A novel link prediction algorithm for reconstructing protein-protein interaction networks by topological similarity

Applying network link prediction in drug discovery: an overview of the literature

INTRODUCTION

Network representation can give a holistic view of relationships for biomedical entities through network topology. Link prediction estimates the probability of link formation between the pair of unconnected nodes. In the drug discovery process, the link prediction method not only enables the detection of connectivity patterns but also predicts the effects of one biomedical entity to multiple entities simultaneously and vice versa, which is useful for many applications.

AREAS COVERED

The authors provide a comprehensive overview of network link prediction in drug discovery. Link prediction methodologies such as similarity-based approaches, embedding-based approaches, probabilistic model-based approaches, and preprocessing methods are summarized with examples. In addition to describing their properties and limitations, the authors discuss the applications of link prediction in drug discovery based on the relationship between biomedical concepts.

EXPERT OPINION

Link prediction is a powerful method to infer the existence of novel relationships in drug discovery. However, link prediction has been hampered by the sparsity of data and the lack of negative links in biomedical networks. With preprocessing to balance positive and negative samples and the collection of more data, the authors believe it is possible to develop more reliable link prediction methods that can become invaluable tools for successful drug discovery.

Hypergraph reconstruction from uncertain pairwise observations

The network reconstruction task aims to estimate a complex system's structure from various data sources such as time series, snapshots, or interaction counts. Recent work has examined this problem in networks whose relationships involve precisely two entities-the pairwise case. Here, using Bayesian inference, we investigate the general problem of reconstructing a network in which higher-order interactions are also present. We study a minimal example of this problem, focusing on the case of hypergraphs with interactions between pairs and triplets of vertices, measured imperfectly and indirectly. We derive a Metropolis-Hastings-within-Gibbs algorithm for this model to highlight the unique challenges that come with estimating higher-order models. We show that this approach tends to reconstruct empirical and synthetic networks more accurately than an equivalent graph model without higher-order interactions.

The importance of graph databases and graph learning for clinical applications

The increasing amount and complexity of clinical data require an appropriate way of storing and analyzing those data. Traditional approaches use a tabular structure (relational databases) for storing data and thereby complicate storing and retrieving interlinked data from the clinical domain. Graph databases provide a great solution for this by storing data in a graph as nodes (vertices) that are connected by edges (links). The underlying graph structure can be used for the subsequent data analysis (graph learning). Graph learning consists of two parts: graph representation learning and graph analytics. Graph representation learning aims to reduce high-dimensional input graphs to low-dimensional representations. Then, graph analytics uses the obtained representations for analytical tasks like visualization, classification, link prediction and clustering which can be used to solve domain-specific problems. In this survey, we review current state-of-the-art graph database management systems, graph learning algorithms and a variety of graph applications in the clinical domain. Furthermore, we provide a comprehensive use case for a clearer understanding of complex graph learning algorithms. Graphical abstract.

A link prediction-based recommendation system using transactional data

Recommending relevant items to users has become an important task in many systems due to the increased amount of data produced. For this purpose, transaction datasets such as credit card transactions and e-commerce purchase histories can be used in recommendation systems to understand underlying user interests by exploiting user-item interactions, which can be a powerful signal to perform this task. This study proposes a link prediction-based recommendation system combining graph representation learning algorithms and gradient boosting classifiers for transaction datasets. The proposed system generates a network where nodes correspond to users and items, and links represent their interactions. A use case scenario is examined on a credit card transaction dataset as a merchant prediction task that predicts the merchants where users can make purchases in the next month. Performances of common network embedding extraction techniques and classifier models are evaluated via various experiments conducted and based on these evaluations, a novel system is proposed, and a matrix factorization-based alternative recommendation method is compared with the proposed model. The proposed method has shown superior performance to the alternative method in terms of receiver operating characteristic curves, area under the curve, and mean average precision metrics. The use of transactional data for a recommendation system is found to be a powerful approach to making relevant recommendations.

Normalized L3-based link prediction in protein-protein interaction networks

BACKGROUND

Protein-protein interaction (PPI) data is an important type of data used in functional genomics. However, high-throughput experiments are often insufficient to complete the PPI interactome of different organisms. Computational techniques are thus used to infer missing data, with link prediction being one such approach that uses the structure of the network of PPIs known so far to identify non-edges whose addition to the network would make it more sound, according to some underlying assumptions. Recently, a new idea called the L3 principle introduced biological motivation into PPI link predictions, yielding predictors that are superior to general-purpose link predictors for complex networks. Interestingly, the L3 principle can be interpreted in another way, so that other signatures of PPI networks can also be characterized for PPI predictions. This alternative interpretation uncovers candidate PPIs that the current L3-based link predictors may not be able to fully capture, underutilizing the L3 principle.

RESULTS

In this article, we propose a formulation of link predictors that we call NormalizedL3 (L3N) which addresses certain missing elements within L3 predictors in the perspective of network modeling. Our computational validations show that the L3N predictors are able to find missing PPIs more accurately (in terms of true positives among the predicted PPIs) than the previously proposed methods on several datasets from the literature, including BioGRID, STRING, MINT, and HuRI, at the cost of using more computation time in some of the cases. In addition, we found that L3-based link predictors (including L3N) ranked a different pool of PPIs higher than the general-purpose link predictors did. This suggests that different types of PPIs can be predicted based on different topological assumptions, and that even better PPI link predictors may be obtained in the future by improved network modeling.

