Protein-protein Interaction Datasets Research Articles

Predicting protein complexes in protein interaction networks using Mapper and graph convolution networks

Protein complexes are groups of interacting proteins that are central to multiple biological processes. Studying protein complexes can enhance our understanding of cellular functions and malfunctions and thus support the development of effective disease treatments. High-throughput experimental techniques allow the generation of large-scale protein-protein interaction datasets. Accordingly, various computational approaches to predict protein complexes from protein-protein interactions were presented in the literature. They are typically based on networks in which nodes and edges represent proteins and their interactions, respectively. State-of-the-art approaches mainly rely on clustering static networks to identify complexes. However, since protein interactions are highly dynamic in nature, recent approaches seek to model such dynamics by typically integrating gene expression data and identifying protein complexes accordingly. We propose MComplex, a method that utilizes time-series gene expression with interaction data to generate a temporal network which is passed to a generative adversarial network whose generator is a graph convolutional network. This creates embeddings which are then analyzed using a modified graph-based version of the Mapper algorithm to predict corresponding protein complexes. We test our approach on multiple benchmark datasets and compare identified complexes against gold-standard protein complex datasets. Our results show that MComplex outperforms existing methods in several evaluation aspects, namely recall and maximum matching ratio as well as a composite score covering aggregated evaluation measures. The code and data are available for free download from https://github.com/LeonardoDaou/MComplex.

Anti-symmetric framework for balanced learning of protein-protein interactions.

Protein-protein interactions (PPIs) are essential for the regulation and facilitation of virtually all biological processes. Computational tools, particularly those based on deep learning, are preferred for the efficient prediction of PPIs. Despite recent progress, two challenges remain unresolved: (i) the imbalanced nature of PPI characteristics is often ignored and (ii) there exists a high computational cost associated with capturing long-range dependencies within protein data, typically exhibiting quadratic complexity relative to the length of the protein sequence. Here, we propose an anti-symmetric graph learning model, BaPPI, for the balanced prediction of PPIs and extrapolation of the involved patterns in PPI network. In BaPPI, the contextualized information of protein data is efficiently handled by an attention-free mechanism formed by recurrent convolution operator. The anti-symmetric graph convolutional network is employed to model the uneven distribution within PPI networks, aiming to learn a more robust and balanced representation of the relationships between proteins. Ultimately, the model is updated using asymmetric loss. The experimental results on classical baseline datasets demonstrate that BaPPI outperforms four state-of-the-art PPI prediction methods. In terms of Micro-F1, BaPPI exceeds the second-best method by 6.5% on SHS27K and 5.3% on SHS148K. Further analysis of the generalization ability and patterns of predicted PPIs also demonstrates our model's generalizability and robustness to the imbalanced nature of PPI datasets. The source code of this work is publicly available at https://github.com/ttan6729/BaPPI.

Effectiveness and Efficiency: Label-Aware Hierarchical Subgraph Learning for Protein–Protein Interaction

The study of protein–protein interactions (PPIs) holds immense significance in understanding various biological activities, as well as in drug discovery and disease diagnosis. Existing deep learning methods for PPI prediction, including graph neural networks (GNNs), have been widely employed as the solutions, while they often experience a decline in performance in the real world. We claim that the topological shortcut is one of the key problems contributing negatively to the performance, according to our analysis. By modeling the PPIs as a graph with protein as nodes and interactions as edge types, the prevailing models tend to learn the pattern of nodes’ degrees rather than intrinsic sequence-structure profiles, leading to the problem termed topological shortcut. The huge data growth of PPI leads to intensive computational costs and challenges computing devices, causing infeasibility in practice. To address the discussed problems, we propose a label-aware hierarchical subgraph learning method (laruGL-PPI) that can effectively infer PPIs while being interpretable. Specifically, we introduced edge-based subgraph sampling to effectively alleviate the problems of topological shortcuts and high computing costs. Besides, the inner-outer connections of PPIs are modeled as a hierarchical graph, together with the dependencies between interaction types constructed by a label graph. Extensive experiments conducted across various scales of PPI datasets have conclusively demonstrated that the laruGL-PPI method surpasses the most advanced PPI prediction techniques currently available, particularly in the testing of unseen proteins. Also, our model can recognize crucial sites of proteins, such as surface sites for binding and active sites for catalysis.

