the Cysteine Serine Rich Nuclear Protein (CSRNP) family, conserved in metazoans, with a transcriptional transactivation activity. According to previous studies, this protein promotes apoptosis and Wnt signaling-mediated neural crest differentiation in vertebrates. However, no analysis has been conducted to determine what other genes it might control, especially in connection with cell survival and apoptosis. To partly answer this question, this work analyzes the role of Drosophila DAxud1 using Targeted-DamID-seq (TaDa-seq), which allows whole genome screening to determine in which regions it is most frequently found. This analysis confirmed the presence of DAxud1 in groups of pro-apoptotic and Wnt pathway genes, as previously described; furthermore, stress resistance genes that coding heat shock protein (HSP) family genes were found as hsp70, hsp67, and hsp26. The enrichment of DAxud1 also identified a DNA-binding motif (AYATACATAYATA) that is frequently found in the promoters of these genes. Surprisingly, the following analyses demonstrated that DAxud1 exerts a repressive role on these genes, which are necessary for cell survival. This is coupled with the pro-apoptotic and cell cycle arrest roles of DAxud1, in which repression of hsp70 complements the maintenance of tissue homeostasis through cell survival modulation.

Diffusion Alignment Coefficient (DAC): A Novel Similarity Metric for Protein-Protein Interaction Network

Interaction networks can be used to predict the functions of unknown proteins using known interactions and proteins with known functions. Many graph theory or diffusion-based methods have been proposed, using the assumption that the topological properties of a protein in a network are related to its biological function. Here we seek to improve function prediction by finding more similar neighbors with a new diffusion-based alignment technique to overcome the topological information loss of the node. In this study, we introduce the Diffusion Alignment Coefficient (DAC) algorithm, which combines diffusion, longest common subsequence, and longest common substring techniques to measure the similarity of two nodes in protein interaction networks. As a proof of concept, our experiments, conducted on a real PPI networks S.cerevisiae and Homo Sapiens, demonstrated that our method obtained better results than competitors for MIPS and MSigDB Collections hallmark gene set functional categories. This is the first study to develop a measure of node function similarity using alignment to consider the positions of nodes in protein-protein interaction networks. According to the experimental results, the use of spatial information belonging to the nodes in the network has a positive effect on the detection of more functionally similar neighboring nodes.

Identification of Cancer Driver Genes by Integrating Multiomics Data with Graph Neural Networks

Cancer is a heterogeneous disease that is driven by the accumulation of both genetic and nongenetic alterations, so integrating multiomics data and extracting effective information from them is expected to be an effective way to predict cancer driver genes. In this paper, we first generate comprehensive instructive features for each gene from genomic, epigenomic, transcriptomic levels together with protein–protein interaction (PPI)-networks-derived attributes and then propose a novel semisupervised deep graph learning framework GGraphSAGE to predict cancer driver genes according to the impact of the alterations on a biological system. When applied to eight tumor types, experimental results suggest that GGraphSAGE outperforms several state-of-the-art computational methods for driver genes identification. Moreover, it broadens our current understanding of cancer driver genes from multiomics level and identifies driver genes specific to the tumor type rather than pan-cancer. We expect GGraphSAGE to open new avenues in precision medicine and even further predict drivers for other complex diseases.

An extensive survey on the use of supervised machine learning techniques in the past two decades for prediction of drug side effects

Approved drugs for sale must be effective and safe, implying that the drug's advantages outweigh its known harmful side effects. Side effects (SE) of drugs are one of the common reasons for drug failure that may halt the whole drug discovery pipeline. The side effects might vary from minor concerns like a runny nose to potentially life-threatening issues like liver damage, heart attack, and death. Therefore, predicting the side effects of the drug is vital in drug development, discovery, and design. Supervised machine learning-based side effects prediction task has recently received much attention since it reduces time, chemical waste, design complexity, risk of failure, and cost. The advancement of supervised learning approaches for predicting side effects have emerged as essential computational tools. Supervised machine learning technique provides early information on drug side effects to develop an effective drug based on drug properties. Still, there are several challenges to predicting drug side effects. Thus, a near-exhaustive survey is carried out in this paper on the use of supervised machine learning approaches employed in drug side effects prediction tasks in the past two decades. In addition, this paper also summarized the drug descriptor required for the side effects prediction task, commonly utilized drug properties sources, computational models, and their performances. Finally, the research gap, open problems, and challenges for the further supervised learning-based side effects prediction task have been discussed.

Graph embedding and Gaussian mixture variational autoencoder network for end-to-end analysis of single-cell RNA sequencing data

Single-cell RNA sequencing (scRNA-seq) is a revolutionary technology to determine the precise gene expression of individual cells and identify cell heterogeneity and subpopulations. However, technical limitations of scRNA-seq lead to heterogeneous and sparse data. Here, we present autoCell, a deep-learning approach for scRNA-seq dropout imputation and feature extraction. autoCell is a variational autoencoding network that combines graph embedding and a probabilistic depth Gaussian mixture model to infer the distribution of high-dimensional, sparse scRNA-seq data. We validate autoCell on simulated datasets and biologically relevant scRNA-seq. We show that interpolation of autoCell improves the performance of existing tools in identifying cell developmental trajectories of human preimplantation embryos. We identify disease-associated astrocytes (DAAs) and reconstruct DAA-specific molecular networks and ligand-receptor interactions involved in cell-cell communications using Alzheimer's disease as a prototypical example. autoCell provides a toolbox for end-to-end analysis of scRNA-seq data, including visualization, clustering, imputation, and disease-specific gene network identification.

A binary interaction map between turnip mosaic virus and Arabidopsis thaliana proteomes

Viruses are obligate intracellular parasites that have co-evolved with their hosts to establish an intricate network of protein-protein interactions. Here, we followed a high-throughput yeast two-hybrid screening to identify 378 novel protein-protein interactions between turnip mosaic virus (TuMV) and its natural host Arabidopsis thaliana. We identified the RNA-dependent RNA polymerase NIb as the viral protein with the largest number of contacts, including key salicylic acid-dependent transcription regulators. We verified a subset of 25 interactions in planta by bimolecular fluorescence complementation assays. We then constructed and analyzed a network comprising 399 TuMV-A. thaliana interactions together with intravirus and intrahost connections. In particular, we found that the host proteins targeted by TuMV are enriched in different aspects of plant responses to infections, are more connected and have an increased capacity to spread information throughout the cell proteome, display higher expression levels, and have been subject to stronger purifying selection than expected by chance. The proviral or antiviral role of ten host proteins was validated by characterizing the infection dynamics in the corresponding mutant plants, supporting a proviral role for the transcriptional regulator TGA1. Comparison with similar studies with animal viruses, highlights shared fundamental features in their mode of action.

