Full Characterization of Localization Diversity in the Human Protein Interactome

Special Issue "Deployment of Proteomics Approaches in Biomedical Research"

Many angles of personalized medicine, such as diagnostic improvements, systems biology [...].

Computational Protein Design - Where it goes?

Proteins have been playing a critical role in the regulation of diverse biological processes related to human life. With the increasing demand, functional proteins are sparse in this immense sequence space. Therefore, protein design has become an important task in various fields, including medicine, food, energy, materials, etc. Directed evolution has recently led to significant achievements. Molecular modification of proteins through directed evolution technology has significantly advanced the fields of enzyme engineering, metabolic engineering, medicine, and beyond. However, it is impossible to identify desirable sequences from a large number of synthetic sequences alone. As a result, computational methods, including data-driven machine learning and physics-based molecular modeling, have been introduced to protein engineering to produce more functional proteins. This review focuses on recent advances in computational protein design, highlighting the applicability of different approaches as well as their limitations.

Co-expression module analysis reveals high expression homogeneity for both coding and non-coding genes in sepsis

Sepsis is a life-threatening condition characterized by a harmful host response to infection with organ dysfunction. Annually about 20 million people are dead owing to sepsis and its mortality rates is as high as 20%. However, no studies have been carried out to investigate sepsis from the system biology point of view, as previous research predominantly focused on individual genes without considering their interactions and associations. Here, we conducted a comprehensive exploration of genome-wide expression alterations in both mRNAs and long non-coding RNAs (lncRNAs) in sepsis, using six microarray datasets. Co-expression networks were conducted to identify mRNA and lncRNA modules, respectively. Comparing these sepsis modules with normal modules, we observed a homogeneous expression pattern within the mRNA/lncRNA members, with the majority of them displaying consistent expression direction. Moreover, we identified consistent modules across diverse datasets, consisting of 20 common mRNA members and two lncRNAs, namely CHRM3-AS2 and PRKCQ-AS1, which are potential regulators of sepsis. Our results reveal that the up-regulated common mRNAs are mainly involved in the processes of neutrophil mediated immunity, while the down-regulated mRNAs and lncRNAs are significantly overrepresented in T-cell mediated immunity functions. This study sheds light on the co-expression patterns of mRNAs and lncRNAs in sepsis, providing a novel perspective and insight into the sepsis transcriptome, which may facilitate the exploration of candidate therapeutic targets and molecular biomarkers for sepsis.

A Robust and Generalizable Immune-Related Signature for Sepsis Diagnostics

High-throughput sequencing can detect tens of thousands of genes in parallel, providing opportunities for improving the diagnostic accuracy of multiple diseases including sepsis, which is an aggressive inflammatory response to infection that can cause organ failure and death. Early screening of sepsis is essential in clinic, but no effective diagnostic biomarkers are available yet. Here, we present a novel method, Recurrent Logistic Regression, to identify diagnostic biomarkers for sepsis from the blood transcriptome data. A panel including five immune-related genes, LRRN3, IL2RB, FCER1A, TLR5, and S100A12, are determined as diagnostic biomarkers (LIFTS) for sepsis. LIFTS discriminates patients with sepsis from normal controls in high accuracy (AUROC = 0.9959 on average; IC = [0.9722-1.0]) on nine validation cohorts across three independent platforms, which outperforms existing markers. Our analysis determined an accurate prediction model and reproducible transcriptome biomarkers that can lay a foundation for clinical diagnostic tests and biological mechanistic studies.