Integrative Bioinformatics Analysis for Targeting Hub Genes in Hepatocellular Carcinoma Treatment

Background: The damage in the liver and hepatocytes is where the primary liver cancer begins, and this is referred to as Hepatocellular Carcinoma (HCC). One of the best methods for detecting changes in gene expression of hepatocellular carcinoma is through bioinformatics approaches. Objective: This study aimed to identify potential drug target(s) hubs mediating HCC progression using computational approaches through gene expression and protein-protein interaction datasets. Methodology: Four datasets related to HCC were acquired from the GEO database, and Differentially Expressed Genes (DEGs) were identified. Using Evenn, the common genes were chosen. Using the Fun Rich tool, functional associations among the genes were identified. Further, protein-protein interaction networks were predicted using STRING, and hub genes were identified using Cytoscape. The selected hub genes were subjected to GEPIA and Shiny GO analysis for survival analysis and functional enrichment studies for the identified hub genes. The up-regulating genes were further studied for immunohistopathological studies using HPA to identify gene/protein expression in normal vs HCC conditions. Drug Bank and Drug Gene Interaction Database were employed to find the reported drug status and targets. Finally, STITCH was performed to identify the functional association between the drugs and the identified hub genes. Results: The GEO2R analysis for the considered datasets identified 735 upregulating and 284 downregulating DEGs. Functional gene associations were identified through the Fun Rich tool. Further, PPIN network analysis was performed using STRING. A comparative study was carried out between the experimental evidence and the other seven data evidence in STRING, revealing that most proteins in the network were involved in protein-protein interactions. Further, through Cytoscape plugins, the ranking of the genes was analyzed, and densely connected regions were identified, resulting in the selection of the top 20 hub genes involved in HCC pathogenesis. The identified hub genes were: KIF2C, CDK1, TPX2, CEP55, MELK, TTK, BUB1, NCAPG, ASPM, KIF11, CCNA2, HMMR, BUB1B, TOP2A, CENPF, KIF20A, NUSAP1, DLGAP5, PBK, and CCNB2. Further, GEPIA and Shiny GO analyses provided insights into survival ratios and functional enrichment studied for the hub genes. The HPA database studies further found that upregulating genes were involved in changes in protein expression in Normal vs HCC tissues. These findings indicated that hub genes were certainly involved in the progression of HCC. STITCH database studies uncovered that existing drug molecules, including sorafenib, regorafenib, cabozantinib, and lenvatinib, could be used as leads to identify novel drugs, and identified hub genes could also be considered as potential and promising drug targets as they are involved in the gene-chemical interaction networks. Conclusion: The present study involved various integrated bioinformatics approaches, analyzing gene expression and protein-protein interaction datasets, resulting in the identification of 20 topranked hubs involved in the progression of HCC. They are KIF2C, CDK1, TPX2, CEP55, MELK, TTK, BUB1, NCAPG, ASPM, KIF11, CCNA2, HMMR, BUB1B, TOP2A, CENPF, KIF20A, NUSAP1, DLGAP5, PBK, and CCNB2. Gene-chemical interaction network studies uncovered that existing drug molecules, including sorafenib, regorafenib, cabozantinib, and lenvatinib, can be used as leads to identify novel drugs, and the identified hub genes can be promising drug targets. The current study underscores the significance of targeting these hub genes and utilizing existing molecules to generate new molecules to combat liver cancer effectively and can be further explored in terms of drug discovery research to develop treatments for HCC.

