An efficient method for protein function annotation based on multilayer protein networks

State of the art in protein-protein interactions within the fungi kingdom

Proteins rarely exert their function by themselves. Protein-protein interactions (PPIs) regulate virtually every biological process that takes place in a cell. Such interactions are targets for new therapeutic agents against all sorts of diseases, through the screening and design of a variety of inhibitors. Here we discuss several aspects of PPIs that contribute to prediction of protein function and drug discovery. As the high-throughput techniques continue to release biological data, targets for fungal therapeutics that rely on PPIs are being proposed worldwide. Computational approaches have reduced the time taken to develop new therapeutic approaches. The near future brings the possibility of developing new PPI and interaction network inhibitors and a revolution in the way we treat fungal diseases.

A systematic review of state-of-the-art strategies for machine learning-based protein function prediction

New drug discovery is inseparable from the discovery of drug targets, and the vast majority of the known targets are proteins. At the same time, proteins are essential structural and functional elements of living cells necessary for the maintenance of all forms of life. Therefore, protein functions have become the focus of many pharmacological and biological studies. Traditional experimental techniques are no longer adequate for rapidly growing annotation of protein sequences, and approaches to protein function prediction using computational methods have emerged and flourished. A significant trend has been to use machine learning to achieve this goal. In this review, approaches to protein function prediction based on the sequence, structure, protein-protein interaction (PPI) networks, and fusion of multi-information sources are discussed. The current status of research on protein function prediction using machine learning is considered, and existing challenges and prominent breakthroughs are discussed to provide ideas and methods for future studies.

Improving automatic GO annotation with semantic similarity

BACKGROUND

Automatic functional annotation of proteins is an open research problem in bioinformatics. The growing number of protein entries in public databases, for example in UniProtKB, poses challenges in manual functional annotation. Manual annotation requires expert human curators to search and read related research articles, interpret the results, and assign the annotations to the proteins. Thus, it is a time-consuming and expensive process. Therefore, designing computational tools to perform automatic annotation leveraging the high quality manual annotations that already exist in UniProtKB/SwissProt is an important research problem RESULTS: In this paper, we extend and adapt the GrAPFI (graph-based automatic protein function inference) (Sarker et al. in BMC Bioinform 21, 2020; Sarker et al., in: Proceedings of 7th international conference on complex networks and their applications, Cambridge, 2018) method for automatic annotation of proteins with gene ontology (GO) terms renaming it as GrAPFI-GO. The original GrAPFI method uses label propagation in a similarity graph where proteins are linked through the domains, families, and superfamilies that they share. Here, we also explore various types of similarity measures based on common neighbors in the graph. Moreover, GO terms are arranged in a hierarchical manner according to semantic parent-child relations. Therefore, we propose an efficient pruning and post-processing technique that integrates both semantic similarity and hierarchical relations between the GO terms. We produce experimental results comparing the GrAPFI-GO method with and without considering common neighbors similarity. We also test the performance of GrAPFI-GO and other annotation tools for GO annotation on a benchmark of proteins with and without the proposed pruning and post-processing procedure.

CONCLUSION

Our results show that the proposed semantic hierarchical post-processing potentially improves the performance of GrAPFI-GO and of other annotation tools as well. Thus, GrAPFI-GO exposes an original efficient and reusable procedure, to exploit the semantic relations among the GO terms in order to improve the automatic annotation of protein functions.

Rule-Based Pruning and In Silico Identification of Essential Proteins in Yeast PPIN

Proteins are vital for the significant cellular activities of living organisms. However, not all of them are essential. Identifying essential proteins through different biological experiments is relatively more laborious and time-consuming than the computational approaches used in recent times. However, practical implementation of conventional scientific methods sometimes becomes challenging due to poor performance impact in specific scenarios. Thus, more developed and efficient computational prediction models are required for essential protein identification. An effective methodology is proposed in this research, capable of predicting essential proteins in a refined yeast protein–protein interaction network (PPIN). The rule-based refinement is done using protein complex and local interaction density information derived from the neighborhood properties of proteins in the network. Identification and pruning of non-essential proteins are equally crucial here. In the initial phase, careful assessment is performed by applying node and edge weights to identify and discard the non-essential proteins from the interaction network. Three cut-off levels are considered for each node and edge weight for pruning the non-essential proteins. Once the PPIN has been filtered out, the second phase starts with two centralities-based approaches: (1) local interaction density (LID) and (2) local interaction density with protein complex (LIDC), which are successively implemented to identify the essential proteins in the yeast PPIN. Our proposed methodology achieves better performance in comparison to the existing state-of-the-art techniques.

