An application of node and edge nonlinear hypergraph centrality to a protein complex hypernetwork

The use of graph centrality measures applied to biological networks, such as protein interaction networks, underpins much research into identifying key players within biological processes. This approach however is restricted to dyadic interactions and it is well-known that in many instances interactions are polyadic. In this study we illustrate the merit of using hypergraph centrality applied to a hypernetwork as an alternative. Specifically, we review and propose an extension to a recently introduced node and edge nonlinear hypergraph centrality model which provides mutually dependent node and edge centralities. A Saccharomyces Cerevisiae protein complex hypernetwork is used as an example application with nodes representing proteins and hyperedges representing protein complexes. The resulting rankings of the nodes and edges are considered to see if they provide insight into the essentiality of the proteins and complexes. We find that certain variations of the model predict essentiality more accurately and that the degree-based variation illustrates that the centrality-lethality rule extends to a hypergraph setting. In particular, through exploitation of the models flexibility, we identify small sets of proteins densely populated with essential proteins. One of the key advantages of applying this model to a protein complex hypernetwork is that it also provides a classification method for protein complexes, unlike previous approaches which are only concerned with classifying proteins.

ACDMBI: A deep learning model based on community division and multi-source biological information fusion predicts essential proteins

Accurately identifying essential proteins is vital for drug research and disease diagnosis. Traditional centrality methods and machine learning approaches often face challenges in accurately discerning essential proteins, primarily relying on information derived from protein-protein interaction (PPI) networks. Despite attempts by some researchers to integrate biological data and PPI networks for predicting essential proteins, designing effective integration methods remains a challenge. In response to these challenges, this paper presents the ACDMBI model, specifically designed to overcome the aforementioned issues. ACDMBI is comprised of two key modules: feature extraction and classification. In terms of capturing relevant information, we draw insights from three distinct data sources. Initially, structural features of proteins are extracted from the PPI network through community division. Subsequently, these features are further optimized using Graph Convolutional Networks (GCN) and Graph Attention Networks (GAT). Moving forward, protein features are extracted from gene expression data utilizing Bidirectional Long Short-Term Memory networks (BiLSTM) and a multi-head self-attention mechanism. Finally, protein features are derived by mapping subcellular localization data to a one-dimensional vector and processing it through fully connected layers. In the classification phase, we integrate features extracted from three different data sources, crafting a multi-layer deep neural network (DNN) for protein classification prediction. Experimental results on brewing yeast data showcase the ACDMBI model's superior performance, with AUC reaching 0.9533 and AUPR reaching 0.9153. Ablation experiments further reveal that the effective integration of features from diverse biological information significantly boosts the model's performance.

A protein network refinement method based on module discovery and biological information

BACKGROUND

The identification of essential proteins can help in understanding the minimum requirements for cell survival and development to discover drug targets and prevent disease. Nowadays, node ranking methods are a common way to identify essential proteins, but the poor data quality of the underlying PIN has somewhat hindered the identification accuracy of essential proteins for these methods in the PIN. Therefore, researchers constructed refinement networks by considering certain biological properties of interacting protein pairs to improve the performance of node ranking methods in the PIN. Studies show that proteins in a complex are more likely to be essential than proteins not present in the complex. However, the modularity is usually ignored for the refinement methods of the PINs.

METHODS

Based on this, we proposed a network refinement method based on module discovery and biological information. The idea is, first, to extract the maximal connected subgraph in the PIN, and to divide it into different modules by using Fast-unfolding algorithm; then, to detect critical modules according to the orthologous information, subcellular localization information and topology information within each module; finally, to construct a more refined network (CM-PIN) by using the identified critical modules.

RESULTS

To evaluate the effectiveness of the proposed method, we used 12 typical node ranking methods (LAC, DC, DMNC, NC, TP, LID, CC, BC, PR, LR, PeC, WDC) to compare the overall performance of the CM-PIN with those on the S-PIN, D-PIN and RD-PIN. The experimental results showed that the CM-PIN was optimal in terms of the identification number of essential proteins, precision-recall curve, Jackknifing method and other criteria, and can help to identify essential proteins more accurately.

Bowel function and inflammation: Is motility the other side of the coin?

Digestion and intestinal absorption allow the body to sustain itself and are the emblematic functions of the bowel. On the flip side, functions also arise from its role as an interface with the environment. Indeed, the gut houses microorganisms, collectively known as the gut microbiota, which interact with the host, and is the site of complex immune activities. Its role in human pathology is complex and scientific evidence is progressively elucidating the functions of the gut, especially regarding the pathogenesis of chronic intestinal diseases and inflammatory conditions affecting various organs and systems. This editorial aims to highlight and relate the factors involved in the pathogenesis of intestinal and systemic inflammation.

Integration of graph neural networks and genome-scale metabolic models for predicting gene essentiality

Genome-scale metabolic models are powerful tools for understanding cellular physiology. Flux balance analysis (FBA), in particular, is an optimization-based approach widely employed for predicting metabolic phenotypes. In model microbes such as Escherichia coli, FBA has been successful at predicting essential genes, i.e. those genes that impair survival when deleted. A central assumption in this approach is that both wild type and deletion strains optimize the same fitness objective. Although the optimality assumption may hold for the wild type metabolic network, deletion strains are not subject to the same evolutionary pressures and knock-out mutants may steer their metabolism to meet other objectives for survival. Here, we present FlowGAT, a hybrid FBA-machine learning strategy for predicting essentiality directly from wild type metabolic phenotypes. The approach is based on graph-structured representation of metabolic fluxes predicted by FBA, where nodes correspond to enzymatic reactions and edges quantify the propagation of metabolite mass flow between a reaction and its neighbours. We integrate this information into a graph neural network that can be trained on knock-out fitness assay data. Comparisons across different model architectures reveal that FlowGAT predictions for E. coli are close to those of FBA for several growth conditions. This suggests that essentiality of enzymatic genes can be predicted by exploiting the inherent network structure of metabolism. Our approach demonstrates the benefits of combining the mechanistic insights afforded by genome-scale models with the ability of deep learning to infer patterns from complex datasets.

RFEM: A framework for essential microRNA identification in mice based on rotation forest and multiple feature fusion

With the increasing number of microRNAs (miRNAs), identifying essential miRNAs has become an important task that needs to be solved urgently. However, there are few computational methods for essential miRNA identification. Here, we proposed a novel framework called Rotation Forest for Essential MicroRNA identification (RFEM) to predict the essentiality of miRNAs in mice. We first constructed 1,264 miRNA features of all miRNA samples by fusing 38 miRNA features obtained from the PESM paper and 1,226 miRNA functional features calculated based on miRNA-target gene interactions. Then, we employed 182 training samples with 1,264 features to train the rotation forest model, which was applied to compute the essentiality scores of the candidate samples. The main innovations of RFEM were as follows: 1) miRNA functional features were introduced to enrich the diversity of miRNA features; 2) the rotation forest model used decision tree as the base classifier and could increase the difference among base classifiers through feature transformation to achieve better ensemble results. Experimental results show that RFEM significantly outperformed two previous models with the AUC (AUPR) of 0.942 (0.944) in three comparison experiments under 5-fold cross validation, which proved the model's reliable performance. Moreover, ablation study was further conducted to demonstrate the effectiveness of the novel miRNA functional features. Additionally, in the case studies of assessing the essentiality of unlabeled miRNAs, experimental literature confirmed that 7 of the top 10 predicted miRNAs have crucial biological functions in mice. Therefore, RFEM would be a reliable tool for identifying essential miRNAs.