Co-training based prediction of multi-label protein–protein interactions

Prediction of protein–protein interaction (PPI) types enhances the comprehension of the underlying structural characteristics and functions of proteins, which gives rise to a multi-label classification problem. The nominal features describe the physicochemical characteristics of proteins directly, establishing a more robust correlation with the interaction types between proteins than ordered features. Motivated by this, we propose a multi-label PPI prediction model referred to as CoMPPI (Co-training based Multi-Label prediction of Protein–Protein Interaction). This approach aims to maximize the utility of both ordered and nominal features extracted from protein sequences. Specifically, CoMPPI incorporates graph convolutional network (GCN) and 1D convolution operation to process the complementary subsets of features individually, leveraging both local and contextualized information in a more efficient way. In addition, two multi-type PPI datasets were constructed to eliminate the duplication in previous datasets. We compare the performance of CoMPPI with three state-of-the-art methods on three datasets partitioned using distinct schemes (Breadth-first search, Depth-first search, and Random), CoMPPI consistently outperforms the other methods across all cases, demonstrating improvements ranging from 3.81% to 32.40% in Micro-F1. The subsequent ablation experiment confirms the efficacy of employing the co-training framework for multi-label PPI prediction, indicating promising avenues for future advancements in this domain.

A comprehensive dataset of protein-protein interactions and ligand binding pockets for advancing drug discovery

This dataset represents a collection of pocket-centric structural data related to protein-protein interactions (PPIs) and PPI-related ligand binding sites. The dataset includes high-quality structural information on more than 23,000 pockets, 3,700 proteins on more than 500 organisms, and nearly 3500 ligands that can aid researchers in the fields of bioinformatics, structural biology, and drug discovery. It encompasses a diverse set of PPI complexes with more than 1,700 unique protein families including some with associated ligands, enabling detailed investigations into molecular interactions at the atomic level. This article introduces an indispensable resource designed to unlock the full potential of PPIs while pioneering a novel metric for pocket similarity for hypothesizing protein partners repurposing.

Protein–protein interaction network-based integration of GWAS and functional data for blood pressure regulation analysis

BackgroundIt is valuable to analyze the genome-wide association studies (GWAS) data for a complex disease phenotype in the context of the protein–protein interaction (PPI) network, as the related pathophysiology results from the function of interacting polyprotein pathways. The analysis may include the design and curation of a phenotype-specific GWAS meta-database incorporating genotypic and eQTL data linking to PPI and other biological datasets, and the development of systematic workflows for PPI network-based data integration toward protein and pathway prioritization. Here, we pursued this analysis for blood pressure (BP) regulation.MethodsThe relational scheme of the implemented in Microsoft SQL Server BP-GWAS meta-database enabled the combined storage of: GWAS data and attributes mined from GWAS Catalog and the literature, Ensembl-defined SNP-transcript associations, and GTEx eQTL data. The BP-protein interactome was reconstructed from the PICKLE PPI meta-database, extending the GWAS-deduced network with the shortest paths connecting all GWAS-proteins into one component. The shortest-path intermediates were considered as BP-related. For protein prioritization, we combined a new integrated GWAS-based scoring scheme with two network-based criteria: one considering the protein role in the reconstructed by shortest-path (RbSP) interactome and one novel promoting the common neighbors of GWAS-prioritized proteins. Prioritized proteins were ranked by the number of satisfied criteria.ResultsThe meta-database includes 6687 variants linked with 1167 BP-associated protein-coding genes. The GWAS-deduced PPI network includes 1065 proteins, with 672 forming a connected component. The RbSP interactome contains 1443 additional, network-deduced proteins and indicated that essentially all BP-GWAS proteins are at most second neighbors. The prioritized BP-protein set was derived from the union of the most BP-significant by any of the GWAS-based or the network-based criteria. It included 335 proteins, with ~ 2/3 deduced from the BP PPI network extension and 126 prioritized by at least two criteria. ESR1 was the only protein satisfying all three criteria, followed in the top-10 by INSR, PTN11, CDK6, CSK, NOS3, SH2B3, ATP2B1, FES and FINC, satisfying two. Pathway analysis of the RbSP interactome revealed numerous bioprocesses, which are indeed functionally supported as BP-associated, extending our understanding about BP regulation.ConclusionsThe implemented workflow could be used for other multifactorial diseases.

Multi-modal features-based human-herpesvirus protein-protein interaction prediction by using LightGBM.