Identification of key immune genes for sepsis-induced ARDS based on bioinformatics analysis

Regarding the extremely high mortality caused by sepsis-induced acute respiratory distress syndrome (ARDS), it is urgent to develop new biomarkers of sepsis-induced ARDS for treatment. Here, 532 differential expression genes (DEGs) related to sepsis and 433 DEGs related to sepsis-induced ARDS were screened in the GSE32707 dataset. Compared with sepsis samples, sepsis ARDS samples showed a higher infiltration of activated memory CD4 T cells and naive B cells, but a relatively lower infiltration of CD8 T cells. The pink and green modules which are significantly associated with sepsis-induced ARDS were then screened through co-expression network analysis. Differentially up-regulated GYPE and aberrantly down-regulated HSPB1, were subsequently found in the pink module of ARDS. CD81 and RPL22, two differentially low-expressed genes peculiar to ARDS, were identified in the green module. The function of CD81 was verified at the cellular level, and it was found that the up-regulation of CD81 in A549 could alleviate the LPS-induced injury of A549 cells. More importantly, the overexpressed CD81 can also increase the content of CD4⁺ CD25⁺ Foxp3⁺ Treg in Jurkat cells, and after the co-culture of overexpressed CD81 Jurkat cells with LPS treatment A549 cells, the LPS-induced lung epithelial cell damage can be improved. Overall, four new plasma biomarker candidates were found in sepsis-induced ARDS, and we verified that CD81 may play critical roles in the biological and immunological processes of sepsis-induced ARDS.

Blood Circulating miRNA Pairs as a Robust Signature for Early Detection of Esophageal Cancer

Esophageal cancer (EC) is a common malignant tumor in the digestive system which is often diagnosed at the middle and late stages. Noninvasive diagnosis using circulating miRNA as biomarkers enables accurate detection of early-stage EC to reduce mortality. We built a diagnostic signature consisting of four miRNA pairs for the early detection of EC using individualized Pairwise Analysis of Gene Expression (iPAGE). Profiling of miRNA expression identified 496 miRNA pairs with significant relative expression change. Four miRNA pairs consistently selected from LASSO were used to construct the final diagnostic model. The performance of the signature was validated using two independent datasets, yielding both AUCs and PRCs over 0.99. Furthermore, precision, recall, and F-score were also evaluated for clinical application, when a fixed threshold is given, resulting in all the scores are larger than 0.92 in the training set, test set, and two validation sets. Our results suggested that the 4-miRNA signature is a new biomarker for the early diagnosis of patients with EC. The clinical use of this signature would have improved the detection of EC for earlier therapy and more favorite prognosis.

Spatiotemporal 22q11.21 Protein Network Implicates DGCR8-Dependent MicroRNA Biogenesis as a Risk for Late-Fetal Cortical Development in Psychiatric Diseases

The chromosome 22q11.21 copy number variant (CNV) is a vital risk factor that can be a genetic predisposition to neurodevelopmental disorders (NDD). As the 22q11.21 CNV affects multiple genes, causal disease genes and mechanisms affected are still poorly understood. Thus, we aimed to identify the most impactful 22q11.21 CNV genes and the potential impacted human brain regions, developmental stages and signaling pathways. We constructed the spatiotemporal dynamic networks of 22q11.21 CNV genes using the brain developmental transcriptome and physical protein–protein interactions. The affected brain regions, developmental stages, driver genes and pathways were subsequently investigated via integrated bioinformatics analysis. As a result, we first identified that 22q11.21 CNV genes affect the cortical area mainly during late fetal periods. Interestingly, we observed that connections between a driver gene, DGCR8, and its interacting partners, MECP2 and CUL3, also network hubs, only existed in the network of the late fetal period within the cortical region, suggesting their functional specificity during brain development. We also confirmed the physical interaction result between DGCR8 and CUL3 by liquid chromatography-tandem mass spectrometry. In conclusion, our results could suggest that the disruption of DGCR8-dependent microRNA biogenesis plays a vital role in NDD for late fetal cortical development.