Host Transcriptome Analysis of Ferret Tissues Following Henipavirus Infection

Ferrets are commonly used as experimental models of infection for a variety of viruses due to their susceptibility to human respiratory viruses and the close resemblance of pathological outcomes found in human infections. Even though ferret-specific reagents are limited, the use of ferrets as a preclinical experimental model of infection has gained considerable interest since the publication of the ferret transcriptome and draft ferret genome. These advances have made it feasible to easily perform whole-genome gene expression analysis in the ferret infection model. Here, we describe methods for genome-wide gene expression analysis using RNA sequence (RNAseq) data obtained from the lung and brain tissues obtained from experimental infections of Hendra (HeV) and Nipah (NiV) viruses in ferrets. We provide detailed methods for RNAseq and representative data for host gene expression profiles of the lung tissues that show early activation of interferon pathways and later activation of inflammation-related pathways.

Multiple pathways promote microtubule stabilization in senescent intestinal epithelial cells

Intestinal epithelial cells are critical for gastrointestinal homeostasis. However, their function declines during aging. The aging-related loss of organ performance is largely driven by the increase in senescent cells. To date, the hallmarks and molecular mechanisms related to cellular senescence are not fully understood. Microtubules control epithelial functions, and we identified microtubule stabilization as a phenotypic marker of senescent intestinal epithelial cells. The senescence inducer determined the pathway to microtubule stabilization. Specifically, enhanced microtubule stability was associated with α-tubulin hyperacetylation or increased abundance of the microtubule-binding protein tau. We show further that overexpression of MAPT, which encodes tau, augmented microtubule stability in intestinal epithelial cells. Notably, pharmacological microtubule stabilization was sufficient to induce cellular senescence. Taken together, this study provides new insights into the molecular mechanisms that control epithelial cell homeostasis. Our results support the concept that microtubule stability serves as a critical cue to trigger intestinal epithelial cell senescence.

Robust and accurate prediction of self-interacting proteins from protein sequence information by exploiting weighted sparse representation based classifier

BACKGROUND

Self-interacting proteins (SIPs), two or more copies of the protein that can interact with each other expressed by one gene, play a central role in the regulation of most living cells and cellular functions. Although numerous SIPs data can be provided by using high-throughput experimental techniques, there are still several shortcomings such as in time-consuming, costly, inefficient, and inherently high in false-positive rates, for the experimental identification of SIPs even nowadays. Therefore, it is more and more significant how to develop efficient and accurate automatic approaches as a supplement of experimental methods for assisting and accelerating the study of predicting SIPs from protein sequence information.

RESULTS

In this paper, we present a novel framework, termed GLCM-WSRC (gray level co-occurrence matrix-weighted sparse representation based classification), for predicting SIPs automatically based on protein evolutionary information from protein primary sequences. More specifically, we firstly convert the protein sequence into Position Specific Scoring Matrix (PSSM) containing protein sequence evolutionary information, exploiting the Position Specific Iterated BLAST (PSI-BLAST) tool. Secondly, using an efficient feature extraction approach, i.e., GLCM, we extract abstract salient and invariant feature vectors from the PSSM, and then perform a pre-processing operation, the adaptive synthetic (ADASYN) technique, to balance the SIPs dataset to generate new feature vectors for classification. Finally, we employ an efficient and reliable WSRC model to identify SIPs according to the known information of self-interacting and non-interacting proteins.

CONCLUSIONS

Extensive experimental results show that the proposed approach exhibits high prediction performance with 98.10% accuracy on the yeast dataset, and 91.51% accuracy on the human dataset, which further reveals that the proposed model could be a useful tool for large-scale self-interacting protein prediction and other bioinformatics tasks detection in the future.

Comprehensive molecular characterization of hypertension-related genes in cancer

Background

During cancer treatment, patients have a significantly higher risk of developing cardiovascular complications such as hypertension. In this study, we investigated the internal relationships between hypertension and different types of cancer.

Methods

First, we comprehensively characterized the involvement of 10 hypertension-related genes across 33 types of cancer. The somatic copy number alteration (CNA) and single nucleotide variant (SNV) of each gene were identified for each type of cancer. Then, the expression patterns of hypertension-related genes were analyzed across 14 types of cancer. The hypertension-related genes were aberrantly expressed in different types of cancer, and some were associated with the overall survival of patients or the cancer stage. Subsequently, the interactions between hypertension-related genes and clinically actionable genes (CAGs) were identified by analyzing the co-expressions and protein–protein interactions.

Results

We found that certain hypertension-related genes were correlated with CAGs. Next, the pathways associated with hypertension-related genes were identified. The positively correlated pathways included epithelial to mesenchymal transition, hormone androgen receptor, and receptor tyrosine kinase, and the negatively correlated pathways included apoptosis, cell cycle, and DNA damage response. Finally, the correlations between hypertension-related genes and drug sensitivity were evaluated for different drugs and different types of cancer. The hypertension-related genes were all positively or negatively correlated with the resistance of cancer to the majority of anti-cancer drugs. These results highlight the importance of hypertension-related genes in cancer.

Conclusions

This study provides an approach to characterize the relationship between hypertension-related genes and cancers in the post-genomic era.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40959-022-00136-z.