The identification of human-herpesvirus protein-protein interactions (PPIs) is an essential and important entry point to understand the mechanisms of viral infection, especially in malignant tumor patients with common herpesvirus infection. While natural language processing (NLP)-based embedding techniques have emerged as powerful approaches, the application of multi-modal embedding feature fusion to predict human-herpesvirus PPIs is still limited. Here, we established a multi-modal embedding feature fusion-based LightGBM method to predict human-herpesvirus PPIs. In particular, we applied document and graph embedding approaches to represent sequence, network and function modal features of human and herpesviral proteins. Training our LightGBM models through our compiled non-rigorous and rigorous benchmarking datasets, we obtained significantly better performance compared to individual-modal features. Furthermore, our model outperformed traditional feature encodings-based machine learning methods and state-of-the-art deep learning-based methods using various benchmarking datasets. In a transfer learning step, we show that our model that was trained on human-herpesvirus PPI dataset without cytomegalovirus data can reliably predict human-cytomegalovirus PPIs, indicating that our method can comprehensively capture multi-modal fusion features of protein interactions across various herpesvirus subtypes. The implementation of our method is available at https://github.com/XiaodiYangpku/MultimodalPPI/.

Proteome-wide assessment of human interactome as a source of capturing domain-motif and domain-domain interactions.

Protein-protein interactions (PPIs) play a crucial role in various biological processes by establishing domain-motif (DMI) and domain-domain interactions (DDIs). While the existence of real DMIs/DDIs is generally assumed, it is rarely tested; therefore, this study extensively compared high-throughput methods and public PPI repositories as sources for DMI and DDI prediction based on the assumption that the human interactome provides sufficient data for the reliable identification of DMIs and DDIs. Different datasets from leading high-throughput methods (Yeast two-hybrid [Y2H], Affinity Purification coupled Mass Spectrometry [AP-MS], and Co-fractionation-coupled Mass Spectrometry) were assessed for their ability to capture DMIs and DDIs using known DMI/DDI information. High-throughput methods were not notably worse than PPI databases and, in some cases, appeared better. In conclusion, all PPI datasets demonstrated significant enrichment in DMIs and DDIs (p-value <0.001), establishing Y2H and AP-MS as reliable methods for predicting these interactions. This study provides valuable insights for biologists in selecting appropriate methods for predicting DMIs, ultimately aiding in SLiM discovery.

An End-to-end Knowledge Graph Fused Graph Neural Network for Accurate Protein-Protein Interactions Prediction.

Protein-protein interactions (PPIs) are essential to understanding cellular mechanisms, signaling networks, disease processes, and drug development, as they represent the physical contacts and functional associations between proteins. Recent advances have witnessed the achievements of artificial intelligence (AI) methods aimed at predicting PPIs. However, these approaches often handle the intricate web of relationships and mechanisms among proteins, drugs, diseases, ribonucleic acid (RNA), and protein structures in a fragmented or superficial manner. This is typically due to the limitations of non-end-to-end learning frameworks, which can lead to sub-optimal feature extraction and fusion, thereby compromising the prediction accuracy. To address these deficiencies, this paper introduces a novel end-to-end learning model, the Knowledge Graph Fused Graph Neural Network (KGF-GNN). This model comprises three integral components: (1) Protein Associated Network (PAN) Construction: We begin by constructing a PAN that extensively captures the diverse relationships and mechanisms linking proteins with drugs, diseases, RNA, and protein structures. (2) Graph Neural Network for Feature Extraction: A Graph Neural Network (GNN) is then employed to distill both topological and semantic features from the PAN, alongside another GNN designed to extract topological features directly from observed PPI networks. (3) Multi-layer Perceptron for Feature Fusion: Finally, a multi-layer perceptron integrates these varied features through end-to-end learning, ensuring that the feature extraction and fusion processes are both comprehensive and optimized for PPI prediction. Extensive experiments conducted on real-world PPI datasets validate the effectiveness of our proposed KGF-GNN approach, which not only achieves high accuracy in predicting PPIs but also significantly surpasses existing state-of-the-art models. This work not only enhances our ability to predict PPIs with a higher precision but also contributes to the broader application of AI in Bioinformatics, offering profound implications for biological research and therapeutic development.