An efficient strategy for identifying essential proteins based on homology, subcellular location and protein-protein interaction information

High throughput biological experiments are expensive and time consuming. For the past few years, many computational methods based on biological information have been proposed and widely used to understand the biological background. However, the processing of biological information data inevitably produces false positive and false negative data, such as the noise in the Protein-Protein Interaction (PPI) networks and the noise generated by the integration of a variety of biological information. How to solve these noise problems is the key role in essential protein predictions. An Identifying Essential Proteins model based on non-negative Matrix Symmetric tri-Factorization and multiple biological information (IEPMSF) is proposed in this paper, which utilizes only the PPI network proteins common neighbor characters to develop a weighted network, and uses the non-negative matrix symmetric tri-factorization method to find more potential interactions between proteins in the network so as to optimize the weighted network. Then, using the subcellular location and lineal homology information, the starting score of proteins is determined, and the random walk algorithm with restart mode is applied to the optimized network to mark and rank each protein. We tested the suggested forecasting model against current representative approaches using a public database. Experiment shows high efficiency of new method in essential proteins identification. The effectiveness of this method shows that it can dramatically solve the noise problems that existing in the multi-source biological information itself and cased by integrating them.

PFmulDL: a novel strategy enabling multi-class and multi-label protein function annotation by integrating diverse deep learning methods

Bioinformatic annotation of protein function is essential but extremely sophisticated, which asks for extensive efforts to develop effective prediction method. However, the existing methods tend to amplify the representativeness of the families with large number of proteins by misclassifying the proteins in the families with small number of proteins. That is to say, the ability of the existing methods to annotate proteins in the 'rare classes' remains limited. Herein, a new protein function annotation strategy, PFmulDL, integrating multiple deep learning methods, was thus constructed. First, the recurrent neural network was integrated, for the first time, with the convolutional neural network to facilitate the function annotation. Second, a transfer learning method was introduced to the model construction for further improving the prediction performances. Third, based on the latest data of Gene Ontology, the newly constructed model could annotate the largest number of protein families comparing with the existing methods. Finally, this newly constructed model was found capable of significantly elevating the prediction performance for the 'rare classes' without sacrificing that for the 'major classes'. All in all, due to the emerging requirements on improving the prediction performance for the proteins in 'rare classes', this new strategy would become an essential complement to the existing methods for protein function prediction. All the models and source codes are freely available and open to all users at: https://github.com/idrblab/PFmulDL.

High-Performance Deep Learning Toolbox for Genome-Scale Prediction of Protein Structure and Function

Computational biology is one of many scientific disciplines ripe for innovation and acceleration with the advent of high-performance computing (HPC). In recent years, the field of machine learning has also seen significant benefits from adopting HPC practices. In this work, we present a novel HPC pipeline that incorporates various machine-learning approaches for structure-based functional annotation of proteins on the scale of whole genomes. Our pipeline makes extensive use of deep learning and provides computational insights into best practices for training advanced deep-learning models for high-throughput data such as proteomics data. We showcase methodologies our pipeline currently supports and detail future tasks for our pipeline to envelop, including large-scale sequence comparison using SAdLSA and prediction of protein tertiary structures using AlphaFold2.

Machine Learning for Prediction of Drug Targets in Microbe Associated Cardiovascular Diseases by Incorporating Host-pathogen Interaction Network Parameters

Host-pathogen interactions play a crucial role in invasion, infection, and induction of immune response in humans. In this work, four machine learning algorithms, namely Logistic regression, K-nearest neighbor, Support Vector Machine, and Random Forest were implemented for the classification of drug targets. The algorithms were trained using 3400 hosts and 3800 pathogen drug and non-drug target proteins as learning instances. For each protein, 68 pathogen and 73 host features were computed that included sequence, structure, biological and host-pathogen network centrality characteristics. The Random Forest classifier model achieved the best accuracy after 10-fold cross-validation. 99 % accuracy was achieved with a ROC-AUC score of 0.99±0.01 for both pathogen and host training sets. The Eigenvector Centrality of host-pathogen interactions and host-host interactions was the top feature in performing classification of pathogen and host targets respectively. Other features important for classification were the presence of catalytic and binding sites, low instability/aliphatic index, and cellular location. The Random Forest classifier was then used for prediction of drug targets involved in Microbe Associated Cardiovascular Diseases. 331 host and 743 pathogen proteins were predicted as drug targets by the random forest model and can be validated experimentally for therapeutic intervention in Microbe Associated Cardiovascular Diseases.