MEG-PPIS: a fast protein-protein interaction site prediction method based on multi-scale graph information and equivariant graph neural network

MOTIVATION

Protein-protein interaction sites (PPIS) are crucial for deciphering protein action mechanisms and related medical research, which is the key issue in protein action research. Recent studies have shown that graph neural networks have achieved outstanding performance in predicting PPIS. However, these studies often neglect the modeling of information at different scales in the graph and the symmetry of protein molecules within three-dimensional space.

RESULTS

In response to this gap, this article proposes the MEG-PPIS approach, a PPIS prediction method based on multi-scale graph information and E(n) equivariant graph neural network (EGNN). There are two channels in MEG-PPIS: the original graph and the subgraph obtained by graph pooling. The model can iteratively update the features of the original graph and subgraph through the weight-sharing EGNN. Subsequently, the max-pooling operation aggregates the updated features of the original graph and subgraph. Ultimately, the model feeds node features into the prediction layer to obtain prediction results. Comparative assessments against other methods on benchmark datasets reveal that MEG-PPIS achieves optimal performance across all evaluation metrics and gets the fastest runtime. Furthermore, specific case studies demonstrate that our method can predict more true positive and true negative sites than the current best method, proving that our model achieves better performance in the PPIS prediction task.

AVAILABILITY AND IMPLEMENTATION

The data and code are available at https://github.com/dhz234/MEG-PPIS.git.

Untangling the Context-Specificity of Essential Genes by Means of Machine Learning: A Constructive Experience

Gene essentiality is a genetic concept crucial for a comprehensive understanding of life and evolution. In the last decade, many essential genes (EGs) have been determined using different experimental and computational approaches, and this information has been used to reduce the genomes of model organisms. A growing amount of evidence highlights that essentiality is a property that depends on the context. Because of their importance in vital biological processes, recognising context-specific EGs (csEGs) could help for identifying new potential pharmacological targets and to improve precision therapeutics. Since most of the computational procedures proposed to identify and predict EGs neglect their context-specificity, we focused on this aspect, providing a theoretical and experimental overview of the literature, data and computational methods dedicated to recognising csEGs. To this end, we adapted existing computational methods to exploit a specific context (the kidney tissue) and experimented with four different prediction methods using the labels provided by four different identification approaches. The considerations derived from the analysis of the obtained results, confirmed and validated also by further experiments for a different tissue context, provide the reader with guidance on exploiting existing tools for achieving csEGs identification and prediction.

Protein-protein interaction site prediction by model ensembling with hybrid feature and self-attention

BACKGROUND

Protein-protein interactions (PPIs) are crucial in various biological functions and cellular processes. Thus, many computational approaches have been proposed to predict PPI sites. Although significant progress has been made, these methods still have limitations in encoding the characteristics of each amino acid in sequences. Many feature extraction methods rely on the sliding window technique, which simply merges all the features of residues into a vector. The importance of some key residues may be weakened in the feature vector, leading to poor performance.

RESULTS

We propose a novel sequence-based method for PPI sites prediction. The new network model, PPINet, contains multiple feature processing paths. For a residue, the PPINet extracts the features of the targeted residue and its context separately. These two types of features are processed by two paths in the network and combined to form a protein representation, where the two types of features are of relatively equal importance. The model ensembling technique is applied to make use of more features. The base models are trained with different features and then ensembled via stacking. In addition, a data balancing strategy is presented, by which our model can get significant improvement on highly unbalanced data.

CONCLUSION

The proposed method is evaluated on a fused dataset constructed from Dset186, Dset_72, and PDBset_164, as well as the public Dset_448 dataset. Compared with current state-of-the-art methods, the performance of our method is better than the others. In the most important metrics, such as AUPRC and recall, it surpasses the second-best programmer on the latter dataset by 6.9% and 4.7%, respectively. We also demonstrated that the improvement is essentially due to using the ensemble model, especially, the hybrid feature. We share our code for reproducibility and future research at https://github.com/CandiceCong/StackingPPINet .

Over-representation analysis of angiogenic factors in immunosuppressive mechanisms in neoplasms and neurological conditions during COVID-19

BACKGROUND

Recent studies emphasized the necessity to identify key (human) biological processes and pathways targeted by the Coronaviridae family of viruses, especially Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Coronavirus Disease (COVID-19) caused up to 33-55 % death rates in COVID-19 patients with malignant neoplasms and Alzheimer's disease. Given this scenario, we identified biological processes and pathways involved in various diseases which are most likely affected by COVID-19.

METHODS

The COVID-19 DisGeNET data set (v4.0) contains the associations between various diseases and human genes known to interact with viruses from Coronaviridae family and were obtained from the IntAct Coronavirus data set annotated with DisGeNET data. We constructed the disease-gene network to identify genes that are involved in various comorbid diseased states. Communities from the disease-gene network were identified using Louvain method and functional enrichment through over-representation analysis methodology was used to discover significant biological processes and pathways shared between COVID-19 and other diseases.

RESULT

The COVID-19 DisGeNET data set (v4.0) comprised of 828 human genes and 10,473 diseases (including various phenotypes) that together constituted nodes in the disease-gene network. Each of the 70,210 edges connects a human gene with an associated disease. The top 10 genes linked to most number of diseases were VEGFA, BCL2, CTNNB1, ALB, COX2, AGT, HLA-A, HMOX1, FGF2 and COMT. The most vulnerable group of patients thus discovered had comorbid conditions such as carcinomas, malignant neoplasms and Alzheimer's disease. Finally, we identified 15 potentially useful biological processes and pathways for improved therapies. Vascular endothelial growth factor (VEGF) is the key mediator of angiogenesis in cancer. It is widely distributed in the brain and plays a crucial role in brain inflammation regulating the level of angiopoietins. With a degree of 1899, VEGFA was associated with maximum number of diseases in the disease-gene network. Previous studies have indicated that increased levels of VEGFA in the blood results in dyspnea, Pulmonary Edema (PE), Acute Lung Injury (ALI) and Acute Respiratory Distress Syndrome (ARDS). In case of COVID-19 patients with neoplasms and other neurological symptoms, our results indicate VEGFA as a therapeutic target for inflammation suppression. As VEGFs are known to disproportionately affect cancer patients, improving endothelial permeability and vasodilation with anti-VEGF therapy could lead to suppression of inflammation and also improve oxygenation. As an outcome of our study, we make case for clinical investigations towards anti-VEGF therapies for such comorbid conditions affected by COVID-19 for better therapeutic outcomes.

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.

'Bingo'-a large language model- and graph neural network-based workflow for the prediction of essential genes from protein data

The identification and characterization of essential genes are central to our understanding of the core biological functions in eukaryotic organisms, and has important implications for the treatment of diseases caused by, for example, cancers and pathogens. Given the major constraints in testing the functions of genes of many organisms in the laboratory, due to the absence of in vitro cultures and/or gene perturbation assays for most metazoan species, there has been a need to develop in silico tools for the accurate prediction or inference of essential genes to underpin systems biological investigations. Major advances in machine learning approaches provide unprecedented opportunities to overcome these limitations and accelerate the discovery of essential genes on a genome-wide scale. Here, we developed and evaluated a large language model- and graph neural network (LLM-GNN)-based approach, called 'Bingo', to predict essential protein-coding genes in the metazoan model organisms Caenorhabditis elegans and Drosophila melanogaster as well as in Mus musculus and Homo sapiens (a HepG2 cell line) by integrating LLM and GNNs with adversarial training. Bingo predicts essential genes under two 'zero-shot' scenarios with transfer learning, showing promise to compensate for a lack of high-quality genomic and proteomic data for non-model organisms. In addition, the attention mechanisms and GNNExplainer were employed to manifest the functional sites and structural domain with most contribution to essentiality. In conclusion, Bingo provides the prospect of being able to accurately infer the essential genes of little- or under-studied organisms of interest, and provides a biological explanation for gene essentiality.

