Predicting glioblastoma prognosis networks using weighted gene co-expression network analysis on TCGA data

Identification of the hub genes related to adipose tissue metabolism of bovine

Due to the demand for high-quality animal protein, there has been consistent interest in how to obtain more high-quality beef. As well-known, the adipose content of beef has a close connection with the taste and quality of beef, and cattle with different energy or protein diet have corresponding effects on the lipid metabolism of beef. Thus, we performed weighted gene co-expression network analysis (WGCNA) with subcutaneous adipose genes from Norwegian red heifers fed different diets to identify hub genes regulating bovine lipid metabolism. For this purpose, the RNA sequencing data of subcutaneous adipose tissue of 12-month-old Norwegian red heifers (n = 48) with different energy or protein levels were selected from the GEO database, and 7,630 genes with the largest variation were selected for WGCNA analysis. Then, three modules were selected as hub genes candidate modules according to the correlation between modules and phenotypes, including pink, magenta and grey60 modules. GO and KEGG enrichment analysis showed that genes were related to metabolism, and participated in Rap, MAPK, AMPK, VEGF signaling pathways, and so forth. Combined gene interaction network analysis using Cytoscape software, eight hub genes of lipid metabolism were identified, including TIA1, LOC516108, SNAPC4, CPSF2, ZNF574, CLASRP, MED15 and U2AF2. Further, the expression levels of hub genes in the cattle tissue were also measured to verify the results, and we found hub genes in higher expression in muscle and adipose tissue in adult cattle. In summary, we predicted the key genes of lipid metabolism in the subcutaneous adipose tissue that were affected by the intake of various energy diets to find the hub genes that coordinate lipid metabolism, which provide a theoretical basis for regulating beef quality.

TPSC: a module detection method based on topology potential and spectral clustering in weighted networks and its application in gene co-expression module discovery

BACKGROUND

Gene co-expression networks are widely studied in the biomedical field, with algorithms such as WGCNA and lmQCM having been developed to detect co-expressed modules. However, these algorithms have limitations such as insufficient granularity and unbalanced module size, which prevent full acquisition of knowledge from data mining. In addition, it is difficult to incorporate prior knowledge in current co-expression module detection algorithms.

RESULTS

In this paper, we propose a novel module detection algorithm based on topology potential and spectral clustering algorithm to detect co-expressed modules in gene co-expression networks. By testing on TCGA data, our novel method can provide more complete coverage of genes, more balanced module size and finer granularity than current methods in detecting modules with significant overall survival difference. In addition, the proposed algorithm can identify modules by incorporating prior knowledge.

CONCLUSION

In summary, we developed a method to obtain as much as possible information from networks with increased input coverage and the ability to detect more size-balanced and granular modules. In addition, our method can integrate data from different sources. Our proposed method performs better than current methods with complete coverage of input genes and finer granularity. Moreover, this method is designed not only for gene co-expression networks but can also be applied to any general fully connected weighted network.

Comparative Genomic Analysis of the DUF34 Protein Family Suggests Role as a Metal Ion Chaperone or Insertase

Members of the DUF34 (domain of unknown function 34) family, also known as the NIF3 protein superfamily, are ubiquitous across superkingdoms. Proteins of this family have been widely annotated as "GTP cyclohydrolase I type 2" through electronic propagation based on one study. Here, the annotation status of this protein family was examined through a comprehensive literature review and integrative bioinformatic analyses that revealed varied pleiotropic associations and phenotypes. This analysis combined with functional complementation studies strongly challenges the current annotation and suggests that DUF34 family members may serve as metal ion insertases, chaperones, or metallocofactor maturases. This general molecular function could explain how DUF34 subgroups participate in highly diversified pathways such as cell differentiation, metal ion homeostasis, pathogen virulence, redox, and universal stress responses.

Dissecting Prognosis Modules and Biomarkers in Glioblastoma Based on Weighted Gene Co-Expression Network Analysis

Introduction

As one of the most prevalent and malignant brain cancers, glioblastoma multiforme (GBM) presents a poor prognosis and the molecular mechanisms remain poorly understood. Consequently, molecular research, including various biomarkers, is essential to exploit the occurrence and development of glioma.

Methods

Weighted gene co-expression network analysis (WGCNA) was used to construct gene co-expression modules and networks based on the Chinese Glioma Genome Atlas (CGGA) glioblastoma specimens. Then, protein–protein interaction (PPI) and gene ontology (GO) analyses were performed to mine hub genes. RT-PCR and immunohistochemistry were employed to examine the expression level of GRPR, CXCL5, and CXCL11 in glioma patients.