Interpretable deep learning translation of GWAS and multi-omics findings to identify pathobiology and drug repurposing in Alzheimer's disease

Translating human genetic findings (genome-wide association studies [GWAS]) to pathobiology and therapeutic discovery remains a major challenge for Alzheimer's disease (AD). We present a network topology-based deep learning framework to identify disease-associated genes (NETTAG). We leverage non-coding GWAS loci effects on quantitative trait loci, enhancers and CpG islands, promoter regions, open chromatin, and promoter flanking regions under the protein-protein interactome. Via NETTAG, we identified 156 AD-risk genes enriched in druggable targets. Combining network-based prediction and retrospective case-control observations with 10 million individuals, we identified that usage of four drugs (ibuprofen, gemfibrozil, cholecalciferol, and ceftriaxone) is associated with reduced likelihood of AD incidence. Gemfibrozil (an approved lipid regulator) is significantly associated with 43% reduced risk of AD compared with simvastatin using an active-comparator design (95% confidence interval 0.51-0.63, p < 0.0001). In summary, NETTAG offers a deep learning methodology that utilizes GWAS and multi-genomic findings to identify pathobiology and drug repurposing in AD.

Multi-Omics Approaches and Resources for Systems-Level Gene Function Prediction in the Plant Kingdom

In higher plants, the complexity of a system and the components within and among species are rapidly dissected by omics technologies. Multi-omics datasets are integrated to infer and enable a comprehensive understanding of the life processes of organisms of interest. Further, growing open-source datasets coupled with the emergence of high-performance computing and development of computational tools for biological sciences have assisted in silico functional prediction of unknown genes, proteins and metabolites, otherwise known as uncharacterized. The systems biology approach includes data collection and filtration, system modelling, experimentation and the establishment of new hypotheses for experimental validation. Informatics technologies add meaningful sense to the output generated by complex bioinformatics algorithms, which are now freely available in a user-friendly graphical user interface. These resources accentuate gene function prediction at a relatively minimal cost and effort. Herein, we present a comprehensive view of relevant approaches available for system-level gene function prediction in the plant kingdom. Together, the most recent applications and sought-after principles for gene mining are discussed to benefit the plant research community. A realistic tabulation of plant genomic resources is included for a less laborious and accurate candidate gene discovery in basic plant research and improvement strategies.

NESM: a network embedding method for tumor stratification by integrating multi-omics data

Tumor stratification plays an important role in cancer diagnosis and individualized treatment. Recent developments in high-throughput sequencing technologies have produced huge amounts of multi-omics data, making it possible to stratify cancer types using multiple molecular datasets. We introduce a Network Embedding method for tumor Stratification by integrating Multi-omics data. Network Embedding method for tumor Stratification by integrating Multi-omics pregroup the samples, integrate the gene features and somatic mutation corresponding to cancer types within each group to construct patient features, and then integrate all groups to obtain comprehensive patient information. The gene features contain network topology information, because it is extracted by integrating deoxyribonucleic acid methylation, messenger ribonucleic acid expression data, and protein–protein interactions through network embedding method. On the one hand, a supervised learning method Light Gradient Boosting Machine is used to classify cancer types based on patient features. When compared with other 3 methods, Network Embedding method for tumor Stratification by integrating Multi-omics has the highest AUC in most cancer types. The average AUC for stratifying cancer types is 0.91, indicating that the patient features extracted by Network Embedding method for tumor Stratification by integrating Multi-omics are effective for tumor stratification. On the other hand, an unsupervised clustering algorithm Density-Based Spatial Clustering of Applications with Noise is utilized to divide single cancer subtypes. The vast majority of the subtypes identified by Network Embedding method for tumor Stratification by integrating Multi-omics are significantly associated with patient survival.

Systems Medicine Design based on Systems Biology Approaches and Deep Neural Network for Gastric Cancer

Gastric cancer (GC) is the third leading cause of cancer death in the world. It is associated with the stimulation of microenvironment, aberrant epigenetic modification, and chronic inflammation. However, few researches discuss the GC molecular progression mechanisms from the perspective of the system level. In this study, we proposed a systems medicine design procedure to identify essential biomarkers and find corresponding drugs for GC. At first, we did big database mining to construct candidate protein-protein interaction network (PPIN) and candidate gene regulation network (GRN). Second, by leveraging the next-generation sequencing (NGS) data, we performed system modeling and applied system identification and model selection to obtain real genome-wide genetic and epigenetic networks (GWGENs). To make the real GWGENs easy to analyze, the principal network projection method was used to extract the core signaling pathways denoted by KEGG pathways. Subsequently, based on the identified biomarkers, we trained a deep neural network of drug-target interaction (DeepDTI) with supervised learning and filtered our candidate drugs considering drug regulation ability and drug sensitivity. With the proposed systematic strategy, we not only shed the light on the progression of GC but also suggested potential multiple-molecule drugs efficiently.

TAIGET: A small-molecule target identification and annotation web server

Background: Accurate target identification of small molecules and downstream target annotation are important in pharmaceutical research and drug development.

Methods: We present TAIGET, a friendly and easy to operate graphical web interface, which consists of a docking module based on AutoDock Vina and LeDock, a target screen module based on a Bayesian–Gaussian mixture model (BGMM), and a target annotation module derived from >14,000 cancer-related literature works.

Results: TAIGET produces binding poses by selecting ≤5 proteins at a time from the UniProt ID-PDB network and submitting ≤3 ligands at a time with the SMILES format. Once the identification process of binding poses is complete, TAIGET then screens potential targets based on the BGMM. In addition, three medical experts and 10 medical students curated associations among drugs, genes, gene regulation, cancer outcome phenotype, 2,170 cancer cell types, and 73 cancer types from the PubMed literature, with the aim to construct a target annotation module. A target-related PPI network can be visualized by an interactive interface.

Conclusion: This online tool significantly lowers the entry barrier of virtual identification of targets for users who are not experts in the technical aspects of virtual drug discovery. The web server is available free of charge at http://www.taiget.cn/.

Pan-cancer proteomic map of 949 human cell lines

The proteome provides unique insights into disease biology beyond the genome and transcriptome. A lack of large proteomic datasets has restricted the identification of new cancer biomarkers. Here, proteomes of 949 cancer cell lines across 28 tissue types are analyzed by mass spectrometry. Deploying a workflow to quantify 8,498 proteins, these data capture evidence of cell-type and post-transcriptional modifications. Integrating multi-omics, drug response, and CRISPR-Cas9 gene essentiality screens with a deep learning-based pipeline reveals thousands of protein biomarkers of cancer vulnerabilities that are not significant at the transcript level. The power of the proteome to predict drug response is very similar to that of the transcriptome. Further, random downsampling to only 1,500 proteins has limited impact on predictive power, consistent with protein networks being highly connected and co-regulated. This pan-cancer proteomic map (ProCan-DepMapSanger) is a comprehensive resource available at https://cellmodelpassports.sanger.ac.uk.