Using PlaPPISite to Predict and Analyze Plant Protein-Protein Interaction Sites

Proteome-wide characterization of protein-protein interactions (PPIs) is crucial to understand the functional roles of protein machinery within cells systematically. With the accumulation of PPI data in different plants, the interaction details of binary PPIs, such as the three-dimensional (3D) structural contexts of interaction sites/interfaces, are urgently demanded. To meet this requirement, we have developed a comprehensive and easy-to-use database called PlaPPISite ( http://zzdlab.com/plappisite/index.php ) to present interaction details for 13 plant interactomes. Here, we provide a clear guide on how to search and view protein interaction details through the PlaPPISite database. Firstly, the running environment of our database is introduced. Secondly, the input file format is briefly introduced. Moreover, we discussed which information related to interaction sites can be achieved through several examples. In addition, some notes about PlaPPISite are also provided. More importantly, we would like to emphasize the importance of interaction site information in plant systems biology through this user guide of PlaPPISite. In particular, the easily accessible 3D structures of PPIs in the coming post-AlphaFold2 era will definitely boost the application of plant interactome to decipher the molecular mechanisms of many fundamental biological issues.

Link Prediction Based on Stochastic Information Diffusion

Link prediction (LP) in networks aims at determining future interactions among elements; it is a critical machine-learning tool in different domains, ranging from genomics to social networks to marketing, especially in e-commerce recommender systems. Although many LP techniques have been developed in the prior art, most of them consider only static structures of the underlying networks, rarely incorporating the network's information flow. Exploiting the impact of dynamic streams, such as information diffusion, is still an open research topic for LP. Information diffusion allows nodes to receive information beyond their social circles, which, in turn, can influence the creation of new links. In this work, we analyze the LP effects through two diffusion approaches, susceptible-infected-recovered and independent cascade. As a result, we propose the progressive-diffusion (PD) method for LP based on nodes' propagation dynamics. The proposed model leverages a stochastic discrete-time rumor model centered on each node's propagation dynamics. It presents low-memory and low-processing footprints and is amenable to parallel and distributed processing implementation. Finally, we also introduce an evaluation metric for LP methods considering both the information diffusion capacity and the LP accuracy. Experimental results on a series of benchmarks attest to the proposed method's effectiveness compared with the prior art in both criteria.

Link Prediction on Complex Networks: An Experimental Survey

Complex networks have been used widely to model a large number of relationships. The outbreak of COVID-19 has had a huge impact on various complex networks in the real world, for example global trade networks, air transport networks, and even social networks, known as racial equality issues caused by the spread of the epidemic. Link prediction plays an important role in complex network analysis in that it can find missing links or predict the links which will arise in the future in the network by analyzing the existing network structures. Therefore, it is extremely important to study the link prediction problem on complex networks. There are a variety of techniques for link prediction based on the topology of the network and the properties of entities. In this work, a new taxonomy is proposed to divide the link prediction methods into five categories and a comprehensive overview of these methods is provided. The network embedding-based methods, especially graph neural network-based methods, which have attracted increasing attention in recent years, have been creatively investigated as well. Moreover, we analyze thirty-six datasets and divide them into seven types of networks according to their topological features shown in real networks and perform comprehensive experiments on these networks. We further analyze the results of experiments in detail, aiming to discover the most suitable approach for each kind of network.

deepHPI: a comprehensive deep learning platform for accurate prediction and visualization of host-pathogen protein-protein interactions

Host-pathogen protein interactions (HPPIs) play vital roles in many biological processes and are directly involved in infectious diseases. With the outbreak of more frequent pandemics in the last couple of decades, such as the recent outburst of Covid-19 causing millions of deaths, it has become more critical to develop advanced methods to accurately predict pathogen interactions with their respective hosts. During the last decade, experimental methods to identify HPIs have been used to decipher host-pathogen systems with the caveat that those techniques are labor-intensive, expensive and time-consuming. Alternatively, accurate prediction of HPIs can be performed by the use of data-driven machine learning. To provide a more robust and accurate solution for the HPI prediction problem, we have developed a deepHPI tool based on deep learning. The web server delivers four host-pathogen model types: plant-pathogen, human-bacteria, human-virus and animal-pathogen, leveraging its operability to a wide range of analyses and cases of use. The deepHPI web tool is the first to use convolutional neural network models for HPI prediction. These models have been selected based on a comprehensive evaluation of protein features and neural network architectures. The best prediction models have been tested on independent validation datasets, which achieved an overall Matthews correlation coefficient value of 0.87 for animal-pathogen using the combined pseudo-amino acid composition and conjoint triad (PAAC_CT) features, 0.75 for human-bacteria using the combined pseudo-amino acid composition, conjoint triad and normalized Moreau-Broto feature (PAAC_CT_NMBroto), 0.96 for human-virus using PAAC_CT_NMBroto and 0.94 values for plant-pathogen interactions using the combined pseudo-amino acid composition, composition and transition feature (PAAC_CTDC_CTDT). Our server running deepHPI is deployed on a high-performance computing cluster that enables large and multiple user requests, and it provides more information about interactions discovered. It presents an enriched visualization of the resulting host-pathogen networks that is augmented with external links to various protein annotation resources. We believe that the deepHPI web server will be very useful to researchers, particularly those working on infectious diseases. Additionally, many novel and known host-pathogen systems can be further investigated to significantly advance our understanding of complex disease-causing agents. The developed models are established on a web server, which is freely accessible at http://bioinfo.usu.edu/deepHPI/.

Predicting Possible New Links to Future Global Plastic Waste Trade Networks

China’s waste plastic ban has sparked a discussion about how the global plastic waste trade (GPWT) will develop in the future. To answer this question, this article uses the link forecasting and QAP method to predict and analyze the possible development trend of the GPWT in the future. The research results show that GPWT has certain stability and sustainability; although plastic waste trade has narrowed under the ban, it still has the potential trend of reconnecting the same type of links. Specifically, from a regional perspective, the future trade of new plastic waste trade will be dominated by cross-regional trade. Plastic waste may continue to flow to countries in the Asian–Pacific, Middle East, and African regions, while European countries will strengthen the internal recycling and processing of plastic waste. From the perspective of the national income level, the establishment of the new relationship will show an evolutionary trend in which high-income countries are dominated and the scale of trade between non-high-income countries expands. In addition, the differences in the level of economic development, liner transport connectivity, and the proportion of mismanagement of plastic waste among countries has a positive effect on the establishment of a new relationship in the GPWT, while tariff rates have an inhibitory effect. In general, the GPWT will still exist in the future, which requires the international community to guide the GPWT to promote the recycling and reuse of plastic waste in a real sense and adjust the unreasonable trade model.

