Deep variational graph autoencoders for novel host-directed therapy options against COVID-19.

The COVID-19 pandemic has been keeping asking urgent questions with respect to therapeutic options. Existing drugs that can be repurposed promise rapid implementation in practice because of their prior approval. Conceivably, there is still room for substantial improvement, because most advanced artificial intelligence techniques for screening drug repositories have not been exploited so far. We construct a comprehensive network by combining year-long curated drug-protein/protein-protein interaction data on the one hand, and most recent SARS-CoV-2 protein interaction data on the other hand. We learn the structure of the resulting encompassing molecular interaction network and predict missing links using variational graph autoencoders (VGAEs), as a most advanced deep learning technique that has not been explored so far. We focus on hitherto unknown links between drugs and human proteins that play key roles in the replication cycle of SARS-CoV-2. Thereby, we establish novel host-directed therapy (HDT) options whose utmost plausibility is confirmed by realistic simulations. As a consequence, many of the predicted links are likely to be crucial for the virus to thrive on the one hand, and can be targeted with existing drugs on the other hand.

Systematic mining of patterns of polysubstance use in a nationwide population survey.

OBJECTIVES: To identify patterns of association and transition in polysubstance use based on National Survey of Drug Use and Health (NSDUH) in the United States. METHODS: We developed a new computational platform for PolySubstance Use data Mining for Associations and Transitions (PSUMAnT). It is based on the computation of weighted support, a measure of popularity, for the use of every combination of one or more substances, termed as a drugset, over a period of 5 decades (1965-2014) based on NSDUH data. It uses an efficient bitstring representation with exact and approximate string matching capabilities to search for patterns of association between drugsets and demographics of user groups at different time-intervals. Moreover, it introduces a quantitative definition of a rule of transition between pairs of substances used within a given time-interval, and provides a function for mining them. RESULTS: We identified the frequent drugsets from individual substance use database, and determined their representation among different demographic groups at different intervals. An interesting pattern of use of pain relievers and tranquilizers was detected for the age-group of 26-34 years. In addition, transition rules for heroin use in the last decade (2004-2015) of the given data were mined. CONCLUSIONS: Computation of weighted supports over time for every possible combination of substances in the survey, and their association with specific user groups, allows PSUMAnT to generate and test novel, interesting hypotheses in polysubstance use. PSUMAnT can be used for mining combinations of substances used among diverse demographic groups including those that have received less attention in this problem.

LSH-GAN enables in-silico generation of cells for small sample high dimensional scRNA-seq data.

A fundamental problem of downstream analysis of scRNA-seq data is the unavailability of enough cell samples compare to the feature size. This is mostly due to the budgetary constraint of single cell experiments or simply because of the small number of available patient samples. Here, we present an improved version of generative adversarial network (GAN) called LSH-GAN to address this issue by producing new realistic cell samples. We update the training procedure of the generator of GAN using locality sensitive hashing which speeds up the sample generation, thus maintains the feasibility of applying the standard procedures of downstream analysis. LSH-GAN outperforms the benchmarks for realistic generation of quality cell samples. Experimental results show that generated samples of LSH-GAN improves the performance of the downstream analysis such as feature (gene) selection and cell clustering. Overall, LSH-GAN therefore addressed the key challenges of small sample scRNA-seq data analysis.

A Regularized Multi-Task Learning Approach for Cell Type Detection in Single-Cell RNA Sequencing Data.