Results

We confirmed two gene modules by protein–protein interaction networks. Functional enrichment analysis was performed to identify the significance of gene modules. Prognostic biomarkers GRPR, CXCL5, and CXCL11 related to the survival time of GBM samples were mined in The Cancer Genome Atlas (TCGA) dataset. qRT-PCR revealed that GRPR, CXCL5, and CXCL11 led to a significant increase in GBM sample compared to control.

Conclusion

In this study, we developed and confirmed three mRNA signatures (GRPR, CXCL5, and CXCL11) for evaluating overall survival in GBM patients. Our research assists in existing understanding of GBM diagnosis and prognosis.

MicroRNA-744-5p inhibits glioblastoma malignancy by suppressing replication factor C subunit 2

Glioblastoma (GBM) is the most common malignant primary brain tumor, accounting for ~57% of all gliomas and 48% of all malignant primary central nervous system tumors in the United States. Abnormal expression of the replication factor C subunit 2 (RFC2) gene and microRNA (miR)-744-5p is associated with tumorigenic characteristics, including cellular proliferation, migration and invasiveness. However, the mechanism underlying the interaction between miR-744-5p and RFC2 in GBM remains unknown. Reverse transcription-quantitative (RT-q) PCR analysis of RFC2 and miR-744-5p was performed using GBM tumor tissues and cells, and the association between miR-744-5p and RFC2 was determined by dual-luciferase reporter assay. Cell Counting Kit 8, 5-bromo-2-deoxyuridine (BrdU), wound-healing and cellular adhesion assays, as well as the detection of caspase-3 activity and western blotting were used to detect cellular proliferation, migration and adhesion, caspase-3 activity, and Bax and Bcl-2 protein expression, respectively, in GBM cells. The results of the present study demonstrated that RFC2 expression was increased in GBM tissues and cell lines. Overexpression of RFC2 promoted cellular proliferation, migration, adhesion and an increase in Bcl-2 protein levels, and suppressed cellular caspase-3 activity and Bax protein expression, while silencing RFC2 resulted in the opposite effect. The effects of miR-744-5p inhibition were similar to those of RFC2 overexpression. Moreover, miR-744-5p was found to target RFC2 in GBM cells, and inhibiting the expression of RFC2 suppressed GBM tumorigenesis. In conclusion, the present study demonstrated that miR-744-5p targets RFC2 and suppresses the progression of GBM by repressing cellular proliferation, migration and Bcl-2 protein expression, and effectively promoting caspase-3 activity and Bax protein expression. These findings suggest a new target for the clinical treatment and improved prognosis of patients with GBM in the future.

PLEK2, RRM2, GCSH: A Novel WWOX-Dependent Biomarker Triad of Glioblastoma at the Crossroads of Cytoskeleton Reorganization and Metabolism Alterations

Glioblastoma is one of the deadliest human cancers. Its malignancy depends on cytoskeleton reorganization, which is related to, e.g., epithelial-to-mesenchymal transition and metastasis. The malignant phenotype of glioblastoma is also affected by the WWOX gene, which is lost in nearly a quarter of gliomas. Although the role of WWOX in the cytoskeleton rearrangement has been found in neural progenitor cells, its function as a modulator of cytoskeleton in gliomas was not investigated. Therefore, this study aimed to investigate the role of WWOX and its collaborators in cytoskeleton dynamics of glioblastoma. Methodology on RNA-seq data integrated the use of databases, bioinformatics tools, web-based platforms, and machine learning algorithm, and the obtained results were validated through microarray data. PLEK2, RRM2, and GCSH were the most relevant WWOX-dependent genes that could serve as novel biomarkers. Other genes important in the context of cytoskeleton (BMP4, CCL11, CUX2, DUSP7, FAM92B, GRIN2B, HOXA1, HOXA10, KIF20A, NF2, SPOCK1, TTR, UHRF1, and WT1), metabolism (MTHFD2), or correlation with WWOX (COL3A1, KIF20A, RNF141, and RXRG) were also discovered. For the first time, we propose that changes in WWOX expression dictate a myriad of alterations that affect both glioblastoma cytoskeleton and metabolism, rendering new therapeutic possibilities.

Identification of breast cancer prognostic modules via differential module selection based on weighted gene Co-expression network analysis

Breast cancer is a complex cancer which includes many different subtypes. Identifying prognostic modules, i.e., functionally related gene networks that play crucial roles in cancer development is essential in breast cancer study. Different subtypes of breast cancer correspond to different treatment methods. The purpose of this study is to use a new method to divide breast cancer into different prognostic modules, so as to provide scientific basis for improving clinical management. The method is based on comparing similarities between modules detected from different weighted gene co-expression networks. The method was applied on genomic data of breast cancer from The Cancer Genome Atlas database and was applied to select differential modules between two groups of patients with significant differences in survival times. It was compared with a previously proposed module selection method. The result shows that our method outperforms the previously proposed one. Moreover, within the identified two differential modules, the first one is highly enriched with genes involved in hormone responds, the second one is highly related with biological process engaged in M-phase. The two modules were further validated by log-rank test in the validation dataset GSE3494. Both of the two modules show significantly different with p-values less than 0.02. The identified two modules confirmed previous findings including importance of biological networks in breast cancer involved in hormone response and M-phase. Out of the top twenty hub genes in the two modules, fifteen genes were previously shown to be prognostic markers for breast cancer.