Progresses and challenges in link prediction

Link prediction is a paradigmatic problem in network science, which aims at estimating the existence likelihoods of nonobserved links, based on known topology. After a brief introduction of the standard problem and evaluation metrics of link prediction, this review will summarize representative progresses about local similarity indices, link predictability, network embedding, matrix completion, ensemble learning, and some others, mainly extracted from related publications in the last decade. Finally, this review will outline some long-standing challenges for future studies.

Performance Assessment of the Network Reconstruction Approaches on Various Interactomes

Beyond the list of molecules, there is a necessity to collectively consider multiple sets of omic data and to reconstruct the connections between the molecules. Especially, pathway reconstruction is crucial to understanding disease biology because abnormal cellular signaling may be pathological. The main challenge is how to integrate the data together in an accurate way. In this study, we aim to comparatively analyze the performance of a set of network reconstruction algorithms on multiple reference interactomes. We first explored several human protein interactomes, including PathwayCommons, OmniPath, HIPPIE, iRefWeb, STRING, and ConsensusPathDB. The comparison is based on the coverage of each interactome in terms of cancer driver proteins, structural information of protein interactions, and the bias toward well-studied proteins. We next used these interactomes to evaluate the performance of network reconstruction algorithms including all-pair shortest path, heat diffusion with flux, personalized PageRank with flux, and prize-collecting Steiner forest (PCSF) approaches. Each approach has its own merits and weaknesses. Among them, PCSF had the most balanced performance in terms of precision and recall scores when 28 pathways from NetPath were reconstructed using the listed algorithms. Additionally, the reference interactome affects the performance of the network reconstruction approaches. The coverage and disease- or tissue-specificity of each interactome may vary, which may result in differences in the reconstructed networks.

Who Will Establish New Trade Relations? Looking for Potential Relationship in International Nickel Trade

Nickel ore sand and its concentrate are the main sources of raw nickel materials in various countries. Due to its uneven distribution throughout the world, the international trade of nickel ore sand is also unstable. Looking for potential links in the changing international nickel ore trade can help governments find potential partners, make strategic preparations in advance, and quickly find new partners when original trade relationships break down. In this paper, we build an international nickel ore trade network using a link prediction method to find potential trade relations between countries. The results show that China and Italy, China and Denmark, China and Indonesia, and China and India are most likely to establish trade relations within five years. Finally, according to the research results, suggestions regarding the international nickel ore trade are proposed.

D-SCRIPT translates genome to phenome with sequence-based, structure-aware, genome-scale predictions of protein-protein interactions

We combine advances in neural language modeling and structurally motivated design to develop D-SCRIPT, an interpretable and generalizable deep-learning model, which predicts interaction between two proteins using only their sequence and maintains high accuracy with limited training data and across species. We show that a D-SCRIPT model trained on 38,345 human PPIs enables significantly improved functional characterization of fly proteins compared with the state-of-the-art approach. Evaluating the same D-SCRIPT model on protein complexes with known 3D structure, we find that the inter-protein contact map output by D-SCRIPT has significant overlap with the ground truth. We apply D-SCRIPT to screen for PPIs in cow (Bos taurus) at a genome-wide scale and focusing on rumen physiology, identify functional gene modules related to metabolism and immune response. The predicted interactions can then be leveraged for function prediction at scale, addressing the genome-to-phenome challenge, especially in species where little data are available.

Node Similarity Based Graph Convolution for Link Prediction in Biological Networks

BACKGROUND

Link prediction is an important and well-studied problem in network biology. Recently, graph representation learning methods, including Graph Convolutional Network (GCN)-based node embedding have drawn increasing attention in link prediction.

MOTIVATION

An important component of GCN-based network embedding is the convolution matrix, which is used to propagate features across the network. Existing algorithms use the degree-normalized adjacency matrix for this purpose, as this matrix is closely related to the graph Laplacian, capturing the spectral properties of the network. In parallel, it has been shown that GCNs with a single layer can generate more robust embeddings by reducing the number of parameters. Laplacian-based convolution is not well suited to single layered GCNs, as it limits the propagation of information to immediate neighbors of a node.

RESULTS

Capitalizing on the rich literature on unsupervised link prediction, we propose using node similarity based convolution matrices in GCNs to compute node embeddings for link prediction. We consider eight representative node similarity measures (Common Neighbors, Jaccard Index, Adamic-Adar, Resource Allocation, Hub Depressed Index, Hub Promoted Index, Sorenson Index, Salton Index) for this purpose. We systematically compare the performance of the resulting algorithms against GCNs that use the degree-normalized adjacency matrix for convolution, as well as other link prediction algorithms. In our experiments, we use three link prediction tasks involving biomedical networks: drug-disease association (DDA) prediction, drug-drug interaction (DDI) prediction, protein-protein interaction (PPI) prediction. Our results show that node similarity-based convolution matrices significantly improve the link prediction performance of GCN-based embeddings.

CONCLUSION

As sophisticated machine learning frameworks are increasingly employed in biological applications, historically well-established methods can be useful in making a head-start.