Cell type prediction is one of the most challenging goals in single-cell RNA sequencing (scRNA-seq) data. Existing methods use unsupervised learning to identify signature genes in each cluster, followed by a literature survey to look up those genes for assigning cell types. However, finding potential marker genes in each cluster is cumbersome, which impedes the systematic analysis of single-cell RNA sequencing data. To address this challenge, we proposed a framework based on regularized multi-task learning (RMTL) that enables us to simultaneously learn the subpopulation associated with a particular cell type. Learning the structure of subpopulations is treated as a separate task in the multi-task learner. Regularization is used to modulate the multi-task model (e.g., W 1, W 2, … W t ) jointly, according to the specific prior. For validating our model, we trained it with reference data constructed from a single-cell RNA sequencing experiment and applied it to a query dataset. We also predicted completely independent data (the query dataset) from the reference data which are used for training. We have checked the efficacy of the proposed method by comparing it with other state-of-the-art techniques well known for cell type detection. Results revealed that the proposed method performed accurately in detecting the cell type in scRNA-seq data and thus can be utilized as a useful tool in the scRNA-seq pipeline.

A copula based topology preserving graph convolution network for clustering of single-cell RNA-seq data.

Annotation of cells in single-cell clustering requires a homogeneous grouping of cell populations. There are various issues in single cell sequencing that effect homogeneous grouping (clustering) of cells, such as small amount of starting RNA, limited per-cell sequenced reads, cell-to-cell variability due to cell-cycle, cellular morphology, and variable reagent concentrations. Moreover, single cell data is susceptible to technical noise, which affects the quality of genes (or features) selected/extracted prior to clustering. Here we introduce sc-CGconv (copula based graph convolution network for single clustering), a stepwise robust unsupervised feature extraction and clustering approach that formulates and aggregates cell-cell relationships using copula correlation (Ccor), followed by a graph convolution network based clustering approach. sc-CGconv formulates a cell-cell graph using Ccor that is learned by a graph-based artificial intelligence model, graph convolution network. The learned representation (low dimensional embedding) is utilized for cell clustering. sc-CGconv features the following advantages. a. sc-CGconv works with substantially smaller sample sizes to identify homogeneous clusters. b. sc-CGconv can model the expression co-variability of a large number of genes, thereby outperforming state-of-the-art gene selection/extraction methods for clustering. c. sc-CGconv preserves the cell-to-cell variability within the selected gene set by constructing a cell-cell graph through copula correlation measure. d. sc-CGconv provides a topology-preserving embedding of cells in low dimensional space.

sc-REnF: An entropy guided robust feature selection for single-cell RNA-seq data.

Annotation of cells in single-cell clustering requires a homogeneous grouping of cell populations. Since single-cell data are susceptible to technical noise, the quality of genes selected prior to clustering is of crucial importance in the preliminary steps of downstream analysis. Therefore, interest in robust gene selection has gained considerable attention in recent years. We introduce sc-REnF [robust entropy based feature (gene) selection method], aiming to leverage the advantages of $R{\prime}{e}nyi$ and $Tsallis$ entropies in gene selection for single cell clustering. Experiments demonstrate that with tuned parameter ($q$), $R{\prime}{e}nyi$ and $Tsallis$ entropies select genes that improved the clustering results significantly, over the other competing methods. sc-REnF can capture relevancy and redundancy among the features of noisy data extremely well due to its robust objective function. Moreover, the selected features/genes can able to determine the unknown cells with a high accuracy. Finally, sc-REnF yields good clustering performance in small sample, large feature scRNA-seq data. Availability: The sc-REnF is available at https://github.com/Snehalikalall/sc-REnF.

Pan-cancer classification by regularized multi-task learning.