Mantis: flexible and consensus-driven genome annotation

Background

The rapid development of the (meta-)omics fields has produced an unprecedented amount of high-resolution and high-fidelity data. Through the use of these datasets we can infer the role of previously functionally unannotated proteins from single organisms and consortia. In this context, protein function annotation can be described as the identification of regions of interest (i.e., domains) in protein sequences and the assignment of biological functions. Despite the existence of numerous tools, challenges remain in terms of speed, flexibility, and reproducibility. In the big data era, it is also increasingly important to cease limiting our findings to a single reference, coalescing knowledge from different data sources, and thus overcoming some limitations in overly relying on computationally generated data from single sources.

Results

We implemented a protein annotation tool, Mantis, which uses database identifiers intersection and text mining to integrate knowledge from multiple reference data sources into a single consensus-driven output. Mantis is flexible, allowing for the customization of reference data and execution parameters, and is reproducible across different research goals and user environments. We implemented a depth-first search algorithm for domain-specific annotation, which significantly improved annotation performance compared to sequence-wide annotation. The parallelized implementation of Mantis results in short runtimes while also outputting high coverage and high-quality protein function annotations.

Conclusions

Mantis is a protein function annotation tool that produces high-quality consensus-driven protein annotations. It is easy to set up, customize, and use, scaling from single genomes to large metagenomes. Mantis is available under the MIT license at https://github.com/PedroMTQ/mantis.

In silico annotation of unreviewed acetylcholinesterase (AChE) in some lepidopteran insect pest species reveals the causes of insecticide resistance

Lepidoptera is the second most diverse insect order outnumbered only by the Coeleptera. Acetylcholinesterase (AChE) is the major target site for insecticides. Extensive use of insecticides, to inhibit the function of this enzyme, have resulted in the development of insecticide resistance. Complete knowledge of the target proteins is very important to know the cause of resistance. Computational annotation of insect acetylcholinesterase can be helpful for the characterization of this important protein. Acetylcholinesterase of fourteen lepidopteran insect pest species was annotated by using different bioinformatics tools. AChE in all the species was hydrophilic and thermostable. All the species showed lower values for instability index except L. orbonalis, S. exigua and T. absoluta. Highest percentage of Arg, Asp, Asn, Gln and Cys were recorded in P. rapae. High percentage of Cys and Gln might be reason for insecticide resistance development in P. rapae. Phylogenetic analysis revealed the AChE in T. absoluta, L. orbonalis and S. exigua are closely related and emerged from same primary branch. Three functional motifs were predicted in eleven species while only two were found in L. orbonalis, S. exigua and T. absoluta. AChE in eleven species followed secretory pathway and have signal peptides. No signal peptides were predicted for S. exigua, L. orbonalis and T. absoluta and follow non secretory pathway. Arginine methylation and cysteine palmotylation was found in all species except S. exigua, L. orbonalis and T. absoluta. Glycosylphosphatidylinositol (GPI) anchor was predicted in only nine species.

NPF:network propagation for protein function prediction

Background

The accurate annotation of protein functions is of great significance in elucidating the phenomena of life, treating disease and developing new medicines. Various methods have been developed to facilitate the prediction of these functions by combining protein interaction networks (PINs) with multi-omics data. However, it is still challenging to make full use of multiple biological to improve the performance of functions annotation.

Results

We presented NPF (Network Propagation for Functions prediction), an integrative protein function predicting framework assisted by network propagation and functional module detection, for discovering interacting partners with similar functions to target proteins. NPF leverages knowledge of the protein interaction network architecture and multi-omics data, such as domain annotation and protein complex information, to augment protein-protein functional similarity in a propagation manner. We have verified the great potential of NPF for accurately inferring protein functions. According to the comprehensive evaluation of NPF, it delivered a better performance than other competing methods in terms of leave-one-out cross-validation and ten-fold cross validation.