Integrated multi-omics analysis identifies epigenetic alteration related to neurodegeneration development in post-traumatic stress disorder patients

INTRODUCTION

Post-traumatic stress disorder (PTSD), is associated with an elevated risk of neurodegenerative disorders, but the molecular mechanism was not wholly identified. Aberrant methylation status and miRNA expression pattern have been identified to be associated with PTSD, but their complex regulatory networks remain largely unexplored.

METHODS

The purpose of this study was to identify the key genes/pathways related to neurodegenerative disorder development in PTSD by evaluating epigenetic regulatory signature (DNA methylation and miRNA) using an integrative bioinformatic analysis. We integrated DNA expression array data with miRNA and DNA methylation array data - obtained from the GEO database- to evaluate the epigenetic regulatory mechanisms.

RESULTS

Our results indicated that target genes of dysregulated miRNAs were significantly related to several neurodegenerative diseases. Several dysregulated genes in the neurodegeneration pathways interacted with some members of the miR-17 and miR-15/107 families. Our analysis indicated that APP/CaN/NFATs signaling pathway was dysregulated in the peripheral blood samples of PTSD. Besides, the DNMT3a and KMT2D genes, as the encoding DNA and histone methyltransferase enzymes, were upregulated, and DNA methylation and miRNA regulators were proposed as critical molecular mechanisms. Our study found dysregulation of circadian rhythm as the CLOCK gene was upregulated and hypomethylated at TSS1500 CpGs S_shores and was also a target of several dysregulated miRNAs.

CONCLUSION

In conclusion, we found evidence of a negative feedback loop between stress oxidative, circadian rhythm dysregulation, miR-17 and miR-15/107 families, some essential genes involved in neuronal and brain cell health, and KMT2D/DNMT3a in the peripheral blood samples of PTSD.

A Construction Method of Dynamic Protein Interaction Networks by Using Relevant Features of Gene Expression Data

Essential proteins play an important role in various life activities and are considered to be a vital part of the organism. Gene expression data are an important dataset to construct dynamic protein-protein interaction networks (DPIN). The existing methods for the construction of DPINs generally utilize all features (or the features in a cycle) of the gene expression data. However, the features observed from successive time points tend to be highly correlated, and thus there are some redundant and irrelevant features in the gene expression data, which will influence the quality of the constructed network and the predictive performance of essential proteins. To address this problem, we propose a construction method of DPINs by using selected relevant features rather than continuous and periodic features. We adopt an improved unsupervised feature selection method based on Laplacian algorithm to remove irrelevant and redundant features from gene expression data, then integrate the chosen relevant features into the static protein-protein interaction network (SPIN) to construct a more concise and effective DPIN (FS-DPIN). To evaluate the effectiveness of the FS-DPIN, we apply 15 network-based centrality methods on the FS-DPIN and compare the results with those on the SPIN and the existing DPINs. Then the predictive performance of the 15 centrality methods is validated in terms of sensitivity, specificity, positive predictive value, negative predictive value, F-measure, accuracy, Jackknife and AUPRC. The experimental results show that the FS-DPIN is superior to the existing DPINs in the identification accuracy of essential proteins.

Integration of text mining and biological network analysis: Identification of essential genes in sulfate-reducing bacteria

The growth and survival of an organism in a particular environment is highly depends on the certain indispensable genes, termed as essential genes. Sulfate-reducing bacteria (SRB) are obligate anaerobes which thrives on sulfate reduction for its energy requirements. The present study used Oleidesulfovibrio alaskensis G20 (OA G20) as a model SRB to categorize the essential genes based on their key metabolic pathways. Herein, we reported a feedback loop framework for gene of interest discovery, from bio-problem to gene set of interest, leveraging expert annotation with computational prediction. Defined bio-problem was applied to retrieve the genes of SRB from literature databases (PubMed, and PubMed Central) and annotated them to the genome of OA G20. Retrieved gene list was further used to enrich protein-protein interaction and was corroborated to the pangenome analysis, to categorize the enriched gene sets and the respective pathways under essential and non-essential. Interestingly, the sat gene (dde_2265) from the sulfur metabolism was the bridging gene between all the enriched pathways. Gene clusters involved in essential pathways were linked with the genes from seleno-compound metabolism, amino acid metabolism, secondary metabolite synthesis, and cofactor biosynthesis. Furthermore, pangenome analysis demonstrated the gene distribution, where 69.83% of the 116 enriched genes were mapped under "persistent," inferring the essentiality of these genes. Likewise, 21.55% of the enriched genes, which involves specially the formate dehydrogenases and metallic hydrogenases, appeared under "shell." Our methodology suggested that semi-automated text mining and network analysis may play a crucial role in deciphering the previously unexplored genes and key mechanisms which can help to generate a baseline prior to perform any experimental studies.

Identifying Tumor-Associated Genes from Bilayer Networks of DNA Methylation Sites and RNAs

Network theory has attracted much attention from the biological community because of its high efficacy in identifying tumor-associated genes. However, most researchers have focused on single networks of single omics, which have less predictive power. With the available multiomics data, multilayer networks can now be used in molecular research. In this study, we achieved this with the construction of a bilayer network of DNA methylation sites and RNAs. We applied the network model to five types of tumor data to identify key genes associated with tumors. Compared with the single network, the proposed bilayer network resulted in more tumor-associated DNA methylation sites and genes, which we verified with prognostic and KEGG enrichment analyses.

Gene transferability and sociality do not correlate with gene connectivity

The connectivity of a gene, defined as the number of interactions a gene's product has with other genes' products, is a key characteristic of a gene. In prokaryotes, the complexity hypothesis predicts that genes which undergo more frequent horizontal transfer will be less connected than genes which are only very rarely transferred. We tested the role of horizontal gene transfer, and other potentially important factors, by examining the connectivity of chromosomal and plasmid genes, across 134 diverse prokaryotic species. We found that (i) genes on plasmids were less connected than genes on chromosomes; (ii) connectivity of plasmid genes was not correlated with plasmid mobility; and (iii) the sociality of genes (cooperative or private) was not correlated with gene connectivity.

A review of biomedical datasets relating to drug discovery: a knowledge graph perspective

Drug discovery and development is a complex and costly process. Machine learning approaches are being investigated to help improve the effectiveness and speed of multiple stages of the drug discovery pipeline. Of these, those that use Knowledge Graphs (KG) have promise in many tasks, including drug repurposing, drug toxicity prediction and target gene-disease prioritization. In a drug discovery KG, crucial elements including genes, diseases and drugs are represented as entities, while relationships between them indicate an interaction. However, to construct high-quality KGs, suitable data are required. In this review, we detail publicly available sources suitable for use in constructing drug discovery focused KGs. We aim to help guide machine learning and KG practitioners who are interested in applying new techniques to the drug discovery field, but who may be unfamiliar with the relevant data sources. The datasets are selected via strict criteria, categorized according to the primary type of information contained within and are considered based upon what information could be extracted to build a KG. We then present a comparative analysis of existing public drug discovery KGs and an evaluation of selected motivating case studies from the literature. Additionally, we raise numerous and unique challenges and issues associated with the domain and its datasets, while also highlighting key future research directions. We hope this review will motivate KGs use in solving key and emerging questions in the drug discovery domain.