A Yeast BiFC-seq Method for Genome-wide Interactome Mapping

Genome-wide physical protein-protein interaction (PPI) mapping remains a major challenge for current technologies. Here, we reported a high-efficiency BiFC-seq method, yeast-enhanced green fluorescent protein-based bimolecular fluorescence complementation (yEGFP-BiFC) coupled with next-generation DNA sequencing, for interactome mapping. We first applied yEGFP-BiFC method to systematically investigate an intraviral network of the Ebola virus. Two-thirds (9/14) of known interactions of EBOV were recaptured, and five novel interactions were discovered. Next, we used the BiFC-seq method to map the interactome of the tumor protein p53. We identified 97 interactors of p53, more than three-quarters of which were novel. Furthermore, in a more complex background, we screened potential interactors by pooling two BiFC libraries together and revealed a network of 229 interactions among 205 proteins. These results show that BiFC-seq is a highly sensitive, rapid, and economical method for genome-wide interactome mapping.

Single-cell network biology characterizes cell type gene regulation for drug repurposing and phenotype prediction in Alzheimer’s disease

Dysregulation of gene expression in Alzheimer’s disease (AD) remains elusive, especially at the cell type level. Gene regulatory network, a key molecular mechanism linking transcription factors (TFs) and regulatory elements to govern gene expression, can change across cell types in the human brain and thus serve as a model for studying gene dysregulation in AD. However, AD-induced regulatory changes across brain cell types remains uncharted. To address this, we integrated single-cell multi-omics datasets to predict the gene regulatory networks of four major cell types, excitatory and inhibitory neurons, microglia and oligodendrocytes, in control and AD brains. Importantly, we analyzed and compared the structural and topological features of networks across cell types and examined changes in AD. Our analysis shows that hub TFs are largely common across cell types and AD-related changes are relatively more prominent in some cell types (e.g., microglia). The regulatory logics of enriched network motifs (e.g., feed-forward loops) further uncover cell type-specific TF-TF cooperativities in gene regulation. The cell type networks are also highly modular and several network modules with cell-type-specific expression changes in AD pathology are enriched with AD-risk genes. The further disease-module-drug association analysis suggests cell-type candidate drugs and their potential target genes. Finally, our network-based machine learning analysis systematically prioritized cell type risk genes likely involved in AD. Our strategy is validated using an independent dataset which showed that top ranked genes can predict clinical phenotypes (e.g., cognitive impairment) of AD with reasonable accuracy. Overall, this single-cell network biology analysis provides a comprehensive map linking genes, regulatory networks, cell types and drug targets and reveals cell-type gene dysregulation in AD.

Alzheimer’s Disease (AD) is the leading cause of dementia. It affects parts of the brain that control language, behavior, and memory. The human brain is comprised of tens of billions of cells, such as neuronal cells that transmit information via electrical and chemical signals, and glial cells that maintain the brain’s immune system. Researchers have found that AD causes changes in the expression of genes within the brain cells. Gene expression is a tightly regulated process involving interconnected networks of multiple genes. Understanding how these gene networks change in AD is critical to identifying genetic biomarkers and potential drug targets. Using genomic data of post-mortem brains diagnosed with AD and healthy individuals, we identified gene networks that play a crucial role in regulating biological processes within neuronal and glial cells. We utilized these gene networks to make predictions on existing FDA approved drugs that could potentially be repurposed for AD. Furthermore, we used a machine learning strategy to identify novel genes that are more likely to be involved in AD pathology. The systems-level approach lends itself to analysis of single-cell genomics data of other human diseases.

Structural Determinants of Yeast Protein-Protein Interaction Interface Evolution at the Residue Level

Interfaces of contact between proteins play important roles in determining the proper structure and function of protein-protein interactions (PPIs). Therefore, to fully understand PPIs, we need to better understand the evolutionary design principles of PPI interfaces. Previous studies have uncovered that interfacial sites are more evolutionarily conserved than other surface protein sites. Yet, little is known about the nature and relative importance of evolutionary constraints in PPI interfaces. Here, we explore constraints imposed by the structure of the microenvironment surrounding interfacial residues on residue evolutionary rate using a large dataset of over 700 structural models of baker's yeast PPIs. We find that interfacial residues are, on average, systematically more conserved than all other residues with a similar degree of total burial as measured by relative solvent accessibility (RSA). Besides, we find that RSA of the residue when the PPI is formed is a better predictor of interfacial residue evolutionary rate than RSA in the monomer state. Furthermore, we investigate four structure-based measures of residue interfacial involvement, including change in RSA upon binding (ΔRSA), number of residue-residue contacts across the interface, and distance from the center or the periphery of the interface. Integrated modeling for evolutionary rate prediction in interfaces shows that ΔRSA plays a dominant role among the four measures of interfacial involvement, with minor, but independent contributions from other measures. These results yield insight into the evolutionary design of interfaces, improving our understanding of the role that structure plays in the molecular evolution of PPIs at the residue level.

SIPGCN: A Novel Deep Learning Model for Predicting Self-Interacting Proteins from Sequence Information Using Graph Convolutional Networks

Protein is the basic organic substance that constitutes the cell and is the material condition for the life activity and the guarantee of the biological function activity. Elucidating the interactions and functions of proteins is a central task in exploring the mysteries of life. As an important protein interaction, self-interacting protein (SIP) has a critical role. The fast growth of high-throughput experimental techniques among biomolecules has led to a massive influx of available SIP data. How to conduct scientific research using the massive amount of SIP data has become a new challenge that is being faced in related research fields such as biology and medicine. In this work, we design an SIP prediction method SIPGCN using a deep learning graph convolutional network (GCN) based on protein sequences. First, protein sequences are characterized using a position-specific scoring matrix, which is able to describe the biological evolutionary message, then their hidden features are extracted by the deep learning method GCN, and, finally, the random forest is utilized to predict whether there are interrelationships between proteins. In the cross-validation experiment, SIPGCN achieved 93.65% accuracy and 99.64% specificity in the human data set. SIPGCN achieved 90.69% and 99.08% of these two indicators in the yeast data set, respectively. Compared with other feature models and previous methods, SIPGCN showed excellent results. These outcomes suggest that SIPGCN may be a suitable instrument for predicting SIP and may be a reliable candidate for future wet experiments.