Conclusions

We demonstrated that network propagation, together with multi-omics data, can both discover more partners with similar function, and is unconstricted by the “small-world” feature of protein interaction networks. We conclude that the performance of function prediction depends greatly on whether we can extract and exploit proper functional information of similarity from protein correlations.

Complete Topological Mapping of a Cellular Protein Interactome Reveals Bow-Tie Motifs as Ubiquitous Connectors of Protein Complexes

The network topology of a protein interactome is shaped by the function of each protein, making it a resource of functional knowledge in tissues and in single cells. Today, this resource is underused, as complete network topology characterization has proved difficult for large protein interactomes. We apply a matrix visualization and decoding approach to a physical protein interactome of a dendritic cell, thereby characterizing its topology with no prior assumptions of structure. We discover 294 proteins, each forming topological motifs called "bow-ties" that tie together the majority of observed protein complexes. The central proteins of these bow-ties have unique network properties, display multifunctional capabilities, are enriched for essential proteins, and are widely expressed in other cells and tissues. Collectively, the bow-tie motifs are a pervasive and previously unnoted topological trend in cellular interactomes. As such, these results provide fundamental knowledge on how intracellular protein connectivity is organized and operates.

A Novel Computational Approach for Identifying Essential Proteins From Multiplex Biological Networks

The identification of essential proteins can help in understanding the minimum requirements for cell survival and development. Ever-increasing amounts of high-throughput data provide us with opportunities to detect essential proteins from protein interaction networks (PINs). Existing network-based approaches are limited by the poor quality of the underlying PIN data, which exhibits high rates of false positive and false negative results. To overcome this problem, researchers have focused on the prediction of essential proteins by combining PINs with other biological data, which has led to the emergence of various interactions between proteins. It remains challenging, however, to use aggregated multiplex interactions within a single analysis framework to identify essential proteins. In this study, we created a multiplex biological network (MON) by initially integrating PINs, protein domains, and gene expression profiles. Next, we proposed a new approach to discover essential proteins by extending the random walk with restart algorithm to the tensor, which provides a data model representation of the MON. In contrast to existing approaches, the proposed MON approach considers for the importance of nodes and the different types of interactions between proteins during the iteration. MON was implemented to identify essential proteins within two yeast PINs. Our comprehensive experimental results demonstrated that MON outperformed 11 other state-of-the-art approaches in terms of precision-recall curve, jackknife curve, and other criteria.

GrAPFI: predicting enzymatic function of proteins from domain similarity graphs

An amendment to this paper has been published and can be accessed via the original article.

Predicting Protein Functions Based on Differential Co-expression and Neighborhood Analysis

Proteins are polypeptides essential in biological processes. Protein physical interactions are complemented by other types of functional relationship data including genetic interactions, knowledge about co-expression, and evolutionary pathways. Existing algorithms integrate protein interaction and gene expression data to retrieve context-specific subnetworks composed of genes/proteins with known and unknown functions. However, most protein function prediction algorithms fail to exploit diverse intrinsic information in feature and label spaces. We develop a novel integrative method based on differential Co-expression analysis and Neighbor-voting algorithm for Protein Function Prediction, namely CNPFP. The method integrates heterogeneous data and exploits intrinsic and latent linkages via global iterative approach and genomic features. CNPFP performs three tasks: clustering, differential co-expression analysis, and predicts protein functions. Our aim is to identify yeast cell cycle-specific proteins linked to differentially expressed proteins in the protein-protein interaction network. To capture intrinsic information, CNPFP selects the most relevant feature subset based on global iterative neighbor-voting algorithm. We identify eight condition-specific modules. The most relevant subnetwork has 87 genes highly enriched with cyclin-dependent kinases, a protein kinase relevant for cell cycle regulation. We present comprehensive annotations for 3538 Saccharomyces cerevisiae proteins. Our method achieves an AUROC of 0.9862, accuracy of 0.9710, and F-score of 0.9691. From the results, we can summarize that exploiting intrinsic nature of protein relationships improves the quality of function prediction. Thus, the proposed method is useful in functional genomics studies.