AVAILABILITY

Our method, SiGraC, is implemented as a Python library and is freely available at https://github.com/mustafaCoskunAgu/SiGraC.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Denoising Protein-Protein interaction network via variational graph auto-encoder for protein complex detection

Identifying protein complexes is an important issue in computational biology, as it benefits the understanding of cellular functions and the design of drugs. In the past decades, many computational methods have been proposed by mining dense subgraphs in Protein-Protein Interaction Networks (PINs). However, the high rate of false positive/negative interactions in PINs prevents accurately detecting complexes directly from the raw PINs. In this paper, we propose a denoising approach for protein complex detection by using variational graph auto-encoder. First, we embed a PIN to vector space by a stacked graph convolutional network (GCN), then decide which interactions in the PIN are credible. If the probability of an interaction being credible is less than a threshold, we delete the interaction. In such a way, we reconstruct a reliable PIN. Following that, we detect protein complexes in the reconstructed PIN by using several typical detection methods, including CPM, Coach, DPClus, GraphEntropy, IPCA and MCODE, and compare the results with those obtained directly from the original PIN. We conduct the empirical evaluation on four yeast PPI datasets (Gavin, Krogan, DIP and Wiphi) and two human PPI datasets (Reactome and Reactomekb), against two yeast complex benchmarks (CYC2008 and MIPS) and three human complex benchmarks (REACT, REACT_uniprotkb and CORE_COMPLEX_human), respectively. Experimental results show that with the reconstructed PINs obtained by our denoising approach, complex detection performance can get obviously boosted, in most cases by over 5%, sometimes even by 200%. Furthermore, we compare our approach with two existing denoising methods (RWS and RedNemo) while varying different matching rates on separate complex distributions. Our results show that in most cases (over 2/3), the proposed approach outperforms the existing methods.

GLIDE: combining local methods and diffusion state embeddings to predict missing interactions in biological networks

Motivation

One of the core problems in the analysis of biological networks is the link prediction problem. In particular, existing interactions networks are noisy and incomplete snapshots of the true network, with many true links missing because those interactions have not yet been experimentally observed. Methods to predict missing links have been more extensively studied for social than for biological networks; it was recently argued that there is some special structure in protein–protein interaction (PPI) network data that might mean that alternate methods may outperform the best methods for social networks. Based on a generalization of the diffusion state distance, we design a new embedding-based link prediction method called global and local integrated diffusion embedding (GLIDE). GLIDE is designed to effectively capture global network structure, combined with alternative network type-specific customized measures that capture local network structure. We test GLIDE on a collection of three recently curated human biological networks derived from the 2016 DREAM disease module identification challenge as well as a classical version of the yeast PPI network in rigorous cross validation experiments.

Results

We indeed find that different local network structure is dominant in different types of biological networks. We find that the simple local network measures are dominant in the highly connected network core between hub genes, but that GLIDE’s global embedding measure adds value in the rest of the network. For example, we make GLIDE-based link predictions from genes known to be involved in Crohn’s disease, to genes that are not known to have an association, and make some new predictions, finding support in other network data and the literature.

Availability and implementation

GLIDE can be downloaded at https://bitbucket.org/kap_devkota/glide.

Supplementary information

Supplementary data are available at Bioinformatics online.

A survey on computational models for predicting protein-protein interactions

Proteins interact with each other to play critical roles in many biological processes in cells. Although promising, laboratory experiments usually suffer from the disadvantages of being time-consuming and labor-intensive. The results obtained are often not robust and considerably uncertain. Due recently to advances in high-throughput technologies, a large amount of proteomics data has been collected and this presents a significant opportunity and also a challenge to develop computational models to predict protein-protein interactions (PPIs) based on these data. In this paper, we present a comprehensive survey of the recent efforts that have been made towards the development of effective computational models for PPI prediction. The survey introduces the algorithms that can be used to learn computational models for predicting PPIs, and it classifies these models into different categories. To understand their relative merits, the paper discusses different validation schemes and metrics to evaluate the prediction performance. Biological databases that are commonly used in different experiments for performance comparison are also described and their use in a series of extensive experiments to compare different prediction models are discussed. Finally, we present some open issues in PPI prediction for future work. We explain how the performance of PPI prediction can be improved if these issues are effectively tackled.

An In Silico Method for Predicting Drug Synergy Based on Multitask Learning

To make better use of all kinds of knowledge to predict drug synergy, it is crucial to successfully establish a drug synergy prediction model and leverage the reconstruction of sparse known drug targets. Therefore, we present an in silico method that predicts the synergy scores of drug pairs based on multitask learning (DSML) that could fuse drug targets, protein-protein interactions, anatomical therapeutic chemical codes, a priori knowledge of drug combinations. To simultaneously reconstruct drug-target protein interactions and synergistic drug combinations, DSML benefits indirectly from the associations with relation through proteins. In cross-validation experiments, DSML improved the ability to predict drug synergy. Moreover, the reconstruction of drug-target interactions and the incorporation of multisource knowledge significantly improved drug combination predictions by a large margin. The potential drug combinations predicted by DSML demonstrate its ability to predict drug synergy.

Pathogenic gene prediction based on network embedding

In disease research, the study of gene-disease correlation has always been an important topic. With the emergence of large-scale connected data sets in biology, we use known correlations between the entities, which may be from different sets, to build a biological heterogeneous network and propose a new network embedded representation algorithm to calculate the correlation between disease and genes, using the correlation score to predict pathogenic genes. Then, we conduct several experiments to compare our method to other state-of-the-art methods. The results reveal that our method achieves better performance than the traditional methods.

Review on Learning and Extracting Graph Features for Link Prediction

Link prediction in complex networks has attracted considerable attention from interdisciplinary research communities, due to its ubiquitous applications in biological networks, social networks, transportation networks, telecommunication networks, and, recently, knowledge graphs. Numerous studies utilized link prediction approaches in order sto find missing links or predict the likelihood of future links as well as employed for reconstruction networks, recommender systems, privacy control, etc. This work presents an extensive review of state-of-art methods and algorithms proposed on this subject and categorizes them into four main categories: similarity-based methods, probabilistic methods, relational models, and learning-based methods. Additionally, a collection of network data sets has been presented in this paper, which can be used in order to study link prediction. We conclude this study with a discussion of recent developments and future research directions.