Classifying pan-cancer samples using gene expression patterns is a crucial challenge for the accurate diagnosis and treatment of cancer patients. Machine learning algorithms have been considered proven tools to perform downstream analysis and capture the deviations in gene expression patterns across diversified diseases. In our present work, we have developed PC-RMTL, a pan-cancer classification model using regularized multi-task learning (RMTL) for classifying 21 cancer types and adjacent normal samples using RNASeq data obtained from TCGA. PC-RMTL is observed to outperform when compared with five state-of-the-art classification algorithms, viz. SVM with the linear kernel (SVM-Lin), SVM with radial basis function kernel (SVM-RBF), random forest (RF), k-nearest neighbours (kNN), and decision trees (DT). The PC-RMTL achieves 96.07% accuracy and 95.80% MCC score for a completely unknown independent test set. The only method that appears as the real competitor is SVM-Lin, which nearly equalizes the accuracy in prediction of PC-RMTL but only when complete feature sets are provided for training; otherwise, PC-RMTL outperformed all other classification models. To the best of our knowledge, this is a significant improvement over all the existing works in pan-cancer classification as they have failed to classify many cancer types from one another reliably. We have also compared gene expression patterns of the top discriminating genes across the cancers and performed their functional enrichment analysis that uncovers several interesting facts in distinguishing pan-cancer samples.

RgCop-A regularized copula based method for gene selection in single-cell RNA-seq data.

Gene selection in unannotated large single cell RNA sequencing (scRNA-seq) data is important and crucial step in the preliminary step of downstream analysis. The existing approaches are primarily based on high variation (highly variable genes) or significant high expression (highly expressed genes) failed to provide stable and predictive feature set due to technical noise present in the data. Here, we propose RgCop, a novel regularized copula based method for gene selection from large single cell RNA-seq data. RgCop utilizes copula correlation (Ccor), a robust equitable dependence measure that captures multivariate dependency among a set of genes in single cell expression data. We formulate an objective function by adding l1 regularization term with Ccor to penalizes the redundant co-efficient of features/genes, resulting non-redundant effective features/genes set. Results show a significant improvement in the clustering/classification performance of real life scRNA-seq data over the other state-of-the-art. RgCop performs extremely well in capturing dependence among the features of noisy data due to the scale invariant property of copula, thereby improving the stability of the method. Moreover, the differentially expressed (DE) genes identified from the clusters of scRNA-seq data are found to provide an accurate annotation of cells. Finally, the features/genes obtained from RgCop is able to annotate the unknown cells with high accuracy.

Identification of key immune regulatory genes in HIV-1 progression.

Human immunodeficiency virus (HIV) infection causes acquired immunodeficiency syndrome (AIDS), one of the most devastating diseases affecting humankind. Here, we have proposed a framework to examine the differences among microarray gene expression data of uninfected and three different HIV-1 infection stages using module preservation statistics. We leverage the advantage of gene co-expression networks (GCN) constructed for each infection stages to detect the topological and structural changes of a group of differentially expressed genes. We examine the relationship among a set of co-expression modules by constructing a module eigengene network considering the overall similarity/dissimilarity among the genes within the modules. We have utilized different module preservation statistics with two composite statistics: "Zsummary" and "MedianRank" to examine the changes in co-expression patterns between modules. We have found several interesting results on the preservation characteristics of gene modules across different stages. Some genes are identified to be preserved in a pair of stages while altering their characteristics across other stages. We further validated the obtained results using permutation test and classification techniques. The biological significances of the obtained modules have also been examined using gene ontology and pathway-based analysis. Additionally, we have identified a set of key immune regulatory hub genes in the associated protein-protein interaction networks (PPINs) of the differentially expressed (DE) genes, which interacts with HIV-1 proteins and are likely to act as potential biomarkers in HIV-1 progression.

Differential Patterns of Social Media Use Associated with Loneliness and Health Outcomes in Selected Socioeconomic Groups.

Loneliness has emerged as a chronic and persistent problem for a considerable fraction of the general population in the developed world. Concurrently, use of online social media by the same societies has steadily increased over the past two decades. The present study analyzed a recent large country-wide loneliness survey of 20,096 adults in the US using an unsupervised approach for systematic identification of clusters of respondents in terms of their social media use and representation among different socioeconomic subgroups. We studied the underlying population heterogeneity with a computational pipeline that was developed to gain insights into cluster- or group-specific patterns of loneliness. In particular, distributions of high loneliness were observed in groups of female users of Facebook and YouTube of certain age, race, marital, and socioeconomic status. For instance, among the group of predominantly YouTube users, we noted that non-Hispanic white female respondents of age 25-44 years who have high school or less education level and are single or never married have more significant high loneliness distribution. In fact, their high loneliness scores also seem to be associated with self-reported poorer physical and mental health outcomes. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s41347-021-00208-4.