Evaluating the Consistency of Gene Methylation in Liver Cancer Using Bisulfite Sequencing Data

Bisulfite sequencing is considered as the gold standard approach for measuring DNA methylation, which acts as a pivotal part in regulating a variety of biological processes without changes in DNA sequences. In this study, we introduced the most prevalent methods for processing bisulfite sequencing data and evaluated the consistency of the data acquired from different measurements in liver cancer. Firstly, we introduced three commonly used bisulfite sequencing assays, i.e., reduced-representation bisulfite sequencing (RRBS), whole-genome bisulfite sequencing (WGBS), and targeted bisulfite sequencing (targeted BS). Next, we discussed the principles and compared different methods for alignment, quality assessment, methylation level scoring, and differentially methylated region identification. After that, we screened differential methylated genes in liver cancer through the three bisulfite sequencing assays and evaluated the consistency of their results. Ultimately, we compared bisulfite sequencing to 450 k beadchip and assessed the statistical similarity and functional association of differentially methylated genes (DMGs) among the four assays. Our results demonstrated that the DMGs measured by WGBS, RRBS, targeted BS and 450 k beadchip are consistently hypo-methylated in liver cancer with high functional similarity.

Long non-coding RNA pairs to assist in diagnosing sepsis

Background

Sepsis is the major cause of death in Intensive Care Unit (ICU) globally. Molecular detection enables rapid diagnosis that allows early intervention to minimize the death rate. Recent studies showed that long non-coding RNAs (lncRNAs) regulate proinflammatory genes and are related to the dysfunction of organs in sepsis. Identifying lncRNA signature with absolute abundance is challenging because of the technical variation and the systematic experimental bias.

Results

Cohorts (n = 768) containing whole blood lncRNA profiling of sepsis patients in the Gene Expression Omnibus (GEO) database were included. We proposed a novel diagnostic strategy that made use of the relative expressions of lncRNA pairs, which are reversed between sepsis patients and normal controls (eg. lncRNA_i > lncRNA_j in sepsis patients and lncRNA_i < lncRNA_j in normal controls), to identify 14 lncRNA pairs as a sepsis diagnostic signature. The signature was then applied to independent cohorts (n = 644) to evaluate its predictive performance across different ages and normalization methods. Comparing to common machine learning models and existing signatures, SepSigLnc consistently attains better performance on the validation cohorts from the same age group (AUC = 0.990 & 0.995 in two cohorts) and across different groups (AUC = 0.878 on average), as well as cohorts processed by an alternative normalization method (AUC = 0.953 on average). Functional analysis demonstrates that the lncRNA pairs in SepsigLnc are functionally similar and tend to implicate in the same biological processes including cell fate commitment and cellular response to steroid hormone stimulus.

Conclusion

Our study identified 14 lncRNA pairs as signature that can facilitate the diagnosis of septic patients at an intervenable point when clinical manifestations are not dramatic. Also, the computational procedure can be generalized to a standard procedure for discovering diagnostic molecule signatures.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-021-07576-4.

Systematic prediction of autophagy-related proteins using Arabidopsis thaliana interactome data

Autophagy is a self-degradative process that is crucial for maintaining cellular homeostasis by removing damaged cytoplasmic components and recycling nutrients. Such an evolutionary conserved proteolysis process is regulated by the autophagy-related (Atg) proteins. The incomplete understanding of plant autophagy proteome and the importance of a proteome-wide understanding of the autophagy pathway prompted us to predict Atg proteins and regulators in Arabidopsis. Here, we developed a systems-level algorithm to identify autophagy-related modules (ARMs) based on protein subcellular localization, protein-protein interactions, and known Atg proteins. This generates a detailed landscape of the autophagic modules in Arabidopsis. We found that the newly identified genes in each ARM tend to be upregulated and coexpressed during the senescence stage of Arabidopsis. We also demonstrated that the Golgi apparatus ARM, ARM13, functions in the autophagy process by module clustering and functional analysis. To verify the in silico analysis, the Atg candidates in ARM13 that are functionally similar to the core Atg proteins were selected for experimental validation. Interestingly, two of the previously uncharacterized proteins identified from the ARM analysis, AGD1 and Sec14, exhibited bona fide association with the autophagy protein complex in plant cells, which provides evidence for a cross-talk between intracellular pathways and autophagy. Thus, the computational framework has facilitated the identification and characterization of plant-specific autophagy-related proteins and novel autophagy proteins/regulators in higher eukaryotes.