ELIMINATOR: essentiality analysis using multisystem networks and integer programming

A gene is considered as essential when it is indispensable for cells to grow and replicate in a certain environment. However, gene essentiality is not a structural property but rather a contextual one, which depends on the specific biological conditions affecting the cell. This circumstantial essentiality of genes is what brings the attention of scientist since we can identify genes essential for cancer cells but not essential for healthy cells. This same contextuality makes their identification extremely challenging. Huge experimental efforts such as Project Achilles where the essentiality of thousands of genes is measured together with a plethora of molecular data (transcriptomics, copy number, mutations, etc.) in over one thousand cell lines can shed light on the causality behind the essentiality of a gene in a given environment. Here, we present an in-silico method for the identification of patient-specific essential genes using constraint-based modelling (CBM). Our method expands the ideas behind traditional CBM to accommodate multisystem networks. In essence, it first calculates the minimum number of lowly expressed genes required to be activated by the cell to sustain life as defined by a set of requirements; and second, it performs an exhaustive in-silico gene knockout to find those that lead to the need of activating additional lowly expressed genes. We validated the proposed methodology using a set of 452 cancer cell lines derived from the Cancer Cell Line Encyclopedia where an exhaustive experimental large-scale gene knockout study using CRISPR (Achilles Project) evaluates the impact of each removal. We also show that the integration of different essentiality predictions per gene, what we called Essentiality Congruity Score, reduces the number of false positives. Finally, we explored our method in a breast cancer patient dataset, and our results showed high concordance with previous publications. These findings suggest that identifying genes whose activity is fundamental to sustain cellular life in a patient-specific manner is feasible using in-silico methods. The patient-level gene essentiality predictions can pave the way for precision medicine by identifying potential drug targets whose deletion can induce death in tumour cells.

A deep learning framework for identifying essential proteins based on multiple biological information

Background

Essential Proteins are demonstrated to exert vital functions on cellular processes and are indispensable for the survival and reproduction of the organism. Traditional centrality methods perform poorly on complex protein–protein interaction (PPI) networks. Machine learning approaches based on high-throughput data lack the exploitation of the temporal and spatial dimensions of biological information.

Results

We put forward a deep learning framework to predict essential proteins by integrating features obtained from the PPI network, subcellular localization, and gene expression profiles. In our model, the node2vec method is applied to learn continuous feature representations for proteins in the PPI network, which capture the diversity of connectivity patterns in the network. The concept of depthwise separable convolution is employed on gene expression profiles to extract properties and observe the trends of gene expression over time under different experimental conditions. Subcellular localization information is mapped into a long one-dimensional vector to capture its characteristics. Additionally, we use a sampling method to mitigate the impact of imbalanced learning when training the model. With experiments carried out on the data of Saccharomyces cerevisiae, results show that our model outperforms traditional centrality methods and machine learning methods. Likewise, the comparative experiments have manifested that our process of various biological information is preferable.

Conclusions

Our proposed deep learning framework effectively identifies essential proteins by integrating multiple biological data, proving a broader selection of subcellular localization information significantly improves the results of prediction and depthwise separable convolution implemented on gene expression profiles enhances the performance.

A Novel Sparse Graph-Regularized Singular Value Decomposition Model and Its Application to Genomic Data Analysis

Learning the gene coexpression pattern is a central challenge for high-dimensional gene expression analysis. Recently, sparse singular value decomposition (SVD) has been used to achieve this goal. However, this model ignores the structural information between variables (e.g., a gene network). The typical graph-regularized penalty can be used to incorporate such prior graph information to achieve more accurate discovery and better interpretability. However, the existing approach fails to consider the opposite effect of variables with negative correlations. In this article, we propose a novel sparse graph-regularized SVD model with absolute operator (AGSVD) for high-dimensional gene expression pattern discovery. The key of AGSVD is to impose a novel graph-regularized penalty ( | u|^T L| u| ). However, such a penalty is a nonconvex and nonsmooth function, so it brings new challenges to model solving. We show that the nonconvex problem can be efficiently handled in a convex fashion by adopting an alternating optimization strategy. The simulation results on synthetic data show that our method is more effective than the existing SVD-based ones. In addition, the results on several real gene expression data sets show that the proposed methods can discover more biologically interpretable expression patterns by incorporating the prior gene network.

ProB-Site: Protein Binding Site Prediction Using Local Features

Protein–protein interactions (PPIs) are responsible for various essential biological processes. This information can help develop a new drug against diseases. Various experimental methods have been employed for this purpose; however, their application is limited by their cost and time consumption. Alternatively, computational methods are considered viable means to achieve this crucial task. Various techniques have been explored in the literature using the sequential information of amino acids in a protein sequence, including machine learning and deep learning techniques. The current efficiency of interaction-site prediction still has growth potential. Hence, a deep neural network-based model, ProB-site, is proposed. ProB-site utilizes sequential information of a protein to predict its binding sites. The proposed model uses evolutionary information and predicted structural information extracted from sequential information of proteins, generating three unique feature sets for every amino acid in a protein sequence. Then, these feature sets are fed to their respective sub-CNN architecture to acquire complex features. Finally, the acquired features are concatenated and classified using fully connected layers. This methodology performed better than state-of-the-art techniques because of the selection of the best features and contemplation of local information of each amino acid.

A Dual Ranking Algorithm Based on the Multiplex Network for Heterogeneous Complex Disease Analysis

Identifying biomarkers of heterogeneous complex diseases has always been one of the focuses in medical research. In previous studies, the powerful network propagation methods have been applied to finding marker genes related to specific diseases, but existing methods are mostly based on a single network, which may be greatly affected by the incompleteness of the network and the ignorance of a large amount of information about physical and functional interactions between biological components. Other methods that directly integrate multiple types of interactions into an aggregate network have the risks that different types of data may conflict with each other and the characteristics and topologies of each individual network are lost. Meanwhile, biomarkers used in clinical trials should have the characteristics of small quantity and strong discriminate ability. In this study, we developed a multiplex network-based dual ranking framework (DualRank) for heterogeneous complex disease analysis. We applied the proposed method to heterogeneous complex diseases for diagnosis, prognosis, and classification. The results showed that DualRank outperformed competing methods and could identify biomarkers with the small quantity, great prediction performance (average AUC = 0.818) and biological interpretability.

Discovery of pan-cancer related genes via integrative network analysis

Identification of cancer-related genes is helpful for understanding the pathogenesis of cancer, developing targeted drugs and creating new diagnostic and therapeutic methods. Considering the complexity of the biological laboratory methods, many network-based methods have been proposed to identify cancer-related genes at the global perspective with the increasing availability of high-throughput data. Some studies have focused on the tissue-specific cancer networks. However, cancers from different tissues may share common features, and those methods may ignore the differences and similarities across cancers during the establishment of modeling. In this work, in order to make full use of global information of the network, we first establish the pan-cancer network via differential network algorithm, which not only contains heterogeneous data across multiple cancer types but also contains heterogeneous data between tumor samples and normal samples. Second, the node representation vectors are learned by network embedding. In contrast to ranking analysis-based methods, with the help of integrative network analysis, we transform the cancer-related gene identification problem into a binary classification problem. The final results are obtained via ensemble classification. We further applied these methods to the most commonly used gene expression data involving six tissue-specific cancer types. As a result, an integrative pan-cancer network and several biologically meaningful results were obtained. As examples, nine genes were ultimately identified as potential pan-cancer-related genes. Most of these genes have been reported in published studies, thus showing our method's potential for application in identifying driver gene candidates for further biological experimental verification.