Discovering key transcriptomic regulators in pancreatic ductal adenocarcinoma using Dirichlet process Gaussian mixture model.

Pancreatic Ductal Adenocarcinoma (PDAC) is the most lethal type of pancreatic cancer, late detection leading to its therapeutic failure. This study aims to determine the key regulatory genes and their impacts on the disease's progression, helping the disease's etiology, which is still mostly unknown. We leverage the landmark advantages of time-series gene expression data of this disease and thereby identified the key regulators that capture the characteristics of gene activity patterns in the cancer progression. We have identified the key gene modules and predicted the functions of top genes from a reconstructed gene association network (GAN). A variation of the partial correlation method is utilized to analyze the GAN, followed by a gene function prediction task. Moreover, we have identified regulators for each target gene by gene regulatory network inference using the dynamical GENIE3 (dynGENIE3) algorithm. The Dirichlet process Gaussian process mixture model and cubic spline regression model (splineTimeR) are employed to identify the key gene modules and differentially expressed genes, respectively. Our analysis demonstrates a panel of key regulators and gene modules that are crucial for PDAC disease progression.

Clustering Patterns Connecting COVID-19 Dynamics and Human Mobility Using Optimal Transport.

Social distancing and stay-at-home are among the few measures that are known to be effective in checking the spread of a pandemic such as COVID-19 in a given population. The patterns of dependency between such measures and their effects on disease incidence may vary dynamically and across different populations. We described a new computational framework to measure and compare the temporal relationships between human mobility and new cases of COVID-19 across more than 150 cities of the United States with relatively high incidence of the disease. We used a novel application of Optimal Transport for computing the distance between the normalized patterns induced by bivariate time series for each pair of cities. Thus, we identified 10 clusters of cities with similar temporal dependencies, and computed the Wasserstein barycenter to describe the overall dynamic pattern for each cluster. Finally, we used city-specific socioeconomic covariates to analyze the composition of each cluster.

DTI-SNNFRA: Drug-target interaction prediction by shared nearest neighbors and fuzzy-rough approximation.

In-silico prediction of repurposable drugs is an effective drug discovery strategy that supplements de-nevo drug discovery from scratch. Reduced development time, less cost and absence of severe side effects are significant advantages of using drug repositioning. Most recent and most advanced artificial intelligence (AI) approaches have boosted drug repurposing in terms of throughput and accuracy enormously. However, with the growing number of drugs, targets and their massive interactions produce imbalanced data which may not be suitable as input to the classification model directly. Here, we have proposed DTI-SNNFRA, a framework for predicting drug-target interaction (DTI), based on shared nearest neighbour (SNN) and fuzzy-rough approximation (FRA). It uses sampling techniques to collectively reduce the vast search space covering the available drugs, targets and millions of interactions between them. DTI-SNNFRA operates in two stages: first, it uses SNN followed by a partitioning clustering for sampling the search space. Next, it computes the degree of fuzzy-rough approximations and proper degree threshold selection for the negative samples' undersampling from all possible interaction pairs between drugs and targets obtained in the first stage. Finally, classification is performed using the positive and selected negative samples. We have evaluated the efficacy of DTI-SNNFRA using AUC (Area under ROC Curve), Geometric Mean, and F1 Score. The model performs exceptionally well with a high prediction score of 0.95 for ROC-AUC. The predicted drug-target interactions are validated through an existing drug-target database (Connectivity Map (Cmap)).

Colored Network Motif Analysis by Dynamic Programming Approach: An Application in Host Pathogen Interaction Network.