MirLocPredictor: A ConvNet-Based Multi-Label MicroRNA Subcellular Localization Predictor by Incorporating k-Mer Positional Information

MicroRNAs (miRNA) are small noncoding RNA sequences consisting of about 22 nucleotides that are involved in the regulation of almost 60% of mammalian genes. Presently, there are very limited approaches for the visualization of miRNA locations present inside cells to support the elucidation of pathways and mechanisms behind miRNA function, transport, and biogenesis. MIRLocator, a state-of-the-art tool for the prediction of subcellular localization of miRNAs makes use of a sequence-to-sequence model along with pretrained k-mer embeddings. Existing pretrained k-mer embedding generation methodologies focus on the extraction of semantics of k-mers. However, in RNA sequences, positional information of nucleotides is more important because distinct positions of the four nucleotides define the function of an RNA molecule. Considering the importance of the nucleotide position, we propose a novel approach (kmerPR2vec) which is a fusion of positional information of k-mers with randomly initialized neural k-mer embeddings. In contrast to existing k-mer-based representation, the proposed kmerPR2vec representation is much more rich in terms of semantic information and has more discriminative power. Using novel kmerPR2vec representation, we further present an end-to-end system (MirLocPredictor) which couples the discriminative power of kmerPR2vec with Convolutional Neural Networks (CNNs) for miRNA subcellular location prediction. The effectiveness of the proposed kmerPR2vec approach is evaluated with deep learning-based topologies (i.e., Convolutional Neural Networks (CNN) and Recurrent Neural Network (RNN)) and by using 9 different evaluation measures. Analysis of the results reveals that MirLocPredictor outperform state-of-the-art methods with a significant margin of 18% and 19% in terms of precision and recall.

Protective Effects of Adiponectin against Cobalt Chloride-Induced Apoptosis of Smooth Muscle Cells via cAMP/PKA Pathway

Adiponectin (APN) is an adipokine secreted from adipose tissue and exhibits biological functions such as microcirculation-regulating, hearing-protective, and antiapoptotic. However, the effect of APN on the apoptosis of spiral arterial smooth muscle cells (SMCs) under hypoxic conditions in vitro is not clear. We used cobalt chloride (CoCl₂) to simulate chemical hypoxia in vitro, and the SMCs were pretreated with APN and then stimulated with CoCl₂. The viability of cells and apoptosis were assessed by CCK-8 and flow cytometry, respectively. Superoxide dismutase (SOD) activity, malondialdehyde (MDA) levels, cAMP level, and the activity of PKA were detected by ELISA. Protein expression and localization were studied by Western blot and immunofluorescence analysis. In the present study, we found that APN exhibits antiapoptosis effects. CoCl₂ exhibited decreased cell viability, increased apoptosis and MDA levels, and decreased SOD activity in a concentration-dependent manner, compared with the control group. Moreover, CoCl₂ upregulated the expression levels of Bax and cleaved caspase-3 and then downregulated Bcl-2 levels in a time-dependent manner. Compared with the CoCl₂ group, the group pretreated with APN had increased cell viability, SOD activity, PKA activity, cAMP level, and PKA expression, but decreased MDA levels and apoptosis. Lastly, the protective effect of APN was blocked by cAMP inhibitor SQ22536 and PKA inhibitor H 89. These results showed that APN protected SMCs against CoCl₂-induced hypoxic injury via the cAMP/PKA signaling pathway.