SkipGNN: predicting molecular interactions with skip-graph networks

Molecular interaction networks are powerful resources for molecular discovery. They are increasingly used with machine learning methods to predict biologically meaningful interactions. While deep learning on graphs has dramatically advanced the prediction prowess, current graph neural network (GNN) methods are mainly optimized for prediction on the basis of direct similarity between interacting nodes. In biological networks, however, similarity between nodes that do not directly interact has proved incredibly useful in the last decade across a variety of interaction networks. Here, we present SkipGNN, a graph neural network approach for the prediction of molecular interactions. SkipGNN predicts molecular interactions by not only aggregating information from direct interactions but also from second-order interactions, which we call skip similarity. In contrast to existing GNNs, SkipGNN receives neural messages from two-hop neighbors as well as immediate neighbors in the interaction network and non-linearly transforms the messages to obtain useful information for prediction. To inject skip similarity into a GNN, we construct a modified version of the original network, called the skip graph. We then develop an iterative fusion scheme that optimizes a GNN using both the skip graph and the original graph. Experiments on four interaction networks, including drug–drug, drug–target, protein–protein, and gene–disease interactions, show that SkipGNN achieves superior and robust performance. Furthermore, we show that unlike popular GNNs, SkipGNN learns biologically meaningful embeddings and performs especially well on noisy, incomplete interaction networks.

Graph embedding-based novel protein interaction prediction via higher-order graph convolutional network

Protein–protein interactions (PPIs) are essential for most biological processes. However, current PPI networks present high levels of noise, sparseness and incompleteness, which limits our ability to understand the cell at the system level from the PPI network. Predicting novel (missing) links in noisy PPI networks is an essential computational method for automatically expanding the human interactome and for identifying biologically legitimate but undetected interactions for experimental determination of PPIs, which is both expensive and time-consuming. Recently, graph convolutional networks (GCN) have shown their effectiveness in modeling graph-structured data, which employ a 1-hop neighborhood aggregation procedure and have emerged as a powerful architecture for node or graph representations. In this paper, we propose a novel node (protein) embedding method by combining GCN and PageRank as the latter can significantly improve the GCN’s aggregation scheme, which has difficulty in extending and exploring topological information of networks across higher-order neighborhoods of each node. Building on this novel node embedding model, we develop a higher-order GCN variational auto-encoder (HO-VGAE) architecture, which can learn a joint node representation of higher-order local and global PPI network topology for novel protein interaction prediction. It is worth noting that our method is based exclusively on network topology, with no protein attributes or extra biological features used. Extensive computational validations on PPI prediction task demonstrate our method without leveraging any additional biological information shows competitive performance—outperforms all existing graph embedding-based link prediction methods in both accuracy and robustness.

Learning an expandable EMR-based medical knowledge network to enhance clinical diagnosis

Electronic medical records (EMRs) contain a wealth of knowledge that can be used to assist doctors in making clinical decisions like disease diagnosis. Constructing a medical knowledge network (MKN) to link medical concepts in EMRs is an effective way to manage this knowledge. The quality of the diagnostic result made by MKN-based clinical decision support system depends on the accuracy of medical knowledge and the completeness of the network. However, collecting knowledge is a long-lasting and cumulative process, which means it's hard to construct a complete MKN with limited data. This study was conducted with the objective of developing an expandable EMR-based MKN to enhance capabilities in making an initial clinical diagnosis. A network of symptom-indicate-disease knowledge in 992 Chinese EMRs (CEMRs) was manually constructed as Original-MKN, and an incremental expansion framework was applied to it to obtain an expandable MKN based on new CEMRs. The framework was composed by: (1) integrating external knowledge extracted from the medical information websites and (2) mining potential knowledge with new EMRs. The framework also adopts a diagnosis-driven learning method to estimate the effectiveness of each knowledge in clinical practice. Experimental results indicate that our expanded MKN achieves a precision of 0.837 for a recall of 0.719 in clinical diagnosis, which outperforms Original-MKN and four classical machine learning methods. Furthermore, both external medical knowledge and potential medical knowledge benefit MKN expansion and disease diagnosis. The proposed incremental expansion framework sustains the MKN learning new knowledge.

Computational Prediction of Probable Single Nucleotide Polymorphism-Cancer Relationships

Genetic variations such as single nucleotide polymorphisms (SNPs) can cause susceptibility to cancer. Although thousands of genetic variants have been identified to be associated with different cancers, the molecular mechanisms of cancer remain unknown. There is not a particular dataset of relationships between cancer and SNPs, as a bipartite network, for computational analysis and prediction. Link prediction as a computational graph analysis method can help us to gain new insight into the network. In this article, after creating a network between cancer and SNPs using SNPedia and Cancer Research UK databases, we evaluated the computational link prediction methods to foresee new SNP-Cancer relationships. Results show that among the popular scoring methods based on network topology, for relation prediction, the preferential attachment (PA) algorithm is the most robust method according to computational and experimental evidence, and some of its computational predictions are corroborated in recent publications. According to the PA predictions, rs1801394-Non-small cell lung cancer, rs4880-Non-small cell lung cancer, and rs1805794-Colorectal cancer are some of the best probable SNP-Cancer associations that have not yet been mentioned in any published article, and they are the most probable candidates for additional laboratory and validation studies. Also, it is feasible to improve the predicting algorithms to produce new predictions in the future.

Incorporating Multisource Knowledge To Predict Drug Synergy Based on Graph Co-regularization

Drug combinations may reduce toxicity and increase therapeutic efficacy, offering a promising strategy to conquer multiple complex diseases. However, due to large-scale combinatorial space, it remains challenging to identify effective combinations. Although many computational methods have focused on predicting drug synergy to reduce combinatorial space, they fail to effectively consider multiple sources of important knowledge. Thus, it is necessary to propose a computational method that can exploit useful information to predict drug synergy. Here, we developed a computational method to predict drug synergy based on graph co-regularization, named DSGCR. By incorporating drug-target network patterns, pharmacological patterns, and prior knowledge of drug combinations, DSGCR performs predictions of synergistic drug combinations. Compared to several existing methods, DSGCR achieves superior performance in predicting drug synergy in terms of various metrics via cross-validation. Additionally, we analyzed the importance of various sources of drug knowledge concerning three DSGCR's scenarios. Finally, the potential of DSGCR to score drug synergy was confirmed by three predicted synergistic drug combinations.