Inference of Molecular Regulatory Systems Using Statistical Path-Consistency Algorithm

One of the key challenges in systems biology and molecular sciences is how to infer regulatory relationships between genes and proteins using high-throughout omics datasets. Although a wide range of methods have been designed to reverse engineer the regulatory networks, recent studies show that the inferred network may depend on the variable order in the dataset. In this work, we develop a new algorithm, called the statistical path-consistency algorithm (SPCA), to solve the problem of the dependence of variable order. This method generates a number of different variable orders using random samples, and then infers a network by using the path-consistent algorithm based on each variable order. We propose measures to determine the edge weights using the corresponding edge weights in the inferred networks, and choose the edges with the largest weights as the putative regulations between genes or proteins. The developed method is rigorously assessed by the six benchmark networks in DREAM challenges, the mitogen-activated protein (MAP) kinase pathway, and a cancer-specific gene regulatory network. The inferred networks are compared with those obtained by using two up-to-date inference methods. The accuracy of the inferred networks shows that the developed method is effective for discovering molecular regulatory systems.

Disclosing Potential Key Genes, Therapeutic Targets and Agents for Non-Small Cell Lung Cancer: Evidence from Integrative Bioinformatics Analysis

Non-small-cell lung cancer (NSCLC) is considered as one of the malignant cancers that causes premature death. The present study aimed to identify a few potential novel genes highlighting their functions, pathways, and regulators for diagnosis, prognosis, and therapies of NSCLC by using the integrated bioinformatics approaches. At first, we picked out 1943 DEGs between NSCLC and control samples by using the statistical LIMMA approach. Then we selected 11 DEGs (CDK1, EGFR, FYN, UBC, MYC, CCNB1, FOS, RHOB, CDC6, CDC20, and CHEK1) as the hub-DEGs (potential key genes) by the protein–protein interaction network analysis of DEGs. The DEGs and hub-DEGs regulatory network analysis commonly revealed four transcription factors (FOXC1, GATA2, YY1, and NFIC) and five miRNAs (miR-335-5p, miR-26b-5p, miR-92a-3p, miR-155-5p, and miR-16-5p) as the key transcriptional and post-transcriptional regulators of DEGs as well as hub-DEGs. We also disclosed the pathogenetic processes of NSCLC by investigating the biological processes, molecular function, cellular components, and KEGG pathways of DEGs. The multivariate survival probability curves based on the expression of hub-DEGs in the SurvExpress web-tool and database showed the significant differences between the low- and high-risk groups, which indicates strong prognostic power of hub-DEGs. Then, we explored top-ranked 5-hub-DEGs-guided repurposable drugs based on the Connectivity Map (CMap) database. Out of the selected drugs, we validated six FDA-approved launched drugs (Dinaciclib, Afatinib, Icotinib, Bosutinib, Dasatinib, and TWS-119) by molecular docking interaction analysis with the respective target proteins for the treatment against NSCLC. The detected therapeutic targets and repurposable drugs require further attention by experimental studies to establish them as potential biomarkers for precision medicine in NSCLC treatment.

EPGAT: Gene Essentiality Prediction With Graph Attention Networks

Identifying essential genes and proteins is a critical step towards a better understanding of human biology and pathology. Computational approaches helped to mitigate experimental constraints by exploring machine learning (ML) methods and the correlation of essentiality with biological information, especially protein-protein interaction (PPI) networks, to predict essential genes. Nonetheless, their performance is still limited, as network-based centralities are not exclusive proxies of essentiality, and traditional ML methods are unable to learn from non-euclidean domains such as graphs. Given these limitations, we proposed EPGAT, an approach for Essentiality Prediction based on Graph Attention Networks (GATs), which are attention-based Graph Neural Networks (GNNs), operating on graph-structured data. Our model directly learns gene essentiality patterns from PPI networks, integrating additional evidence from multiomics data encoded as node attributes. We benchmarked EPGAT for four organisms, including humans, accurately predicting gene essentiality with ROC AUC score ranging from 0.78 to 0.97. Our model significantly outperformed network-based and shallow ML-based methods and achieved a very competitive performance against the state-of-the-art node2vec embedding method. Notably, EPGAT was the most robust approach in scenarios with limited and imbalanced training data. Thus, the proposed approach offers a powerful and effective way to identify essential genes and proteins.

EGRET: edge aggregated graph attention networks and transfer learning improve protein-protein interaction site prediction

MOTIVATION

Protein-protein interactions (PPIs) are central to most biological processes. However, reliable identification of PPI sites using conventional experimental methods is slow and expensive. Therefore, great efforts are being put into computational methods to identify PPI sites.

RESULTS

We present Edge Aggregated GRaph Attention NETwork (EGRET), a highly accurate deep learning-based method for PPI site prediction, where we have used an edge aggregated graph attention network to effectively leverage the structural information. We, for the first time, have used transfer learning in PPI site prediction. Our proposed edge aggregated network, together with transfer learning, has achieved notable improvement over the best alternate methods. Furthermore, we systematically investigated EGRET's network behavior to provide insights about the causes of its decisions.

AVAILABILITY

EGRET is freely available as an open source project at https://github.com/Sazan-Mahbub/EGRET.

CONTACT

shams_bayzid@cse.buet.ac.bd.

Significance-Based Essential Protein Discovery

The identification of essential proteins is an important problem in bioinformatics. During the past decades, many centrality measures and algorithms have been proposed to address this issue. However, existing methods still deserve the following drawbacks: (1) the lack of a context-free and readily interpretable quantification of their centrality values; (2) the difficulty of specifying a proper threshold for their centrality values; (3) the incapability of controlling the quality of reported essential proteins in a statistically sound manner. To overcome the limitations of existing solutions, we tackle the essential protein discovery problem from a significance testing perspective. More precisely, the essential protein discovery problem is formulated as a multiple hypothesis testing problem, where the null hypothesis is that each protein is not an essential protein. To quantify the statistical significance of each protein, we present a p-value calculation method in which both the degree and the local clustering coefficient are used as the test statistic and the Erdös-Rényi model is employed as the random graph model. After calculating the p-value for each protein, the false discovery rate is used as the error rate in the multiple testing correction procedure. Our significance-based essential protein discovery method is named as SigEP, which is tested on both simulated networks and real PPI networks. The experimental results show that our method is able to achieve better performance than those competing algorithms.

PangenomeNet: a pan-genome-based network reveals functional modules on antimicrobial resistome for Escherichia coli strains

BACKGROUND

Discerning genes crucial to antimicrobial resistance (AMR) mechanisms is becoming more and more important to accurately and swiftly identify AMR pathogenic strains. Pangenome-wide association studies (e.g. Scoary) identified numerous putative AMR genes. However, only a tiny proportion of the putative resistance genes are annotated by AMR databases or Gene Ontology. In addition, many putative resistance genes are of unknown function (termed hypothetical proteins). An annotation tool is crucially needed in order to reveal the functional organization of the resistome and expand our knowledge of the AMR gene repertoire.

RESULTS

We developed an approach (PangenomeNet) for building co-functional networks from pan-genomes to infer functions for hypothetical genes. Using Escherichia coli as an example, we demonstrated that it is possible to build co-functional network from its pan-genome using co-inheritance, domain-sharing, and protein-protein-interaction information. The investigation of the network revealed that it fits the characteristics of biological networks and can be used for functional inferences. The subgraph consisting of putative meropenem resistance genes consists of clusters of stress response genes and resistance gene acquisition pathways. Resistome subgraphs also demonstrate drug-specific AMR genes such as beta-lactamase, as well as functional roles shared among multiple classes of drugs, mostly in the stress-related pathways.

CONCLUSIONS

By demonstrating the idea of pan-genome-based co-functional network on the E. coli species, we showed that the network can infer functional roles of the genes, including those without functional annotations, and provides holistic views on the putative antimicrobial resistomes. We hope that the pan-genome network idea can help formulate hypothesis for targeted experimental works.