Identification of the communal pathogenesis and immune landscape between viral myocarditis and dilated cardiomyopathy.

Studies have confirmed that viral myocarditis (VMC) is one of the risk factors for dilated cardiomyopathy (DCM). The molecular mechanisms underlying the progression from VMC to DCM remain unclear and require further investigation. The mRNA microarray datasets GSE57338 (DCM) and GSE1145 (VMC) were obtained from the Gene Expression Omnibus database. The candidate key genes were further screened using weighted correlation network analysis (WGCNA), protein-protein interaction and external dataset validation, and the correlation between the candidate key genes and immune cells and the signalling pathways of the candidate key genes were observed by enrichment analysis and immune infiltration analysis. The expression of key genes was validated in the external dataset GSE35182. The crosstalk genes between DCM and VMC were mainly enriched in 'transcriptional misregulation in cancer', 'FoxO signalling pathway', 'AGE-RAGE signalling pathway in diabetic complications', 'thyroid hormone signalling pathway', 'AMPK signalling pathway', and other signalling pathways. The immune infiltration analysis indicated that VMC was mainly associated with resting dendritic cells and M0 macrophages, while DCM was mainly associated with monocytes, M0 macrophages, CD8+ T cells, resting CD4 memory T cells, naive CD4+ T cells, and resting mast cells. In DCM-related dataset GSE57338 and VMC-related dataset GSE1145, a total of 18 candidate key genes were differentially expressed. BLC6, FOXO1, and UBE2M were identified as the key genes that lead to the progression from VMC to DCM by GSE35182. Three key genes (BLC6, FOXO1, and UBE2M) were identified and provided new insights into the diagnosis and treatment of VMC with DCM.

Prediction of Protein-Protein Interactions Using Vision Transformer and Language Model.

The knowledge of protein-protein interaction (PPI) helps us to understand proteins' functions, the causes and growth of several diseases, and can aid in designing new drugs. The majority of existing PPI research has relied mainly on sequence-based approaches. With the availability of multi-omics datasets (sequence, 3D structure) and advancements in deep learning techniques, it is feasible to develop a deep multi-modal framework that fuses the features learned from different sources of information to predict PPI. In this work, we propose a multi-modal approach utilizing protein sequence and 3D structure. To extract features from the 3D structure of proteins, we use a pre-trained vision transformer model that has been fine-tuned on the structural representation of proteins. The protein sequence is encoded into a feature vector using a pre-trained language model. The feature vectors extracted from the two modalities are fused and then fed to the neural network classifier to predict the protein interactions. To showcase the effectiveness of the proposed methodology, we conduct experiments on two popular PPI datasets, namely, the human dataset and the S. cerevisiae dataset. Our approach outperforms the existing methodologies to predict PPI, including multi-modal approaches. We also evaluate the contributions of each modality by designing uni-modal baselines. We perform experiments with three modalities as well, having gene ontology as the third modality.

Predicting Protein-Protein Interactions Using Sequence and Network Information via Variational Graph Autoencoder.

Protein-protein interactions (PPIs) play a critical role in the proteomics study, and a variety of computational algorithms have been developed to predict PPIs. Though effective, their performance is constrained by high false-positive and false-negative rates observed in PPI data. To overcome this problem, a novel PPI prediction algorithm, namely PASNVGA, is proposed in this work by combining the sequence and network information of proteins via variational graph autoencoder. To do so, PASNVGA first applies different strategies to extract the features of proteins from their sequence and network information, and obtains a more compact form of these features using principal component analysis. In addition, PASNVGA designs a scoring function to measure the higher-order connectivity between proteins and so as to obtain a higher-order adjacency matrix. With all these features and adjacency matrices, PASNVGA trains a variational graph autoencoder model to further learn the integrated embeddings of proteins. The prediction task is then completed by using a simple feedforward neural network. Extensive experiments have been conducted on five PPI datasets collected from different species. Compared with several state-of-the-art algorithms, PASNVGA has been demonstrated as a promising PPI prediction algorithm.