Gene Ontology Semantic Similarity Analysis Using GOSemSim

The GOSemSim package, an R-based tool within the Bioconductor project, offers several methods based on information content and graph structure for measuring semantic similarity among GO terms, gene products and gene clusters. In this chapter, I illustrate the use of GOSemSim on a list of regulators in preimplantation embryos. A step-by-step analysis was provided as well as instructions on interpretation and visualization of the results. GOSemSim is open-source and is available from https://www.bioconductor.org/packages/GOSemSim .

WIPER: Weighted in-Path Edge Ranking for biomolecular association networks

Background

In network biology researchers generate biomolecular networks with candidate genes or proteins experimentally-derived from high-throughput data and known biomolecular associations. Current bioinformatics research focuses on characterizing candidate genes/proteins, or nodes, with network characteristics, e.g., betweenness centrality. However, there have been few research reports to characterize and prioritize biomolecular associations ("edges"), which can represent gene regulatory events essential to biological processes.

Method

We developed Weighted In-Path Edge Ranking (WIPER), a new computational algorithm which can help evaluate all biomolecular interactions/associations ("edges") in a network model and generate a rank order of every edge based on their in-path traversal scores and statistical significance test result. To validate whether WIPER worked as we designed, we tested the algorithm on synthetic network models.

Results

Our results showed WIPER can reliably discover both critical "well traversed in-path edges", which are statistically more traversed than normal edges, and "peripheral in-path edges", which are less traversed than normal edges. Compared with other simple measures such as betweenness centrality, WIPER provides better biological interpretations. In the case study of analyzing postanal pig hearts gene expression, WIPER highlighted new signaling pathways suggestive of cardiomyocyte regeneration and proliferation. In the case study of Alzheimer's disease genetic disorder association, WIPER reports SRC:APP, AR:APP, APP:FYN, and APP:NES edges (gene-gene associations) both statistically and biologically important from PubMed co-citation.

Conclusion

We believe that WIPER will become an essential software tool to help biologists discover and validate essential signaling/regulatory events from high-throughput biology data in the context of biological networks.

Availability

The free WIPER API is described at discovery.informatics.uab.edu/wiper/.

scNPF: an integrative framework assisted by network propagation and network fusion for preprocessing of single-cell RNA-seq data

Background

Single-cell RNA-sequencing (scRNA-seq) is fast becoming a powerful tool for profiling genome-scale transcriptomes of individual cells and capturing transcriptome-wide cell-to-cell variability. However, scRNA-seq technologies suffer from high levels of technical noise and variability, hindering reliable quantification of lowly and moderately expressed genes. Since most downstream analyses on scRNA-seq, such as cell type clustering and differential expression analysis, rely on the gene-cell expression matrix, preprocessing of scRNA-seq data is a critical preliminary step in the analysis of scRNA-seq data.

Results

We presented scNPF, an integrative scRNA-seq preprocessing framework assisted by network propagation and network fusion, for recovering gene expression loss, correcting gene expression measurements, and learning similarities between cells. scNPF leverages the context-specific topology inherent in the given data and the priori knowledge derived from publicly available molecular gene-gene interaction networks to augment gene-gene relationships in a data driven manner. We have demonstrated the great potential of scNPF in scRNA-seq preprocessing for accurately recovering gene expression values and learning cell similarity networks. Comprehensive evaluation of scNPF across a wide spectrum of scRNA-seq data sets showed that scNPF achieved comparable or higher performance than the competing approaches according to various metrics of internal validation and clustering accuracy. We have made scNPF an easy-to-use R package, which can be used as a versatile preprocessing plug-in for most existing scRNA-seq analysis pipelines or tools.

Conclusions

scNPF is a universal tool for preprocessing of scRNA-seq data, which jointly incorporates the global topology of priori interaction networks and the context-specific information encapsulated in the scRNA-seq data to capture both shared and complementary knowledge from diverse data sources. scNPF could be used to recover gene signatures and learn cell-to-cell similarities from emerging scRNA-seq data to facilitate downstream analyses such as dimension reduction, cell type clustering, and visualization.

Electronic supplementary material

The online version of this article (10.1186/s12864-019-5747-5) contains supplementary material, which is available to authorized users.

The application of degree related clustering coefficient in estimating the link predictability and predicting missing links of networks

Though a lot of valuable algorithms of link prediction have been created, it is still difficult to improve the accuracy of link prediction for some networks. Such difficulties may be due to the intrinsic topological features of these networks. To reveal the correlation between the network topology and the link predictability, we generate a group of artificial networks by keeping some structural features of an initial seed network. Based on these artificial networks and some real networks, we find that five topological measures including clustering coefficient, structural consistency, random walk entropy, network diameter, and average path length significantly show their impact on the link predictability. Then, we define a topological score that combines these important topological features. Specifically, it is an integration of structural consistency with degree-related clustering coefficient defined in this work. This topological score exhibits high correlation with the link predictability. Finally, we propose an algorithm for link prediction based on this topological score. Our experiment on eight real networks verifies good performance of this algorithm in link prediction, which supports the reasonability of the new topological score. This work could be insightful for the study of the link predictability.