Essential Protein Recognition via Community Significance

Essential protein plays a vital role in understanding the cellular life. With the advance in high-throughput technologies, a number of protein-protein interaction (PPI) networks have been constructed such that essential proteins can be identified from a system biology perspective. Although a series of network-based essential protein discovery methods have been proposed, these existing methods still have some drawbacks. Recently, it has been shown that the significance-based method SigEP is promising on overcoming the defects that are inherent in currently available essential protein identification methods. However, the SigEP method is developed under the unrealistic Erdös-Rényi (E-R) model and its time complexity is very high. Hence, we propose a new significance-based essential protein recognition method named EPCS in which the essential protein discovery problem is formulated as a community significance testing problem. Experimental results on four PPI networks show that EPCS performs better than nine state-of-the-art essential protein identification methods and the only significance-based essential protein identification method SigEP.

FUNMarker: Fusion Network-Based Method to Identify Prognostic and Heterogeneous Breast Cancer Biomarkers

Breast cancer is a heterogeneous disease with many clinically distinguishable molecular subtypes each corresponding to a cluster of patients. Identification of prognostic and heterogeneous biomarkers for breast cancer is to detect cluster-specific gene biomarkers which can be used for accurate survival prediction of breast cancer outcomes. In this study, we proposed a FUsion Network-based method (FUNMarker) to identify prognostic and heterogeneous breast cancer biomarkers by considering the heterogeneity of patient samples and biological information from multiple sources. To reduce the affect of heterogeneity of patients, samples were first clustered using the K-means algorithm based on the principal components of gene expression. For each cluster, to comprehensively evaluate the influence of genes on breast cancer, genes were weighted from three aspects: biological function, prognostic ability and correlation with known disease genes. Then they were ranked via a label propagation model on a fusion network that combined physical protein interactions from seven types of networks and thus could reduce the impact of incompleteness of interactome. We compared FUNMarker with three state-of-the-art methods and the results showed that biomarkers identified by FUNMarker were biological interpretable and had stronger discriminative power than the existing methods in differentiating patients with different prognostic outcomes.

A graph-traversal approach to identify influential nodes in a network

Influential node identification plays a significant role in understanding network structure and functions. Here we propose a general method for detecting influential nodes in a graph-traversal framework. We evaluate the influence of each node by constructing a breadth-first search (BFS) tree in which the target node is the root node. From the BFS tree, we generate a curve in which the x axis is the level number and the y axis is the cumulative scores of all nodes visited so far. We use the area under the curve value as the final influence score of the target node. Experimental results on various networks across different domains demonstrate that our method can be significantly superior to widely used centrality measures on the task of influential node detection.

•

We propose an influential node detection method, TARank, in a graph-traversal framework

•

We evaluate the influence of each node by constructing a breadth-first search tree

•

TARank is capable of enhancing existing centrality measures

•

TARank can yield new, yet effective, centrality measures as well

The discovery of influential nodes is a fundamental research issue in network science. To quantify the influence of each node in a network, various methods have been presented in the literature. To the best of our knowledge, no previous research efforts address the influential node identification problem from a graph-traversal perspective. To fulfill this void, we propose the TARank method that integrates the information collected from the breadth-first search tree to identify influential nodes. The formulation under the graph-traversal framework opens the door to a fundamentally new type of method of influential node identification. In the future, more effective recognition methods can be expected to be constructed based on this general framework. Since empirical studies have validated the effectiveness of TARank, it would be plausible to employ this method in different applications to reveal new findings.

Protein Integrated Network Analysis to Reveal Potential Drug Targets Against Extended Drug-Resistant Mycobacterium tuberculosis XDR1219

The reconstruction and analysis of the protein-protein interaction (PPI) network is a powerful approach to understand the complex biological and molecular functions in normal and disease states of the cell. The interactome of most organisms is largely unidentified except some model organisms. The current study focused on the construction of PPI network for the human pathogen Mycobacterium tuberculosis (MTB)-resistant strain XDR1219 using computational methods. In this work, a bioinformatics approach was employed to reveal potential drug targets. The pipeline adopted the combination of an extensive integrated network analysis that led to identify 22 key proteins involved in drug resistance, resistant metabolic pathways, virulence, pathogenesis and persistency of the infection. The MTB XDR1219 interactome consists of 11,383 non-redundant PPIs among 1499 proteins covering 38% of the entire MTB XDR1219 proteome. The overall quality of the network was assessed and topological parameters of the PPI were calculated. The predicted interactions were functionally annotated and their relevance was assessed with the functional similarity. The study attempts to present the interactome of previously unidentified MTB XDR1219 and revealed potential drug targets that can be further explored by scientific community.

NetAUC: A network-based multi-biomarker identification method by AUC optimization

Complex diseases are caused by a variety of factors, and their diagnosis, treatment and prognosis are usually difficult. Proteins play an indispensable role in living organisms and perform specific biological functions by interacting with other proteins or biomolecules, their dysfunction may lead to diseases, it is a natural way to mine disease-related biomarkers from protein-protein interaction network. AUC, the area under the receiver operating characteristics (ROC) curve, is regarded as a gold standard to evaluate the effectiveness of a binary classifier, which measures the classification ability of an algorithm under arbitrary distribution or any misclassification cost. In this study, we have proposed a network-based multi-biomarker identification method by AUC optimization (NetAUC), which integrates gene expression and the network information to identify biomarkers for the complex disease analysis. The main purpose is to optimize two objectives simultaneously: maximizing AUC and minimizing the number of selected features. We have applied NetAUC to two types of disease analysis: 1) prognosis of breast cancer, 2) classification of similar diseases. The results show that NetAUC can identify a small panel of disease-related biomarkers which have the powerful classification ability and the functional interpretability.

Integrated Inference of Asymmetric Protein Interaction Networks Using Dynamic Model and Individual Patient Proteomics Data

Recent advances in experimental biology studies have produced large amount of molecular activity data. In particular, individual patient data provide non-time series information for the molecular activities in disease conditions. The challenge is how to design effective algorithms to infer regulatory networks using the individual patient datasets and consequently address the issue of network symmetry. This work is aimed at developing an efficient pipeline to reverse-engineer regulatory networks based on the individual patient proteomic data. The first step uses the SCOUT algorithm to infer the pseudo-time trajectory of individual patients. Then the path-consistent method with part mutual information is used to construct a static network that contains the potential protein interactions. To address the issue of network symmetry in terms of undirected symmetric network, a dynamic model of ordinary differential equations is used to further remove false interactions to derive asymmetric networks. In this work a dataset from triple-negative breast cancer patients is used to develop a protein-protein interaction network with 15 proteins.

Semi-Supervised Learning Using Hierarchical Mixture Models: Gene Essentiality Case Study

Integrating gene-level data is useful for predicting the role of genes in biological processes. This problem has typically focused on supervised classification, which requires large training sets of positive and negative examples. However, training data sets that are too small for supervised approaches can still provide valuable information. We describe a hierarchical mixture model that uses limited positively labeled gene training data for semi-supervised learning. We focus on the problem of predicting essential genes, where a gene is required for the survival of an organism under particular conditions. We applied cross-validation and found that the inclusion of positively labeled samples in a semi-supervised learning framework with the hierarchical mixture model improves the detection of essential genes compared to unsupervised, supervised, and other semi-supervised approaches. There was also improved prediction performance when genes are incorrectly assumed to be non-essential. Our comparisons indicate that the incorporation of even small amounts of existing knowledge improves the accuracy of prediction and decreases variability in predictions. Although we focused on gene essentiality, the hierarchical mixture model and semi-supervised framework is standard for problems focused on prediction of genes or other features, with multiple data types characterizing the feature, and a small set of positive labels.