Systems Drug Discovery for Diffuse Large B Cell Lymphoma Based on Pathogenic Molecular Mechanism via Big Data Mining and Deep Learning Method

Diffuse large B cell lymphoma (DLBCL) is an aggressive heterogeneous disease. The most common subtypes of DLBCL include germinal center b-cell (GCB) type and activated b-cell (ABC) type. To learn more about the pathogenesis of two DLBCL subtypes (i.e., DLBCL ABC and DLBCL GCB), we firstly construct a candidate genome-wide genetic and epigenetic network (GWGEN) by big database mining. With the help of two DLBCL subtypes’ genome-wide microarray data, we identify their real GWGENs via system identification and model order selection approaches. Afterword, the core GWGENs of two DLBCL subtypes could be extracted from real GWGENs by principal network projection (PNP) method. By comparing core signaling pathways and investigating pathogenic mechanisms, we are able to identify pathogenic biomarkers as drug targets for DLBCL ABC and DLBCL GCD, respectively. Furthermore, we do drug discovery considering drug-target interaction ability, drug regulation ability, and drug toxicity. Among them, a deep neural network (DNN)-based drug-target interaction (DTI) model is trained in advance to predict potential drug candidates holding higher probability to interact with identified biomarkers. Consequently, two drug combinations are proposed to alleviate DLBCL ABC and DLBCL GCB, respectively.

Repurposing Histaminergic Drugs in Multiple Sclerosis

Multiple sclerosis is an autoimmune disease with a strong neuroinflammatory component that contributes to severe demyelination, neurodegeneration and lesions formation in white and grey matter of the spinal cord and brain. Increasing attention is being paid to the signaling of the biogenic amine histamine in the context of several pathological conditions. In multiple sclerosis, histamine regulates the differentiation of oligodendrocyte precursors, reduces demyelination, and improves the remyelination process. However, the concomitant activation of histamine H1–H4 receptors can sustain either damaging or favorable effects, depending on the specifically activated receptor subtype/s, the timing of receptor engagement, and the central versus peripheral target district. Conventional drug development has failed so far to identify curative drugs for multiple sclerosis, thus causing a severe delay in therapeutic options available to patients. In this perspective, drug repurposing offers an exciting and complementary alternative for rapidly approving some medicines already approved for other indications. In the present work, we have adopted a new network-medicine-based algorithm for drug repurposing called SAveRUNNER, for quantifying the interplay between multiple sclerosis-associated genes and drug targets in the human interactome. We have identified new histamine drug-disease associations and predicted off-label novel use of the histaminergic drugs amodiaquine, rupatadine, and diphenhydramine among others, for multiple sclerosis. Our work suggests that selected histamine-related molecules might get to the root causes of multiple sclerosis and emerge as new potential therapeutic strategies for the disease.

Comprehensive network medicine-based drug repositioning via integration of therapeutic efficacy and side effects

Despite advances in modern medicine that led to improvements in cardiovascular outcomes, cardiovascular disease (CVD) remains the leading cause of mortality and morbidity globally. Thus, there is an urgent need for new approaches to improve CVD drug treatments. As the development time and cost of drug discovery to clinical application are excessive, alternate strategies for drug development are warranted. Among these are included computational approaches based on omics data for drug repositioning, which have attracted increasing attention. In this work, we developed an adjusted similarity measure implemented by the algorithm SAveRUNNER to reposition drugs for cardiovascular diseases while, at the same time, considering the side effects of drug candidates. We analyzed nine cardiovascular disorders and two side effects. We formulated both disease disorders and side effects as network modules in the human interactome, and considered those drug candidates that are proximal to disease modules but far from side-effects modules as ideal. Our method provides a list of drug candidates for cardiovascular diseases that are unlikely to produce common, adverse side-effects. This approach incorporating side effects is applicable to other diseases, as well.

Function of SYDE C2-RhoGAP family as signaling hubs for neuronal development deduced by computational analysis

Recent investigations of neurological developmental disorders have revealed the Rho-family modulators such as Syde and its interactors as the candidate genes. Although the mammalian Syde proteins are reported to possess GTPase-accelerating activity for RhoA-family proteins, diverse species-specific substrate selectivities and binding partners have been described, presumably based on their evolutionary variance in the molecular organization. A comprehensive in silico analysis of Syde family proteins was performed to elucidate their molecular functions and neurodevelopmental networks. Predicted structural modeling of the RhoGAP domain may account for the molecular constraints to substrate specificity among Rho-family proteins. Deducing conserved binding motifs can extend the Syde interaction network and highlight diverse but Syde isoform-specific signaling pathways in neuronal homeostasis, differentiation, and synaptic plasticity from novel aspects of post-translational modification and proteolysis.

Identifying Gene Signatures for Cancer Drug Repositioning Based on Sample Clustering

Drug repositioning is an important approach for drug discovery. Computational drug repositioning approaches typically use a gene signature to represent a particular disease and connect the gene signature with drug perturbation profiles. Although disease samples, especially from cancer, may be heterogeneous, most existing methods consider them as a homogeneous set to identify differentially expressed genes (DEGs)for further determining a gene signature. As a result, some genes that should be in a gene signature may be averaged off. In this study, we propose a new framework to identify gene signatures for cancer drug repositioning based on sample clustering (GS4CDRSC). GS4CDRSC first groups samples into several clusters based on their gene expression profiles. Second, an existing method is applied to the samples in each cluster for generating a list of DEGs. Then a weighting approach is used to identify an intergrated gene signature from all the lists of DEGs. The integrated gene signature is used to connect with drug perturbation profiles in the Connectivity Map (CMap)database to generate a list of drug candidates. GS4CDRSC has been tested with several cancer datasets and existing methods. The computational results show that GS4CDRSC outperforms those methods without the sample clustering and weighting approaches in terms of both number and rate of predicted known drugs for specific cancers.