Knockdown of lncRNA MALAT1 Alleviates LPS-Induced Acute Lung Injury via Inhibiting Apoptosis Through the miR-194-5p/FOXP2 Axis

Purpose

We aimed to identify and verify the key genes and lncRNAs associated with acute lung injury (ALI) and explore the pathogenesis of ALI. Research showed that lower expression of the lncRNA metastasis-associated lung carcinoma transcript 1 (MALAT1) alleviates lung injury induced by lipopolysaccharide (LPS). Nevertheless, the mechanisms of MALAT1 on cellular apoptosis remain unclear in LPS-stimulated ALI. We investigated the mechanism of MALAT1 in modulating the apoptosis of LPS-induced human pulmonary alveolar epithelial cells (HPAEpiC).

Methods

Differentially expressed lncRNAs between the ALI samples and normal controls were identified using gene expression profiles. ALI-related genes were determined by the overlap of differentially expressed genes (DEGs), genes correlated with lung, genes correlated with key lncRNAs, and genes sharing significantly high proportions of microRNA targets with MALAT1. Quantitative real-time PCR (qPCR) was applied to detect the expression of MALAT1, microRNA (miR)-194-5p, and forkhead box P2 (FOXP2) mRNA in 1 μg/ml LPS-treated HPAEpiC. MALAT1 knockdown vectors, miR-194-5p inhibitors, and ov-FOXP2 were constructed and used to transfect HPAEpiC. The influence of MALAT1 knockdown on LPS-induced HPAEpiC proliferation and apoptosis via the miR-194-5p/FOXP2 axis was determined using Cell counting kit-8 (CCK-8) assay, flow cytometry, and Western blotting analysis, respectively. The interactions between MALAT1, miR-194-5p, and FOXP2 were verified using dual-luciferase reporter gene assay.

Results

We identified a key lncRNA (MALAT1) and three key genes (EYA1, WNT5A, and FOXP2) that are closely correlated with the pathogenesis of ALI. LPS stimulation promoted MALAT1 expression and apoptosis and also inhibited HPAEpiC viability. MALAT1 knockdown significantly improved viability and suppressed the apoptosis of LPS-stimulated HPAEpiC. Moreover, MALAT1 directly targeted miR-194-5p, a downregulated miRNA in LPS-stimulated HPAEpiC, when FOXP2 was overexpressed. MALAT1 knockdown led to the overexpression of miR-194-5p and restrained FOXP2 expression. Furthermore, inhibition of miR-194-5p exerted a rescue effect on MALAT1 knockdown of FOXP2, whereas the overexpression of FOXP2 reversed the effect of MALAT1 knockdown on viability and apoptosis of LPS-stimulated HPAEpiC.

Conclusion

Our results demonstrated that MALAT1 knockdown alleviated HPAEpiC apoptosis by competitively binding to miR-194-5p and then elevating the inhibitory effect on its target FOXP2. These data provide a novel insight into the role of MALAT1 in the progression of ALI and potential diagnostic and therapeutic strategies for ALI patients.

Weighted correlation network bioinformatics uncovers a key molecular biosignature driving the left-sided heart failure

BACKGROUND

Left-sided heart failure (HF) is documented as a key prognostic factor in HF. However, the relative molecular mechanisms underlying left-sided HF is unknown. The purpose of this study is to unearth significant modules, pivotal genes and candidate regulatory components governing the progression of left-sided HF by bioinformatical analysis.

METHODS

A total of 319 samples in GSE57345 dataset were used for weighted gene correlation network analysis (WGCNA). ClusterProfiler package in R was used to conduct functional enrichment for genes uncovered from the modules of interest. Regulatory networks of genes were built using Cytoscape while Enrichr database was used for identification of transcription factors (TFs). The MCODE plugin was used for identifying hub genes in the modules of interest and their validation was performed based on GSE1869 dataset.