Network motifs are subgraphs of a network which are found with significantly higher frequency than that expected in similar random networks. Motifs are small building blocks of a network and they have emerged as a way to uncover topological properties of complex networks. A special yet not much explored type of motif is the 'colored motif' where color (type) of each node, and hence the edges, in the motif is distinguishable from each other. A traditional motif is defined as a recurring structure in a network, whereas colored motif introduces detailed information about the color of the nodes. G-trie is a data structure to efficiently store a given set of subgraphs by exploiting the topological overlaps within them. In this article we have implemented a modified g-trie to store colored subgraphs and developed a method to discover colored motifs. Our method uses an approximate enumeration for counting the subgraphs to reduce the runtime. We have applied our method to find colored motifs of size three in a host pathogen protein-protein interaction network having two types of proteins namely HIV-1 and human proteins, and four types of edges. Here, we have discovered eight motifs, six of which contain both HIV-1 and human proteins, while the remaining two contain only human proteins.

CODC: a Copula-based model to identify differential coexpression.

Differential coexpression has recently emerged as a new way to establish a fundamental difference in expression pattern among a group of genes between two populations. Earlier methods used some scoring techniques to detect changes in correlation patterns of a gene pair in two conditions. However, modeling differential coexpression by means of finding differences in the dependence structure of the gene pair has hitherto not been carried out. We exploit a copula-based framework to model differential coexpression between gene pairs in two different conditions. The Copula is used to model the dependency between expression profiles of a gene pair. For a gene pair, the distance between two joint distributions produced by copula is served as differential coexpression. We used five pan-cancer TCGA RNA-Seq data to evaluate the model that outperforms the existing state of the art. Moreover, the proposed model can detect a mild change in the coexpression pattern across two conditions. For noisy expression data, the proposed method performs well because of the popular scale-invariant property of copula. In addition, we have identified differentially coexpressed modules by applying hierarchical clustering on the distance matrix. The identified modules are analyzed through Gene Ontology terms and KEGG pathway enrichment analysis.

Computational Prediction of HCV-Human Protein-Protein Interaction via Topological Analysis of HCV Infected PPI Modules.

In this paper, we have developed a framework for detection of protein-protein interactions (PPI) between Hepatitis-C virus (HCV) and human proteins based on PPI and gene ontology based information of the HCV infected proteins. First, a bipartite interaction network is formed between HCV proteins and human host proteins. Next, we have analyzed different topological properties of the interaction network and observed that degree of HCV-interacting proteins is significantly higher than non-interacting host proteins. We have also observed that the HCV interacted protein pairs are functionally similar with each other than the non-interacting pairs. Following the observations, we have applied an inference mechanism to predict novel interactions between HCV and human protein. The inference mechanism is based on partitioning the network formed by HCV interacted human proteins and their first neighbors in dense and functionally similar groups using a PPI network clustering algorithm. The groups are then analyzed to predict PPIs. The predicted interaction pairs are validated using literature search in PUBMED. Experimental evidence of over 50% of the predicted pairs are found in existing literatures by searching PUBMED. A Gene Ontology and pathway based analysis is also carried out to validate the identified modules biologically.

Discovering Perturbation of Modular Structure in HIV Progression by Integrating Multiple Data Sources Through Non-Negative Matrix Factorization.

Detecting perturbation in modular structure during HIV-1 disease progression is an important step to understand stage specific infection pattern of HIV-1 virus in human cell. In this article, we proposed a novel methodology on integration of multiple biological information to identify such disruption in human gene module during different stages of HIV-1 infection. We integrate three different biological information: gene expression information, protein-protein interaction information, and gene ontology information in single gene meta-module, through non negative matrix factorization (NMF). As the identified meta-modules inherit those information so, detecting perturbation of these, reflects the changes in expression pattern, in PPI structure and in functional similarity of genes during the infection progression. To integrate modules of different data sources into strong meta-modules, NMF based clustering is utilized here. Perturbation in meta-modular structure is identified by investigating the topological and intramodular properties and putting rank to those meta-modules using a rank aggregation algorithm. We have also analyzed the preservation structure of significant GO terms in which the human proteins of the meta-modules participate. Moreover, we have performed an analysis to show the change of coregulation pattern of identified transcription factors (TFs) over the HIV progression stages.