The genomic landscape of mammal domestication might be orchestrated by selected transcription factors regulating brain and craniofacial development

Domestication transforms once wild animals into tamed animals that can be then exploited by humans. The process entails modifications in the body, cognition, and behavior that are essentially driven by differences in gene expression patterns. Although genetic and epigenetic mechanisms were shown to underlie such differences, less is known about the role exerted by trans-regulatory molecules, notably transcription factors (TFs) in domestication. In this paper, we conducted extensive in silico analyses aimed to clarify the TF landscape of mammal domestication. We first searched the literature, so as to establish a large list of genes selected with domestication in mammals. From this list, we selected genes experimentally demonstrated to exhibit TF functions. We also considered TFs displaying a statistically significant number of targets among the entire list of (domestication) selected genes. This workflow allowed us to identify 5 candidate TFs (SOX2, KLF4, MITF, NR3C1, NR3C2) that were further assessed in terms of biochemical and functional properties. We found that such TFs-of-interest related to mammal domestication are all significantly involved in the development of the brain and the craniofacial region, as well as the immune response and lipid metabolism. A ranking strategy, essentially based on a survey of protein-protein interactions datasets, allowed us to identify SOX2 as the main candidate TF involved in domestication-associated evolutionary changes. These findings should help to clarify the molecular mechanics of domestication and are of interest for future studies aimed to understand the behavioral and cognitive changes associated to domestication.

Topological feature generation for link prediction in biological networks.

Graph or network embedding is a powerful method for extracting missing or potential information from interactions between nodes in biological networks. Graph embedding methods learn representations of nodes and interactions in a graph with low-dimensional vectors, which facilitates research to predict potential interactions in networks. However, most graph embedding methods suffer from high computational costs in the form of high computational complexity of the embedding methods and learning times of the classifier, as well as the high dimensionality of complex biological networks. To address these challenges, in this study, we use the Chopper algorithm as an alternative approach to graph embedding, which accelerates the iterative processes and thus reduces the running time of the iterative algorithms for three different (nervous system, blood, heart) undirected protein-protein interaction (PPI) networks. Due to the high dimensionality of the matrix obtained after the embedding process, the data are transformed into a smaller representation by applying feature regularization techniques. We evaluated the performance of the proposed method by comparing it with state-of-the-art methods. Extensive experiments demonstrate that the proposed approach reduces the learning time of the classifier and performs better in link prediction. We have also shown that the proposed embedding method is faster than state-of-the-art methods on three different PPI datasets.

A correlation coefficient-based feature selection approach for virus-host protein-protein interaction prediction.

Prediction of virus-host protein-protein interactions (PPI) is a broad research area where various machine-learning-based classifiers are developed. Transforming biological data into machine-usable features is a preliminary step in constructing these virus-host PPI prediction tools. In this study, we have adopted a virus-host PPI dataset and a reduced amino acids alphabet to create tripeptide features and introduced a correlation coefficient-based feature selection. We applied feature selection across several correlation coefficient metrics and statistically tested their relevance in a structural context. We compared the performance of feature-selection models against that of the baseline virus-host PPI prediction models created using different classification algorithms without the feature selection. We also tested the performance of these baseline models against the previously available tools to ensure their predictive power is acceptable. Here, the Pearson coefficient provides the best performance with respect to the baseline model as measured by AUPR; a drop of 0.003 in AUPR while achieving a 73.3% (from 686 to 183) reduction in the number of tripeptides features for random forest. The results suggest our correlation coefficient-based feature selection approach, while decreasing the computation time and space complexity, has a limited impact on the prediction performance of virus-host PPI prediction tools.