On the relation of gene essentiality to intron structure: a computational and deep learning approach

Essential genes have been studied by copy number variants and deletions, both associated with introns. The premise of our work is that introns of essential genes have distinct characteristic properties. We provide support for this by training a deep learning model and demonstrating that introns alone can be used to classify essentiality. The model, limited to first introns, performs at an increased level, implicating first introns in essentiality. We identify unique properties of introns of essential genes, finding that their structure protects against deletion and intron-loss events, especially centered on the first intron. We show that GC density is increased in the first introns of essential genes, allowing for increased enhancer activity, protection against deletions, and improved splice site recognition. We find that first introns of essential genes are of remarkably smaller size than their nonessential counterparts, and to protect against common 3' end deletion events, essential genes carry an increased number of (smaller) introns. To demonstrate the importance of the seven features we identified, we train a feature-based model using only these features and achieve high performance.

NIDM: network impulsive dynamics on multiplex biological network for disease-gene prediction

The prediction of genes related to diseases is important to the study of the diseases due to high cost and time consumption of biological experiments. Network propagation is a popular strategy for disease-gene prediction. However, existing methods focus on the stable solution of dynamics while ignoring the useful information hidden in the dynamical process, and it is still a challenge to make use of multiple types of physical/functional relationships between proteins/genes to effectively predict disease-related genes. Therefore, we proposed a framework of network impulsive dynamics on multiplex biological network (NIDM) to predict disease-related genes, along with four variants of NIDM models and four kinds of impulsive dynamical signatures (IDSs). NIDM is to identify disease-related genes by mining the dynamical responses of nodes to impulsive signals being exerted at specific nodes. By a series of experimental evaluations in various types of biological networks, we confirmed the advantage of multiplex network and the important roles of functional associations in disease-gene prediction, demonstrated superior performance of NIDM compared with four types of network-based algorithms and then gave the effective recommendations of NIDM models and IDS signatures. To facilitate the prioritization and analysis of (candidate) genes associated to specific diseases, we developed a user-friendly web server, which provides three kinds of filtering patterns for genes, network visualization, enrichment analysis and a wealth of external links (http://bioinformatics.csu.edu.cn/DGP/NID.jsp). NIDM is a protocol for disease-gene prediction integrating different types of biological networks, which may become a very useful computational tool for the study of disease-related genes.

Anticancer peptides prediction with deep representation learning features

Anticancer peptides constitute one of the most promising therapeutic agents for combating common human cancers. Using wet experiments to verify whether a peptide displays anticancer characteristics is time-consuming and costly. Hence, in this study, we proposed a computational method named identify anticancer peptides via deep representation learning features (iACP-DRLF) using light gradient boosting machine algorithm and deep representation learning features. Two kinds of sequence embedding technologies were used, namely soft symmetric alignment embedding and unified representation (UniRep) embedding, both of which involved deep neural network models based on long short-term memory networks and their derived networks. The results showed that the use of deep representation learning features greatly improved the capability of the models to discriminate anticancer peptides from other peptides. Also, UMAP (uniform manifold approximation and projection for dimension reduction) and SHAP (shapley additive explanations) analysis proved that UniRep have an advantage over other features for anticancer peptide identification. The python script and pretrained models could be downloaded from https://github.com/zhibinlv/iACP-DRLF or from http://public.aibiochem.net/iACP-DRLF/.

Inference of Gene Regulatory Network from Single-Cell Transcriptomic Data Using pySCENIC

With the advent of recent next-generation sequencing (NGS) technologies in genomics, transcriptomics, and epigenomics, profiling single-cell sequencing became possible. The single-cell RNA sequencing (scRNA-seq) is widely used to characterize diverse cell populations and ascertain cell type-specific regulatory mechanisms. The gene regulatory network (GRN) mainly consists of genes and their regulators-transcription factors (TF). Here, we describe the lightning-fast Python implementation of the SCENIC (Single-Cell reEgulatory Network Inference and Clustering) pipeline called pySCENIC. Using single-cell RNA-seq data, it maps TFs onto gene regulatory networks and integrates various cell types to infer cell-specific GRNs. There are two fast and efficient GRN inference algorithms, GRNBoost2 and GENIE3, optionally available with pySCENIC. The pipeline has three steps: (1) identification of potential TF targets based on co-expression; (2) TF-motif enrichment analysis to identify the direct targets (regulons); and (3) scoring the activity of regulons (or other gene sets) on single cell types.

An Ensemble Method to Reconstruct Gene Regulatory Networks Based on Multivariate Adaptive Regression Splines

Gene regulatory networks (GRNs) play a key role in biological processes. However, GRNs are diverse under different biological conditions. Reconstructing gene regulatory networks (GRNs) from gene expression has become an important opportunity and challenge in the past decades. Although there are a lot of existing methods to infer the topology of GRNs, such as mutual information, random forest, and partial least squares, the accuracy is still low due to the noise and high dimension of the expression data. In this paper, we introduce an ensemble Multivariate Adaptive Regression Splines (MARS) based method to reconstruct the directed GRNs from multifactorial gene expression data, called PBMarsNet. PBMarsNet incorporates part mutual information (PMI) to pre-weight the candidate regulatory genes and then uses MARS to detect the nonlinear regulatory links. Moreover, we apply bootstrap to run the MARS multiple times and average the outputs of each MARS as the final score of regulatory links. The results on DREAM4 challenge and DREAM5 challenge datasets show PBMarsNet has a superior performance and generalization over other state-of-the-art methods.

A Deep Learning Framework for Identifying Essential Proteins by Integrating Multiple Types of Biological Information

Computational methods including centrality and machine learning-based methods have been proposed to identify essential proteins for understanding the minimum requirements of the survival and evolution of a cell. In centrality methods, researchers are required to design a score function which is based on prior knowledge, yet is usually not sufficient to capture the complexity of biological information. In machine learning-based methods, some selected biological features cannot represent the complete properties of biological information as they lack a computational framework to automatically select features. To tackle these problems, we propose a deep learning framework to automatically learn biological features without prior knowledge. We use node2vec technique to automatically learn a richer representation of protein-protein interaction (PPI) network topologies than a score function. Bidirectional long short term memory cells are applied to capture non-local relationships in gene expression data. For subcellular localization information, we exploit a high dimensional indicator vector to characterize their feature. To evaluate the performance of our method, we tested it on PPI network of S. cerevisiae. Our experimental results demonstrate that the performance of our method is better than traditional centrality methods and is superior to existing machine learning-based methods. To explore which of the three types of biological information is the most vital element, we conduct an ablation study by removing each component in turn. Our results show that the PPI network embedding contributes most to the improvement. In addition, gene expression profiles and subcellular localization information are also helpful to improve the performance in identification of essential proteins.

Essential gene prediction using limited gene essentiality information-An integrative semi-supervised machine learning strategy

Essential gene prediction helps to find minimal genes indispensable for the survival of any organism. Machine learning (ML) algorithms have been useful for the prediction of gene essentiality. However, currently available ML pipelines perform poorly for organisms with limited experimental data. The objective is the development of a new ML pipeline to help in the annotation of essential genes of less explored disease-causing organisms for which minimal experimental data is available. The proposed strategy combines unsupervised feature selection technique, dimension reduction using the Kamada-Kawai algorithm, and semi-supervised ML algorithm employing Laplacian Support Vector Machine (LapSVM) for prediction of essential and non-essential genes from genome-scale metabolic networks using very limited labeled dataset. A novel scoring technique, Semi-Supervised Model Selection Score, equivalent to area under the ROC curve (auROC), has been proposed for the selection of the best model when supervised performance metrics calculation is difficult due to lack of data. The unsupervised feature selection followed by dimension reduction helped to observe a distinct circular pattern in the clustering of essential and non-essential genes. LapSVM then created a curve that dissected this circle for the classification and prediction of essential genes with high accuracy (auROC > 0.85) even with 1% labeled data for model training. After successful validation of this ML pipeline on both Eukaryotes and Prokaryotes that show high accuracy even when the labeled dataset is very limited, this strategy is used for the prediction of essential genes of organisms with inadequate experimentally known data, such as Leishmania sp. Using a graph-based semi-supervised machine learning scheme, a novel integrative approach has been proposed for essential gene prediction that shows universality in application to both Prokaryotes and Eukaryotes with limited labeled data. The essential genes predicted using the pipeline provide an important lead for the prediction of gene essentiality and identification of novel therapeutic targets for antibiotic and vaccine development against disease-causing parasites.