Modelling Self-Organization in Complex Networks Via a Brain-Inspired Network Automata Theory Improves Link Reliability in Protein Interactomes

Protein interactomes are epitomes of incomplete and noisy networks. Methods for assessing link-reliability using exclusively topology are valuable in network biology, and their investigation facilitates the general understanding of topological mechanisms and models to draw and correct complex network connectivity. Here, I revise and extend the local-community-paradigm (LCP). Initially detected in brain-network topological self-organization and afterward generalized to any complex network, the LCP is a theory to model local-topology-dependent link-growth in complex networks using network automata. Four novel LCP-models are compared versus baseline local-topology-models. It emerges that the reliability of an interaction between two proteins is higher: (i) if their common neighbours are isolated in a complex (local-community) that has low tendency to interact with other external proteins; (ii) if they have a low propensity to link with other proteins external to the local-community. These two rules are mathematically combined in C1*: a proposed mechanistic model that, in fact, outperforms the others. This theoretical study elucidates basic topological rules behind self-organization principia of protein interactomes and offers the conceptual basis to extend this theory to any class of complex networks. The link-reliability improvement, based on the mere topology, can impact many applied domains such as systems biology and network medicine.

Completing sparse and disconnected protein-protein network by deep learning

Background

Protein-protein interaction (PPI) prediction remains a central task in systems biology to achieve a better and holistic understanding of cellular and intracellular processes. Recently, an increasing number of computational methods have shifted from pair-wise prediction to network level prediction. Many of the existing network level methods predict PPIs under the assumption that the training network should be connected. However, this assumption greatly affects the prediction power and limits the application area because the current golden standard PPI networks are usually very sparse and disconnected. Therefore, how to effectively predict PPIs based on a training network that is sparse and disconnected remains a challenge.

Results

In this work, we developed a novel PPI prediction method based on deep learning neural network and regularized Laplacian kernel. We use a neural network with an autoencoder-like architecture to implicitly simulate the evolutionary processes of a PPI network. Neurons of the output layer correspond to proteins and are labeled with values (1 for interaction and 0 for otherwise) from the adjacency matrix of a sparse disconnected training PPI network. Unlike autoencoder, neurons at the input layer are given all zero input, reflecting an assumption of no a priori knowledge about PPIs, and hidden layers of smaller sizes mimic ancient interactome at different times during evolution. After the training step, an evolved PPI network whose rows are outputs of the neural network can be obtained. We then predict PPIs by applying the regularized Laplacian kernel to the transition matrix that is built upon the evolved PPI network. The results from cross-validation experiments show that the PPI prediction accuracies for yeast data and human data measured as AUC are increased by up to 8.4 and 14.9% respectively, as compared to the baseline. Moreover, the evolved PPI network can also help us leverage complementary information from the disconnected training network and multiple heterogeneous data sources. Tested by the yeast data with six heterogeneous feature kernels, the results show our method can further improve the prediction performance by up to 2%, which is very close to an upper bound that is obtained by an Approximate Bayesian Computation based sampling method.

Conclusions

The proposed evolution deep neural network, coupled with regularized Laplacian kernel, is an effective tool in completing sparse and disconnected PPI networks and in facilitating integration of heterogeneous data sources.

Linearity of network proximity measures: implications for set-based queries and significance testing

Motivation

In recent years, various network proximity measures have been proposed to facilitate the use of biomolecular interaction data in a broad range of applications. These applications include functional annotation, disease gene prioritization, comparative analysis of biological systems and prediction of new interactions. In such applications, a major task is the scoring or ranking of the nodes in the network in terms of their proximity to a given set of 'seed' nodes (e.g. a group of proteins that are identified to be associated with a disease, or are deferentially expressed in a certain condition). Many different network proximity measures are utilized for this purpose, and these measures are quite diverse in terms of the benefits they offer.

Results

We propose a unifying framework for characterizing network proximity measures for set-based queries. We observe that many existing measures are linear, in that the proximity of a node to a set of nodes can be represented as an aggregation of its proximity to the individual nodes in the set. Based on this observation, we propose methods for processing of set-based proximity queries that take advantage of sparse local proximity information. In addition, we provide an analytical framework for characterizing the distribution of proximity scores based on reference models that accurately capture the characteristics of the seed set (e.g. degree distribution and biological function). The resulting framework facilitates computation of exact figures for the statistical significance of network proximity scores, enabling assessment of the accuracy of Monte Carlo simulation based estimation methods.

Availability and Implementation

Implementations of the methods in this paper are available at https://bioengine.case.edu/crosstalker which includes a robust visualization for results viewing.

Contact

stm@case.edu or mxk331@case.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

Protein-protein interactions: detection, reliability assessment and applications

Protein-protein interactions (PPIs) participate in all important biological processes in living organisms, such as catalyzing metabolic reactions, DNA replication, DNA transcription, responding to stimuli and transporting molecules from one location to another. To reveal the function mechanisms in cells, it is important to identify PPIs that take place in the living organism. A large number of PPIs have been discovered by high-throughput experiments and computational methods. However, false-positive PPIs have been introduced too. Therefore, to obtain reliable PPIs, many computational methods have been proposed. Generally, these methods can be classified into two categories. One category includes the methods that are designed to determine new reliable PPIs. The other one is designed to assess the reliability of existing PPIs and filter out the unreliable ones. In this article, we review the two kinds of methods for detecting reliable PPIs, and then focus on evaluating the performance of some of these typical methods. Later on, we also enumerate several PPI network-based applications with taking a reliability assessment of the PPI data into consideration. Finally, we will discuss the challenges for obtaining reliable PPIs and future directions of the construction of reliable PPI networks. Our research will provide readers some guidance for choosing appropriate methods and features for obtaining reliable PPIs.

RedNemo: topology-based PPI network reconstruction via repeated diffusion with neighborhood modifications

Motivation

Analysis of protein-protein interaction (PPI) networks provides invaluable insight into several systems biology problems. High-throughput experimental techniques together with computational methods provide large-scale PPI networks. However, a major issue with these networks is their erroneous nature; they contain false-positive interactions and usually many more false-negatives. Recently, several computational methods have been proposed for network reconstruction based on topology, where given an input PPI network the goal is to reconstruct the network by identifying false-positives/-negatives as correctly as possible.