Investigating the Role of Obesity in Prostate Cancer and Identifying Biomarkers for Drug Discovery: Systems Biology and Deep Learning Approaches

Prostate cancer (PCa) is the second most frequently diagnosed cancer for men and is viewed as the fifth leading cause of death worldwide. The body mass index (BMI) is taken as a vital criterion to elucidate the association between obesity and PCa. In this study, systematic methods are employed to investigate how obesity influences the noncutaneous malignancies of PCa. By comparing the core signaling pathways of lean and obese patients with PCa, we are able to investigate the relationships between obesity and pathogenic mechanisms and identify significant biomarkers as drug targets for drug discovery. Regarding drug design specifications, we take drug–target interaction, drug regulation ability, and drug toxicity into account. One deep neural network (DNN)-based drug–target interaction (DTI) model is trained in advance for predicting drug candidates based on the identified biomarkers. In terms of the application of the DNN-based DTI model and the consideration of drug design specifications, we suggest two potential multiple-molecule drugs to prevent PCa (covering lean and obese PCa) and obesity-specific PCa, respectively. The proposed multiple-molecule drugs (apigenin, digoxin, and orlistat) not only help to prevent PCa, suppressing malignant metastasis, but also result in lower production of fatty acids and cholesterol, especially for obesity-specific PCa.

Machine Learning in Causal Inference: Application in Pharmacovigilance

Monitoring adverse drug events or pharmacovigilance has been promoted by the World Health Organization to assure the safety of medicines through a timely and reliable information exchange regarding drug safety issues. We aim to discuss the application of machine learning methods as well as causal inference paradigms in pharmacovigilance. We first reviewed data sources for pharmacovigilance. Then, we examined traditional causal inference paradigms, their applications in pharmacovigilance, and how machine learning methods and causal inference paradigms were integrated to enhance the performance of traditional causal inference paradigms. Finally, we summarized issues with currently mainstream correlation-based machine learning models and how the machine learning community has tried to address these issues by incorporating causal inference paradigms. Our literature search revealed that most existing data sources and tasks for pharmacovigilance were not designed for causal inference. Additionally, pharmacovigilance was lagging in adopting machine learning-causal inference integrated models. We highlight several currently trending directions or gaps to integrate causal inference with machine learning in pharmacovigilance research. Finally, our literature search revealed that the adoption of causal paradigms can mitigate known issues with machine learning models. We foresee that the pharmacovigilance domain can benefit from the progress in the machine learning field.

A full-proteome, interaction-specific characterization of mutational hotspots across human cancers

Rapid accumulation of cancer genomic data has led to the identification of an increasing number of mutational hotspots with uncharacterized significance. Here we present a biologically informed computational framework that characterizes the functional relevance of all 1107 published mutational hotspots identified in approximately 25,000 tumor samples across 41 cancer types in the context of a human 3D interactome network, in which the interface of each interaction is mapped at residue resolution. Hotspots reside in network hub proteins and are enriched on protein interaction interfaces, suggesting that alteration of specific protein-protein interactions is critical for the oncogenicity of many hotspot mutations. Our framework enables, for the first time, systematic identification of specific protein interactions affected by hotspot mutations at the full proteome scale. Furthermore, by constructing a hotspot-affected network that connects all hotspot-affected interactions throughout the whole-human interactome, we uncover genome-wide relationships among hotspots and implicate novel cancer proteins that do not harbor hotspot mutations themselves. Moreover, applying our network-based framework to specific cancer types identifies clinically significant hotspots that can be used for prognosis and therapy targets. Overall, we show that our framework bridges the gap between the statistical significance of mutational hotspots and their biological and clinical significance in human cancers.

Targeting uPA-uPAR interaction to improve intestinal epithelial barrier integrity in inflammatory bowel disease

Background

Loss of intestinal epithelial barrier integrity is a critical component of Inflammatory Bowel Disease (IBD) pathogenesis. Co-expression regulation of ligand-receptor pairs in IBD mucosa has not been systematically studied. Targeting ligand-receptor pairs which are induced in IBD mucosa and function in intestinal epithelial barrier integrity may provide novel therapeutics for IBD.

Methods

We performed transcriptomic meta-analysis on public IBD datasets combined with cell surface protein-protein-interaction (PPI) databases. We explored primary human/mouse intestinal organoids and Caco-2 cells for expression and function studies of uPA-uPAR (prime hits from the meta-analysis). Epithelial barrier integrity was measured by Trans-Epithelial Electrical Resistance (TEER), FITC-Dextran permeability and tight junction assessment. Genetic (CRISPR, siRNA and KO mice) and pharmacological (small molecules, neutralizing antibody and peptide inhibitors) approaches were applied. Mice deficient of uPAR were studied using the Dextran Sulfate Sodium (DSS)-induced colitis model.

Findings

The IBD ligand-receptor meta-analysis led to the discovery of a coordinated upregulation of uPA and uPAR in IBD mucosa. Both genes were significantly upregulated during epithelial barrier breakdown in primary intestinal organoids and decreased during barrier formation. Genetic inhibition of uPAR or uPA, or pharmacologically blocking uPA-uPAR interaction protects against cytokine-induced barrier breakdown. Deficiency of uPAR in epithelial cells leads to enhanced EGF/EGFR signalling, a known regulator of epithelial homeostasis and repair. Mice deficient of uPAR display improved intestinal barrier function in vitro and during DSS-induced colitis in vivo.

Interpretation

Our findings suggest that blocking uPA-uPAR interaction via pharmacological agents protects the epithelial barrier from inflammation-induced damage, indicating a potential therapeutic target for IBD.

Funding

The study was funded by Boehringer Ingelheim.