Robustness and lethality in multilayer biological molecular networks

Robustness is a prominent feature of most biological systems. Most previous related studies have been focused on homogeneous molecular networks. Here we propose a comprehensive framework for understanding how the interactions between genes, proteins and metabolites contribute to the determinants of robustness in a heterogeneous biological network. We integrate heterogeneous sources of data to construct a multilayer interaction network composed of a gene regulatory layer, a protein-protein interaction layer, and a metabolic layer. We design a simulated perturbation process to characterize the contribution of each gene to the overall system's robustness, and find that influential genes are enriched in essential and cancer genes. We show that the proposed mechanism predicts a higher vulnerability of the metabolic layer to perturbations applied to genes associated with metabolic diseases. Furthermore, we find that the real network is comparably or more robust than expected in multiple random realizations. Finally, we analytically derive the expected robustness of multilayer biological networks starting from the degree distributions within and between layers. These results provide insights into the non-trivial dynamics occurring in the cell after a genetic perturbation is applied, confirming the importance of including the coupling between different layers of interaction in models of complex biological systems.

IHP-PING-generating integrated human protein-protein interaction networks on-the-fly

Advances in high-throughput sequencing technologies have resulted in an exponential growth of publicly accessible biological datasets. In the 'big data' driven 'post-genomic' context, much work is being done to explore human protein-protein interactions (PPIs) for a systems level based analysis to uncover useful signals and gain more insights to advance current knowledge and answer specific biological and health questions. These PPIs are experimentally or computationally predicted, stored in different online databases and some of PPI resources are updated regularly. As with many biological datasets, such regular updates continuously render older PPI datasets potentially outdated. Moreover, while many of these interactions are shared between these online resources, each resource includes its own identified PPIs and none of these databases exhaustively contains all existing human PPI maps. In this context, it is essential to enable the integration of or combining interaction datasets from different resources, to generate a PPI map with increased coverage and confidence. To allow researchers to produce an integrated human PPI datasets in real-time, we introduce the integrated human protein-protein interaction network generator (IHP-PING) tool. IHP-PING is a flexible python package which generates a human PPI network from freely available online resources. This tool extracts and integrates heterogeneous PPI datasets to generate a unified PPI network, which is stored locally for further applications.

Supervised learning is an accurate method for network-based gene classification

Background

Assigning every human gene to specific functions, diseases and traits is a grand challenge in modern genetics. Key to addressing this challenge are computational methods, such as supervised learning and label propagation, that can leverage molecular interaction networks to predict gene attributes. In spite of being a popular machine-learning technique across fields, supervised learning has been applied only in a few network-based studies for predicting pathway-, phenotype- or disease-associated genes. It is unknown how supervised learning broadly performs across different networks and diverse gene classification tasks, and how it compares to label propagation, the widely benchmarked canonical approach for this problem.

Results

In this study, we present a comprehensive benchmarking of supervised learning for network-based gene classification, evaluating this approach and a classic label propagation technique on hundreds of diverse prediction tasks and multiple networks using stringent evaluation schemes. We demonstrate that supervised learning on a gene’s full network connectivity outperforms label propagaton and achieves high prediction accuracy by efficiently capturing local network properties, rivaling label propagation’s appeal for naturally using network topology. We further show that supervised learning on the full network is also superior to learning on node embeddings (derived using node2vec), an increasingly popular approach for concisely representing network connectivity. These results show that supervised learning is an accurate approach for prioritizing genes associated with diverse functions, diseases and traits and should be considered a staple of network-based gene classification workflows.

Availability and implementation

The datasets and the code used to reproduce the results and add new gene classification methods have been made freely available.

Contact

arjun@msu.edu

Supplementary information

Supplementary data are available at Bioinformatics online.

Comprehensive host-pathogen protein-protein interaction network analysis

BACKGROUND

Infectious diseases are a cruel assassin with millions of victims around the world each year. Understanding infectious mechanism of viruses is indispensable for their inhibition. One of the best ways of unveiling this mechanism is to investigate the host-pathogen protein-protein interaction network. In this paper we try to disclose many properties of this network. We focus on human as host and integrate experimentally 32,859 interaction between human proteins and virus proteins from several databases. We investigate different properties of human proteins targeted by virus proteins and find that most of them have a considerable high centrality scores in human intra protein-protein interaction network. Investigating human proteins network properties which are targeted by different virus proteins can help us to design multipurpose drugs.

RESULTS

As host-pathogen protein-protein interaction network is a bipartite network and centrality measures for this type of networks are scarce, we proposed seven new centrality measures for analyzing bipartite networks. Applying them to different virus strains reveals unrandomness of attack strategies of virus proteins which could help us in drug design hence elevating the quality of life. They could also be used in detecting host essential proteins. Essential proteins are those whose functions are critical for survival of its host. One of the proposed centralities named diversity of predators, outperforms the other existing centralities in terms of detecting essential proteins and could be used as an optimal essential proteins' marker.

CONCLUSIONS

Different centralities were applied to analyze human protein-protein interaction network and to detect characteristics of human proteins targeted by virus proteins. Moreover, seven new centralities were proposed to analyze host-pathogen protein-protein interaction network and to detect pathogens' favorite host protein victims. Comparing different centralities in detecting essential proteins reveals that diversity of predator (one of the proposed centralities) is the best essential protein marker.

DeepHE: Accurately predicting human essential genes based on deep learning

Accurately predicting essential genes using computational methods can greatly reduce the effort in finding them via wet experiments at both time and resource scales, and further accelerate the process of drug discovery. Several computational methods have been proposed for predicting essential genes in model organisms by integrating multiple biological data sources either via centrality measures or machine learning based methods. However, the methods aiming to predict human essential genes are still limited and the performance still need improve. In addition, most of the machine learning based essential gene prediction methods are lack of skills to handle the imbalanced learning issue inherent in the essential gene prediction problem, which might be one factor affecting their performance. We propose a deep learning based method, DeepHE, to predict human essential genes by integrating features derived from sequence data and protein-protein interaction (PPI) network. A deep learning based network embedding method is utilized to automatically learn features from PPI network. In addition, 89 sequence features were derived from DNA sequence and protein sequence for each gene. These two types of features are integrated to train a multilayer neural network. A cost-sensitive technique is used to address the imbalanced learning problem when training the deep neural network. The experimental results for predicting human essential genes show that our proposed method, DeepHE, can accurately predict human gene essentiality with an average performance of AUC higher than 94%, the area under precision-recall curve (AP) higher than 90%, and the accuracy higher than 90%. We also compare DeepHE with several widely used traditional machine learning models (SVM, Naïve Bayes, Random Forest, and Adaboost) using the same features and utilizing the same cost-sensitive technique to against the imbalanced learning issue. The experimental results show that DeepHE significantly outperforms the compared machine learning models. We have demonstrated that human essential genes can be accurately predicted by designing effective machine learning algorithm and integrating representative features captured from available biological data. The proposed deep learning framework is effective for such task.