RESULTS

A total of six significant modules were identified. Notably, the blue module was confirmed as the most crucially associated with left-sided HF, ischemic heart disease (ISCH) and dilated cardiomyopathy (CMP). Functional enrichment conveyed that genes belonging to this module were mainly those driving the extracellular matrix-associated processes such as extracellular matrix structural constituent and collagen binding. A total of seven transcriptional factors, including Suppressor of Zeste 12 Protein Homolog (SUZ12) and nuclear factor erythroid 2 like 2 (NFE2L2), adrenergic receptor (AR), were identified as possible regulators of coexpression genes identified in the blue module. A total of three key genes (OGN, HTRA1 and MXRA5) were retained after validation of their prognostic value in left-sided HF. The results of functional enrichment confirmed that these key genes were primarily involved in response to transforming growth factor beta and extracellular matrix.

CONCLUSION

We uncovered a candidate gene signature correlated with HF, ISCH and CMP in the left ventricle, which may help provide better prognosis and therapeutic decisions and in HF, ISCH and CMP patients.

Whole blood transcriptomic investigation identifies long non-coding RNAs as regulators in sepsis

Background

Sepsis is a fatal disease referring to the presence of a known or strongly suspected infection coupled with systemic and uncontrolled immune activation causing multiple organ failure. However, current knowledge of the role of lncRNAs in sepsis is still extremely limited.

Methods

We performed an in silico investigation of the gene coexpression pattern for the patients response to all-cause sepsis in consecutive intensive care unit (ICU) admissions. Sepsis coexpression gene modules were identified using WGCNA and enrichment analysis. lncRNAs were determined as sepsis biomarkers based on the interactions among lncRNAs and the identified modules.

Results

Twenty-three sepsis modules, including both differentially expressed modules and prognostic modules, were identified from the whole blood RNA expression profiling of sepsis patients. Five lncRNAs, FENDRR, MALAT1, TUG1, CRNDE, and ANCR, were detected as sepsis regulators based on the interactions among lncRNAs and the identified coexpression modules. Furthermore, we found that CRNDE and MALAT1 may act as miRNA sponges of sepsis related miRNAs to regulate the expression of sepsis modules. Ultimately, FENDRR, MALAT1, TUG1, and CRNDE were reannotated using three independent lncRNA expression datasets and validated as differentially expressed lncRNAs.

Conclusion

The procedure facilitates the identification of prognostic biomarkers and novel therapeutic strategies of sepsis. Our findings highlight the importance of transcriptome modularity and regulatory lncRNAs in the progress of sepsis.

Phase-Separated Multienzyme Biosynthesis

Liquid-liquid phase separation forms condensates that feature a highly concentrated liquid phase, a defined yet dynamic boundary, and dynamic exchange at and across the boundary. Phase transition drives the formation of dynamic multienzyme complexes in cells, for example, the purinosome, which forms subcellular macrobodies responsible for de novo purine biosynthesis. Here, we construct synthetic versions of multienzyme biosynthetic systems by assembling enzymes in protein condensates. A synthetic protein phase separation system using component proteins from postsynaptic density in neuronal synapses, GKAP, Shank, and Homer provides the scaffold for assembly. Three sets of guest proteins: a pair of fluorescent proteins (CFP and YFP), three sequential enzymes in menaquinone biosynthesis pathway (MenF, MenD, and MenH), and two enzymes in terpene biosynthesis pathway (Idi and IspA) are assembled via peptide-peptide interactions in the condensate. First, we discover that coassembly of CFP and YFP exhibited a broad distribution of the FRET signal within the condensate. Second, a spontaneous enrichment of the rate-limiting enzyme MenD in the condensate is sufficient to increase the 2-succinyl-6-hydroxy-2,4-cyclohexadiene-1-carboxylate production rate by 70%. Third, coassembly of both Idi and IspA in the protein condensate increases the farnesyl pyrophosphate production rate by more than 50%. Altogether, we show here that phase separation significantly accelerates the efficiency of multienzyme biocatalysis.