CAA-PPI: A Computational Feature Design to Predict Protein–Protein Interactions Using Different Encoding Strategies

Protein–protein interactions (PPIs) are involved in an extensive variety of biological procedures, including cell-to-cell interactions, and metabolic and developmental control. PPIs are becoming one of the most important aims of system biology. PPIs act as a fundamental part in predicting the protein function of the target protein and the drug ability of molecules. An abundance of work has been performed to develop methods to computationally predict PPIs as this supplements laboratory trials and offers a cost-effective way of predicting the most likely set of interactions at the entire proteome scale. This article presents an innovative feature representation method (CAA-PPI) to extract features from protein sequences using two different encoding strategies followed by an ensemble learning method. The random forest methodwas used as a classifier for PPI prediction. CAA-PPI considers the role of the trigram and bond of a given amino acid with its nearby ones. The proposed PPI model achieved more than a 98% prediction accuracy with one encoding scheme and more than a 95% prediction accuracy with another encoding scheme for the two diverse PPI datasets, i.e., H. pylori and Yeast. Further, investigations were performed to compare the CAA-PPI approach with existing sequence-based methods and revealed the proficiency of the proposed method with both encoding strategies. To further assess the practical prediction competence, a blind test was implemented on five other species’ datasets independent of the training set, and the obtained results ascertained the productivity of CAA-PPI with both encoding schemes.

Link Prediction with Continuous-Time Classical and Quantum Walks.

Protein-protein interaction (PPI) networks consist of the physical and/or functional interactions between the proteins of an organism, and they form the basis for the field of network medicine. Since the biophysical and high-throughput methods used to form PPI networks are expensive, time-consuming, and often contain inaccuracies, the resulting networks are usually incomplete. In order to infer missing interactions in these networks, we propose a novel class of link prediction methods based on continuous-time classical and quantum walks. In the case of quantum walks, we examine the usage of both the network adjacency and Laplacian matrices for specifying the walk dynamics. We define a score function based on the corresponding transition probabilities and perform tests on six real-world PPI datasets. Our results show that continuous-time classical random walks and quantum walks using the network adjacency matrix can successfully predict missing protein-protein interactions, with performance rivalling the state-of-the-art.

A novel global clustering coefficient-dependent degree centrality (GCCDC) metric for large network analysis using real-world datasets

Nowadays, it is imperative to identify the combination of profitable products (nodes) within large product networks. Data exploration for such broad-spectrum product networks require sophisticated techniques for their analysis and meaningful inferences. The most commonly used techniques are the centrality metrics due to their efficiency in computation. Centrality metrics include Degree Centrality (DC), Closeness Centrality (CC), Betweenness Centrality (BC), Eigenvector Centrality (EVC), Katz Centrality (KC), and the local clustering coefficient-dependent degree centrality (LCCDC or LD). In this research, a novel approach the global clustering coefficient-dependent degree centrality (GCCDC or GD) method has been formulated for the analysis of links of the profitable products (nodes) in a large product network. GCCDC or GD is formulated by using the global clustering coefficient method which is efficient and accurate for large product networks such as product Amazon network. Furthermore, three correlation coefficients that are Pearson’s, Spearman’s, and Kendall’s have been used for evaluation. The results have shown that GD is preferable over LD to avoid uncertainties in computation of results for real-world datasets. To prove the scalability of the novel method, a dataset from different domain (biological yeast protein-protein interaction (PPI) dataset) was also analyzed using similar metrics and shown improved results.

Graph-BERT and language model-based framework for protein–protein interaction identification

Identification of protein–protein interactions (PPI) is among the critical problems in the domain of bioinformatics. Previous studies have utilized different AI-based models for PPI classification with advances in artificial intelligence (AI) techniques. The input to these models is the features extracted from different sources of protein information, mainly sequence-derived features. In this work, we present an AI-based PPI identification model utilizing a PPI network and protein sequences. The PPI network is represented as a graph where each node is a protein pair, and an edge is defined between two nodes if there exists a common protein between these nodes. Each node in a graph has a feature vector. In this work, we have used the language model to extract feature vectors directly from protein sequences. The feature vectors for protein in pairs are concatenated and used as a node feature vector of a PPI network graph. Finally, we have used the Graph-BERT model to encode the PPI network graph with sequence-based features and learn the hidden representation of the feature vector for each node. The next step involves feeding the learned representations of nodes to the fully connected layer, the output of which is fed into the softmax layer to classify the protein interactions. To assess the efficacy of the proposed PPI model, we have performed experiments on several PPI datasets. The experimental results demonstrate that the proposed approach surpasses the existing PPI works and designed baselines in classifying PPI.