A novel method to predict essential proteins based on tensor and HITS algorithm

BACKGROUND

Essential proteins are an important part of the cell and closely related to the life activities of the cell. Hitherto, Protein-Protein Interaction (PPI) networks have been adopted by many computational methods to predict essential proteins. Most of the current approaches focus mainly on the topological structure of PPI networks. However, those methods relying solely on the PPI network have low detection accuracy for essential proteins. Therefore, it is necessary to integrate the PPI network with other biological information to identify essential proteins.

RESULTS

In this paper, we proposed a novel random walk method for identifying essential proteins, called HEPT. A three-dimensional tensor is constructed first by combining the PPI network of Saccharomyces cerevisiae with multiple biological data such as gene ontology annotations and protein domains. Then, based on the newly constructed tensor, we extended the Hyperlink-Induced Topic Search (HITS) algorithm from a two-dimensional to a three-dimensional tensor model that can be utilized to infer essential proteins. Different from existing state-of-the-art methods, the importance of proteins and the types of interactions will both contribute to the essential protein prediction. To evaluate the performance of our newly proposed HEPT method, proteins are ranked in the descending order based on their ranking scores computed by our method and other competitive methods. After that, a certain number of the ranked proteins are selected as candidates for essential proteins. According to the list of known essential proteins, the number of true essential proteins is used to judge the performance of each method. Experimental results show that our method can achieve better prediction performance in comparison with other nine state-of-the-art methods in identifying essential proteins.

CONCLUSIONS

Through analysis and experimental results, it is obvious that HEPT can be used to effectively improve the prediction accuracy of essential proteins by the use of HITS algorithm and the combination of network topology with gene ontology annotations and protein domains, which provides a new insight into multi-data source fusion.

Protein functional annotation of simultaneously improved stability, accuracy and false discovery rate achieved by a sequence-based deep learning

Functional annotation of protein sequence with high accuracy has become one of the most important issues in modern biomedical studies, and computational approaches of significantly accelerated analysis process and enhanced accuracy are greatly desired. Although a variety of methods have been developed to elevate protein annotation accuracy, their ability in controlling false annotation rates remains either limited or not systematically evaluated. In this study, a protein encoding strategy, together with a deep learning algorithm, was proposed to control the false discovery rate in protein function annotation, and its performances were systematically compared with that of the traditional similarity-based and de novo approaches. Based on a comprehensive assessment from multiple perspectives, the proposed strategy and algorithm were found to perform better in both prediction stability and annotation accuracy compared with other de novo methods. Moreover, an in-depth assessment revealed that it possessed an improved capacity of controlling the false discovery rate compared with traditional methods. All in all, this study not only provided a comprehensive analysis on the performances of the newly proposed strategy but also provided a tool for the researcher in the fields of protein function annotation.

Clustering multilayer omics data using MuNCut

BACKGROUND

Omics profiling is now a routine component of biomedical studies. In the analysis of omics data, clustering is an essential step and serves multiple purposes including for example revealing the unknown functionalities of omics units, assisting dimension reduction in outcome model building, and others. In the most recent omics studies, a prominent trend is to conduct multilayer profiling, which collects multiple types of genetic, genomic, epigenetic and other measurements on the same subjects. In the literature, clustering methods tailored to multilayer omics data are still limited. Directly applying the existing clustering methods to multilayer omics data and clustering each layer first and then combing across layers are both "suboptimal" in that they do not accommodate the interconnections within layers and across layers in an informative way.

METHODS

In this study, we develop the MuNCut (Multilayer NCut) clustering approach. It is tailored to multilayer omics data and sufficiently accounts for both across- and within-layer connections. It is based on the novel NCut technique and also takes advantages of regularized sparse estimation. It has an intuitive formulation and is computationally very feasible. To facilitate implementation, we develop the function muncut in the R package NcutYX.

RESULTS

Under a wide spectrum of simulation settings, it outperforms competitors. The analysis of TCGA (The Cancer Genome Atlas) data on breast cancer and cervical cancer shows that MuNCut generates biologically meaningful results which differ from those using the alternatives.

CONCLUSIONS

We propose a more effective clustering analysis of multiple omics data. It provides a new venue for jointly analyzing genetic, genomic, epigenetic and other measurements.