A network-based algorithm for the identification of moonlighting noncoding RNAs and its application in sepsis

Moonlighting proteins provide more options for cells to execute multiple functions without increasing the genome and transcriptome complexity. Although there have long been calls for computational methods for the prediction of moonlighting proteins, no method has been designed for determining moonlighting long noncoding ribonucleicacidz (RNAs) (mlncRNAs). Previously, we developed an algorithm MoonFinder for the identification of mlncRNAs at the genome level based on the functional annotation and interactome data of lncRNAs and proteins. Here, we update MoonFinder to MoonFinder v2.0 by providing an extensive framework for the detection of protein modules and the establishment of RNA-module associations in human. A novel measure, moonlighting coefficient, was also proposed to assess the confidence of an ncRNA acting in a moonlighting manner. Moreover, we explored the expression characteristics of mlncRNAs in sepsis, in which we found that mlncRNAs tend to be upregulated and differentially expressed. Interestingly, the mlncRNAs are mutually exclusive in terms of coexpression when compared to the other lncRNAs. Overall, MoonFinder v2.0 is dedicated to the prediction of human mlncRNAs and thus bears great promise to serve as a valuable R package for worldwide research communities (https://cran.r-project.org/web/packages/MoonFinder/index.html). Also, our analyses provide the first attempt to characterize mlncRNA expression and coexpression properties in adult sepsis patients, which will facilitate the understanding of the interaction and expression patterns of mlncRNAs.

Normalization Methods for the Analysis of Unbalanced Transcriptome Data: A Review

Dozens of normalization methods for correcting experimental variation and bias in high-throughput expression data have been developed during the last two decades. Up to 23 methods among them consider the skewness of expression data between sample states, which are even more than the conventional methods, such as loess and quantile. From the perspective of reference selection, we classified the normalization methods for skewed expression data into three categories, data-driven reference, foreign reference, and entire gene set. We separately introduced and summarized these normalization methods designed for gene expression data with global shift between compared conditions, including both microarray and RNA-seq, based on the reference selection strategies. To our best knowledge, this is the most comprehensive review of available preprocessing algorithms for the unbalanced transcriptome data. The anatomy and summarization of these methods shed light on the understanding and appropriate application of preprocessing methods.

Exploiting locational and topological overlap model to identify modules in protein interaction networks

Background

Clustering molecular network is a typical method in system biology, which is effective in predicting protein complexes or functional modules. However, few studies have realized that biological molecules are spatial-temporally regulated to form a dynamic cellular network and only a subset of interactions take place at the same location in cells.

Results

In this study, considering the subcellular localization of proteins, we first construct a co-localization human protein interaction network (PIN) and systematically investigate the relationship between subcellular localization and biological functions. After that, we propose a Locational and Topological Overlap Model (LTOM) to preprocess the co-localization PIN to identify functional modules. LTOM requires the topological overlaps, the common partners shared by two proteins, to be annotated in the same localization as the two proteins. We observed the model has better correspondence with the reference protein complexes and shows more relevance to cancers based on both human and yeast datasets and two clustering algorithms, ClusterONE and MCL.

Conclusion

Taking into consideration of protein localization and topological overlap can improve the performance of module detection from protein interaction networks.

Electronic supplementary material

The online version of this article (10.1186/s12859-019-2598-7) contains supplementary material, which is available to authorized users.

SMILE: a novel procedure for subcellular module identification with localisation expansion

Computational clustering methods help identify functional modules in protein–protein interaction (PPI) network, in which proteins participate in the same biological pathways or specific functions. Subcellular localisation is crucial for proteins to implement biological functions and each compartment accommodates specific portions of the protein interaction structure. However, the importance of protein subcellular localisation is often neglected in the studies of module identification. In this study, the authors propose a novel procedure, subcellular module identification with localisation expansion (SMILE), to identify super modules that consist of several subcellular modules performing specific biological functions among cell compartments. These super modules identified by SMILE are more functionally diverse and have been verified to be more associated with known protein complexes and biological pathways compared with the modules identified from the global PPI networks in both the compartmentalised PPI and InWeb_InBioMap datasets. The authors’ results reveal that subcellular localisation is a principal feature of functional modules and offers important guidance in detecting biologically meaningful results.