Interpretation of network-based integration from multi-omics longitudinal data

Multi-omics integration is key to fully understand complex biological processes in an holistic manner. Furthermore, multi-omics combined with new longitudinal experimental design can unreveal dynamic relationships between omics layers and identify key players or interactions in system development or complex phenotypes. However, integration methods have to address various experimental designs and do not guarantee interpretable biological results. The new challenge of multi-omics integration is to solve interpretation and unlock the hidden knowledge within the multi-omics data. In this paper, we go beyond integration and propose a generic approach to face the interpretation problem. From multi-omics longitudinal data, this approach builds and explores hybrid multi-omics networks composed of both inferred and known relationships within and between omics layers. With smart node labelling and propagation analysis, this approach predicts regulation mechanisms and multi-omics functional modules. We applied the method on 3 case studies with various multi-omics designs and identified new multi-layer interactions involved in key biological functions that could not be revealed with single omics analysis. Moreover, we highlighted interplay in the kinetics that could help identify novel biological mechanisms. This method is available as an R package netOmics to readily suit any application.

Protein-protein interactions: Methods, databases, and applications in virus-host study

Almost all the cellular processes in a living system are controlled by proteins: They regulate gene expression, catalyze chemical reactions, transport small molecules across membranes, and transmit signal across membranes. Even, a viral infection is often initiated through virus-host protein interactions. Protein-protein interactions (PPIs) are the physical contacts between two or more proteins and they represent complex biological functions. Nowadays, PPIs have been used to construct PPI networks to study complex pathways for revealing the functions of unknown proteins. Scientists have used PPIs to find the molecular basis of certain diseases and also some potential drug targets. In this review, we will discuss how PPI networks are essential to understand the molecular basis of virus-host relationships and several databases which are dedicated to virus-host interaction studies. Here, we present a short but comprehensive review on PPIs, including the experimental and computational methods of finding PPIs, the databases dedicated to virus-host PPIs, and the associated various applications in protein interaction networks of some lethal viruses with their hosts.

Tumor relevant protein functional interactions identified using bipartite graph analyses

An increased surge of -omics data for the diseases such as cancer allows for deriving insights into the affiliated protein interactions. We used bipartite network principles to build protein functional associations of the differentially regulated genes in 18 cancer types. This approach allowed us to combine expression data to functional associations in many cancers simultaneously. Further, graph centrality measures suggested the importance of upregulated genes such as BIRC5, UBE2C, BUB1B, KIF20A and PTH1R in cancer. Pathway analysis of the high centrality network nodes suggested the importance of the upregulation of cell cycle and replication associated proteins in cancer. Some of the downregulated high centrality proteins include actins, myosins and ATPase subunits. Among the transcription factors, mini-chromosome maintenance proteins (MCMs) and E2F family proteins appeared prominently in regulating many differentially regulated genes. The projected unipartite networks of the up and downregulated genes were comprised of 37,411 and 41,756 interactions, respectively. The conclusions obtained by collating these interactions revealed pan-cancer as well as subtype specific protein complexes and clusters. Therefore, we demonstrate that incorporating expression data from multiple cancers into bipartite graphs validates existing cancer associated mechanisms as well as directs to novel interactions and pathways.

Computational repurposing of therapeutic small molecules from cancer to pulmonary hypertension

Cancer therapies are being considered for treating rare noncancerous diseases like pulmonary hypertension (PH), but effective computational screening is lacking. Via transcriptomic differential dependency analyses leveraging parallels between cancer and PH, we mapped a landscape of cancer drug functions dependent upon rewiring of PH gene clusters. Bromodomain and extra-terminal motif (BET) protein inhibitors were predicted to rely upon several gene clusters inclusive of galectin-8 (LGALS8). Correspondingly, LGALS8 was found to mediate the BET inhibitor–dependent control of endothelial apoptosis, an essential role for PH in vivo. Separately, a piperlongumine analog’s actions were predicted to depend upon the iron-sulfur biogenesis gene ISCU. Correspondingly, the analog was found to inhibit ISCU glutathionylation, rescuing oxidative metabolism, decreasing endothelial apoptosis, and improving PH. Thus, we identified crucial drug-gene axes central to endothelial dysfunction and therapeutic priorities for PH. These results establish a wide-ranging, network dependency platform to redefine cancer drugs for use in noncancerous conditions.

Network Biology Approaches to Achieve Precision Medicine in Inflammatory Bowel Disease

Inflammatory bowel disease (IBD) is a chronic immune-mediated condition arising due to complex interactions between multiple genetic and environmental factors. Despite recent advances, the pathogenesis of the condition is not fully understood and patients still experience suboptimal clinical outcomes. Over the past few years, investigators are increasingly capturing multi-omics data from patient cohorts to better characterise the disease. However, reaching clinically translatable endpoints from these complex multi-omics datasets is an arduous task. Network biology, a branch of systems biology that utilises mathematical graph theory to represent, integrate and analyse biological data through networks, will be key to addressing this challenge. In this narrative review, we provide an overview of various types of network biology approaches that have been utilised in IBD including protein-protein interaction networks, metabolic networks, gene regulatory networks and gene co-expression networks. We also include examples of multi-layered networks that have combined various network types to gain deeper insights into IBD pathogenesis. Finally, we discuss the need to incorporate other data sources including metabolomic, histopathological, and high-quality clinical meta-data. Together with more robust network data integration and analysis frameworks, such efforts have the potential to realise the key goal of precision medicine in IBD.

Gene expression signatures predict response to therapy with growth hormone

Recombinant human growth hormone (r-hGH) is used as a therapeutic agent for disorders of growth including growth hormone deficiency (GHD) and Turner syndrome (TS). Treatment is costly and current methods to model response are inexact. GHD (n = 71) and TS patients (n = 43) were recruited to study response to r-hGH over 5 years. Analysis was performed using 1219 genetic markers and baseline (pre-treatment) blood transcriptome. Random forest was used to determine predictive value of transcriptomic data associated with growth response. No genetic marker passed the stringency criteria for prediction. However, we identified an identical set of genes in both GHD and TS whose expression could be used to classify therapeutic response to r-hGH with a high accuracy (AUC > 0.9). Combining transcriptomic markers with clinical phenotype was shown to significantly reduce predictive error. This work could be translated into a single genomic test linked to a prediction algorithm to improve clinical management. Trial registration numbers: NCT00256126 and NCT00699855.