Results

We observe that the existing topology-based network reconstruction algorithms suffer several shortcomings. An important issue is regarding the scalability of their computational requirements, especially in terms of execution times, with the network sizes. They have only been tested on small-scale networks thus far and when applied on large-scale networks of popular PPI databases, the executions require unreasonable amounts of time, or may even crash without producing any output for some instances even after several months of execution. We provide an algorithm, RedNemo, for the topology-based network reconstruction problem. It provides more accurate networks than the alternatives as far as biological qualities measured in terms of most metrics based on gene ontology annotations. The recovery of a high-confidence network modified via random edge removals and rewirings is also better with RedNemo than with the alternatives under most of the experimented removal/rewiring ratios. Furthermore, through extensive tests on databases of varying sizes, we show that RedNemo achieves these results with much better running time performances.

Availability and Implementation

Supplementary material including source code, useful scripts, experimental data and the results are available at http://webprs.khas.edu.tr/~cesim/RedNemo.tar.gz.

Contact

cesim@khas.edu.tr.

Supplementary information

Supplementary data are available at Bioinformatics online.

Discovery of Bladder Cancer-related Genes Using Integrative Heterogeneous Network Modeling of Multi-omics Data

In human health, a fundamental challenge is the identification of disease-related genes. Bladder cancer (BC) is a worldwide malignant tumor, which has resulted in 170,000 deaths in 2010 up from 114,000 in 1990. Moreover, with the emergence of multi-omics data, more comprehensive analysis of human diseases become possible. In this study, we propose a multi-step approach for the identification of BC-related genes by using integrative Heterogeneous Network Modeling of Multi-Omics data (iHNMMO). The heterogeneous network model properly and comprehensively reflects the multiple kinds of relationships between genes in the multi-omics data of BC, including general relationships, unique relationships under BC condition, correlational relationships within each omics and regulatory relationships between different omics. Besides, a network-based propagation algorithm with resistance is utilized to quantize the relationships between genes and BC precisely. The results of comprehensive performance evaluation suggest that iHNMMO significantly outperforms other approaches. Moreover, further analysis suggests that the top ranked genes may be functionally implicated in BC, which also confirms the superiority of iHNMMO. In summary, this study shows that disease-related genes can be better identified through reasonable integration of multi-omics data.

Utilizing networks for differential analysis of chromatin interactions

Chromatin conformation capture with high-throughput sequencing (Hi-C) is a powerful technique to detect genome-wide chromatin interactions. In this paper, we introduce two novel approaches to detect differentially interacting genomic regions between two Hi-C experiments using a network model. To make input data from multiple experiments comparable, we propose a normalization strategy guided by network topological properties. We then devise two measurements, using local and global connectivity information from the chromatin interaction networks, respectively, to assess the interaction differences between two experiments. When multiple replicates are present in experiments, our approaches provide the flexibility for users to either pool all replicates together to therefore increase the network coverage, or to use the replicates in parallel to increase the signal to noise ratio. We show that while the local method works better in detecting changes from simulated networks, the global method performs better on real Hi-C data. The local and global methods, regardless of pooling, are always superior to two existing methods. Furthermore, our methods work well on both unweighted and weighted networks and our normalization strategy significantly improves the performance compared with raw networks without normalization. Therefore, we believe our methods will be useful for identifying differentially interacting genomic regions.

Network propagation: a universal amplifier of genetic associations

Biological networks are powerful resources for the discovery of genes and genetic modules that drive disease. Fundamental to network analysis is the concept that genes underlying the same phenotype tend to interact; this principle can be used to combine and to amplify signals from individual genes. Recently, numerous bioinformatic techniques have been proposed for genetic analysis using networks, based on random walks, information diffusion and electrical resistance. These approaches have been applied successfully to identify disease genes, genetic modules and drug targets. In fact, all these approaches are variations of a unifying mathematical machinery - network propagation - suggesting that it is a powerful data transformation method of broad utility in genetic research.

Mapping, modeling, and characterization of protein-protein interactions on a proteomic scale

Proteins effect a number of biological functions, from cellular signaling, organization, mobility, and transport to catalyzing biochemical reactions and coordinating an immune response. These varied functions are often dependent upon macromolecular interactions, particularly with other proteins. Small-scale studies in the scientific literature report protein-protein interactions (PPIs), but slowly and with bias towards well-studied proteins. In an era where genomic sequence is readily available, deducing genotype-phenotype relationships requires an understanding of protein connectivity at proteome-scale. A proteome-scale map of the protein-protein interaction network provides a global view of cellular organization and function. Here, we discuss a summary of methods for building proteome-scale interactome maps and the current status and implications of mapping achievements. Not only do interactome maps serve as a reference, detailing global cellular function and organization patterns, but they can also reveal the mechanisms altered by disease alleles, highlight the patterns of interaction rewiring across evolution, and help pinpoint biologically and therapeutically relevant proteins. Despite the considerable strides made in proteome-wide mapping, several technical challenges persist. Therefore, future considerations that impact current mapping efforts are also discussed.

A new two-stage method for revealing missing parts of edges in protein-protein interaction networks

With the increasing availability of high-throughput data, various computational methods have recently been developed for understanding the cell through protein-protein interaction (PPI) networks at a systems level. However, due to the incompleteness of the original PPI networks those efforts have been significantly hindered. In this paper, we propose a two stage method to predict underlying links between two originally unlinked protein pairs. First, we measure gene expression and gene functional similarly between unlinked protein pairs on Saccharomyces cerevisiae benchmark network and obtain new constructed networks. Then, we select the significant part of the new predicted links by analyzing the difference between essential proteins that have been identified based on the new constructed networks and the original network. Furthermore, we validate the performance of the new method by using the reliable and comprehensive PPI dataset obtained from the STRING database and compare the new proposed method with four other random walk-based methods. Comparing the results indicates that the new proposed strategy performs well in predicting underlying links. This study provides a general paradigm for predicting new interactions between protein pairs and offers new insights into identifying essential proteins.