A review of computational approaches for analysis of hepatitis C virus-mediated liver diseases.

Chronic infection of hepatitis C virus (HCV) leads to severe life-threatening liver diseases such as cirrhosis of liver, fibrosis and hepatocellular carcinoma (HCC). Severity of the disease infects >180 million people worldwide. In recent years, many computational approaches have been proposed to study and analyze the progression of liver fibrosis, HCC and other liver diseases developed from chronic HCV infection. In this article, we review the literature published in this area of study. Here we categorize all the approaches into two basic groups: analyzing gene expression and studying protein-protein interaction network among HCV-infected human proteins. We also review functional and pathway-enrichment analysis of HCV-interacted human proteins, which gives a clear understanding of functional perturbations leading to hepatocarcinogenesis. Topological analysis of HCV-human protein interaction network and HCV-HCC association network reveals important information of hepatocarcinogenesis progression in liver tissue. We compare the results of topological analysis performed in different studies. Moreover we observe that the HCV-interacted human proteins, which are also responsible for HCC progression, have relatively higher degree and betweenness centrality than that of the other HCV-interacted proteins.

A comprehensive analysis on preservation patterns of gene co-expression networks during Alzheimer's disease progression.

BACKGROUND: Alzheimer's disease (AD) is a chronic neuro-degenerative disruption of the brain which involves in large scale transcriptomic variation. The disease does not impact every regions of the brain at the same time, instead it progresses slowly involving somewhat sequential interaction with different regions. Analysis of the expression patterns of the genes in different regions of the brain influenced in AD surely contribute for a enhanced comprehension of AD pathogenesis and shed light on the early characterization of the disease. RESULTS: Here, we have proposed a framework to identify perturbation and preservation characteristics of gene expression patterns across six distinct regions of the brain ("EC", "HIP", "PC", "MTG", "SFG", and "VCX") affected in AD. Co-expression modules were discovered considering a couple of regions at once. These are then analyzed to know the preservation and perturbation characteristics. Different module preservation statistics and a rank aggregation mechanism have been adopted to detect the changes of expression patterns across brain regions. Gene ontology (GO) and pathway based analysis were also carried out to know the biological meaning of preserved and perturbed modules. CONCLUSIONS: In this article, we have extensively studied the preservation patterns of co-expressed modules in six distinct brain regions affected in AD. Some modules are emerged as the most preserved while some others are detected as perturbed between a pair of brain regions. Further investigation on the topological properties of preserved and non-preserved modules reveals a substantial association amongst "betweenness centrality" and "degree" of the involved genes. Our findings may render a deeper realization of the preservation characteristics of gene expression patterns in discrete brain regions affected by AD.

Identifying protein complexes in PPI network using non-cooperative sequential game.

Identifying protein complexes from protein-protein interaction (PPI) network is an important and challenging task in computational biology as it helps in better understanding of cellular mechanisms in various organisms. In this paper we propose a noncooperative sequential game based model for protein complex detection from PPI network. The key hypothesis is that protein complex formation is driven by mechanism that eventually optimizes the number of interactions within the complex leading to dense subgraph. The hypothesis is drawn from the observed network property named small world. The proposed multi-player game model translates the hypothesis into the game strategies. The Nash equilibrium of the game corresponds to a network partition where each protein either belong to a complex or form a singleton cluster. We further propose an algorithm to find the Nash equilibrium of the sequential game. The exhaustive experiment on synthetic benchmark and real life yeast networks evaluates the structural as well as biological significance of the network partitions.