Deep learning based prediction of reversible HAT/HDAC-specific lysine acetylation

Protein lysine acetylation regulation is an important molecular mechanism for regulating cellular processes and plays critical physiological and pathological roles in cancers and diseases. Although massive acetylation sites have been identified through experimental identification and high-throughput proteomics techniques, their enzyme-specific regulation remains largely unknown. Here, we developed the deep learning-based protein lysine acetylation modification prediction (Deep-PLA) software for histone acetyltransferase (HAT)/histone deacetylase (HDAC)-specific acetylation prediction based on deep learning. Experimentally identified substrates and sites of several HATs and HDACs were curated from the literature to generate enzyme-specific data sets. We integrated various protein sequence features with deep neural network and optimized the hyperparameters with particle swarm optimization, which achieved satisfactory performance. Through comparisons based on cross-validations and testing data sets, the model outperformed previous studies. Meanwhile, we found that protein-protein interactions could enrich enzyme-specific acetylation regulatory relations and visualized this information in the Deep-PLA web server. Furthermore, a cross-cancer analysis of acetylation-associated mutations revealed that acetylation regulation was intensively disrupted by mutations in cancers and heavily implicated in the regulation of cancer signaling. These prediction and analysis results might provide helpful information to reveal the regulatory mechanism of protein acetylation in various biological processes to promote the research on prognosis and treatment of cancers. Therefore, the Deep-PLA predictor and protein acetylation interaction networks could provide helpful information for studying the regulation of protein acetylation. The web server of Deep-PLA could be accessed at http://deeppla.cancerbio.info.

Universal concept signature analysis: genome-wide quantification of new biological and pathological functions of genes and pathways

Identifying new gene functions and pathways underlying diseases and biological processes are major challenges in genomics research. Particularly, most methods for interpreting the pathways characteristic of an experimental gene list defined by genomic data are limited by their dependence on assessing the overlapping genes or their interactome topology, which cannot account for the variety of functional relations. This is particularly problematic for pathway discovery from single-cell genomics with low gene coverage or interpreting complex pathway changes such as during change of cell states. Here, we exploited the comprehensive sets of molecular concepts that combine ontologies, pathways, interactions and domains to help inform the functional relations. We first developed a universal concept signature (uniConSig) analysis for genome-wide quantification of new gene functions underlying biological or pathological processes based on the signature molecular concepts computed from known functional gene lists. We then further developed a novel concept signature enrichment analysis (CSEA) for deep functional assessment of the pathways enriched in an experimental gene list. This method is grounded on the framework of shared concept signatures between gene sets at multiple functional levels, thus overcoming the limitations of the current methods. Through meta-analysis of transcriptomic data sets of cancer cell line models and single hematopoietic stem cells, we demonstrate the broad applications of CSEA on pathway discovery from gene expression and single-cell transcriptomic data sets for genetic perturbations and change of cell states, which complements the current modalities. The R modules for uniConSig analysis and CSEA are available through https://github.com/wangxlab/uniConSig.

A machine learning framework for predicting drug-drug interactions

Understanding drug–drug interactions is an essential step to reduce the risk of adverse drug events before clinical drug co-prescription. Existing methods, commonly integrating heterogeneous data to increase model performance, often suffer from a high model complexity, As such, how to elucidate the molecular mechanisms underlying drug–drug interactions while preserving rational biological interpretability is a challenging task in computational modeling for drug discovery. In this study, we attempt to investigate drug–drug interactions via the associations between genes that two drugs target. For this purpose, we propose a simple f drug target profile representation to depict drugs and drug pairs, from which an l₂-regularized logistic regression model is built to predict drug–drug interactions. Furthermore, we define several statistical metrics in the context of human protein–protein interaction networks and signaling pathways to measure the interaction intensity, interaction efficacy and action range between two drugs. Large-scale empirical studies including both cross validation and independent test show that the proposed drug target profiles-based machine learning framework outperforms existing data integration-based methods. The proposed statistical metrics show that two drugs easily interact in the cases that they target common genes; or their target genes connect via short paths in protein–protein interaction networks; or their target genes are located at signaling pathways that have cross-talks. The unravelled mechanisms could provide biological insights into potential adverse drug reactions of co-prescribed drugs.

From DNA human sequence to the chromatin higher order organisation and its biological meaning: Using biomolecular interaction networks to understand the influence of structural variation on spatial genome organisation and its functional effect

The three-dimensional structure of the human genome has been proven to have a significant functional impact on gene expression. The high-order spatial chromatin is organised first by looping mediated by multiple protein factors, and then it is further formed into larger structures of topologically associated domains (TADs) or chromatin contact domains (CCDs), followed by A/B compartments and finally the chromosomal territories (CTs). The genetic variation observed in human population influences the multi-scale structures, posing a question regarding the functional impact of structural variants reflected by the variability of the genes expression patterns. The current methods of evaluating the functional effect include eQTLs analysis which uses statistical testing of influence of variants on spatially close genes. Rarely, non-coding DNA sequence changes are evaluated by their impact on the biomolecular interaction network (BIN) reflecting the cellular interactome that can be analysed by the classical graph-theoretic algorithms. Therefore, in the second part of the review, we introduce the concept of BIN, i.e. a meta-network model of the complete molecular interactome developed by integrating various biological networks. The BIN meta-network model includes DNA-protein binding by the plethora of protein factors as well as chromatin interactions, therefore allowing connection of genomics with the downstream biomolecular processes present in a cell. As an illustration, we scrutinise the chromatin interactions mediated by the CTCF protein detected in a ChIA-PET experiment in the human lymphoblastoid cell line GM12878. In the corresponding BIN meta-network the DNA spatial proximity is represented as a graph model, combined with the Proteins-Interaction Network (PIN) of human proteome using the Gene Association Network (GAN). Furthermore, we enriched the BIN with the signalling and metabolic pathways and Gene Ontology (GO) terms to assert its functional context. Finally, we mapped the Single Nucleotide Polymorphisms (SNPs) from the GWAS studies and identified the chromatin mutational hot-spots associated with a significant enrichment of SNPs related to autoimmune diseases. Afterwards, we mapped Structural Variants (SVs) from healthy individuals of 1000 Genomes Project and identified an interesting example of the missing protein complex associated with protein Q6GYQ0 due to a deletion on chromosome 14. Such an analysis using the meta-network BIN model is therefore helpful in evaluating the influence of genetic variation on spatial organisation of the genome and its functional effect in a cell.