A bibliometric analysis and visualization of medical data mining research

BACKGROUND

Data mining technology used in the field of medicine has been widely studied by scholars all over the world. But there is little research on medical data mining (MDM) from the perspectives of bibliometrics and visualization, and the research topics and development trends in this field are still unclear.

METHODS

This paper has applied bibliometric visualization software tools, VOSviewer 1.6.10 and CiteSpace V, to study the citation characteristics, international cooperation, author cooperation, and geographical distribution of the MDM.

RESULTS

A total of 1575 documents are obtained, and the most frequent document type is article (1376). SHAN NH is the most productive author, with the highest number of publications of 12, and the Gillies's article (750 times citation) is the most cited paper. The most productive country and institution in MDM is the USA (559) and US FDA (35), respectively. The Journal of Biomedical Informatics, Expert Systems with Applications and Journal of Medical Systems are the most productive journals, which reflected the nature of the research, and keywords "classification (790)" and "system (576)" have the strongest strength. The hot topics in MDM are drug discovery, medical imaging, vaccine safety, and so on. The 3 frontier topics are reporting system, precision medicine, and inflammation, and would be the foci of future research.

CONCLUSION

The present study provides a panoramic view of data mining methods applied in medicine by visualization and bibliometrics. Analysis of authors, journals, institutions, and countries could provide reference for researchers who are fresh to the field in different ways. Researchers may also consider the emerging trends when deciding the direction of their study.

Generalized gene co-expression analysis via subspace clustering using low-rank representation

Background

Gene Co-expression Network Analysis (GCNA) helps identify gene modules with potential biological functions and has become a popular method in bioinformatics and biomedical research. However, most current GCNA algorithms use correlation to build gene co-expression networks and identify modules with highly correlated genes. There is a need to look beyond correlation and identify gene modules using other similarity measures for finding novel biologically meaningful modules.

Results

We propose a new generalized gene co-expression analysis algorithm via subspace clustering that can identify biologically meaningful gene co-expression modules with genes that are not all highly correlated. We use low-rank representation to construct gene co-expression networks and local maximal quasi-clique merger to identify gene co-expression modules. We applied our method on three large microarray datasets and a single-cell RNA sequencing dataset. We demonstrate that our method can identify gene modules with different biological functions than current GCNA methods and find gene modules with prognostic values.

Conclusions

The presented method takes advantage of subspace clustering to generate gene co-expression networks rather than using correlation as the similarity measure between genes. Our generalized GCNA method can provide new insights from gene expression datasets and serve as a complement to current GCNA algorithms.

SALMON: Survival Analysis Learning With Multi-Omics Neural Networks on Breast Cancer

Improved cancer prognosis is a central goal for precision health medicine. Though many models can predict differential survival from data, there is a strong need for sophisticated algorithms that can aggregate and filter relevant predictors from increasingly complex data inputs. In turn, these models should provide deeper insight into which types of data are most relevant to improve prognosis. Deep Learning-based neural networks offer a potential solution for both problems because they are highly flexible and account for data complexity in a non-linear fashion. In this study, we implement Deep Learning-based networks to determine how gene expression data predicts Cox regression survival in breast cancer. We accomplish this through an algorithm called SALMON (Survival Analysis Learning with Multi-Omics Neural Networks), which aggregates and simplifies gene expression data and cancer biomarkers to enable prognosis prediction. The results revealed improved performance when more omics data were used in model construction. Rather than use raw gene expression values as model inputs, we innovatively use eigengene modules from the result of gene co-expression network analysis. The corresponding high impact co-expression modules and other omics data are identified by feature selection technique, then examined by conducting enrichment analysis and exploiting biological functions, escalated the interpretation of input feature from gene level to co-expression modules level. Our study shows the feasibility of discovering breast cancer related co-expression modules, sketch a blueprint of future endeavors on Deep Learning-based survival analysis. SALMON source code is available at https://github.com/huangzhii/SALMON/.

Identification of glioblastoma gene prognosis modules based on weighted gene co-expression network analysis

Background

Glioblastoma multiforme, the most prevalent and aggressive brain tumour, has a poor prognosis. The molecular mechanisms underlying gliomagenesis remain poorly understood. Therefore, molecular research, including various markers, is necessary to understand the occurrence and development of glioma.

Method

Weighted gene co-expression network analysis (WGCNA) was performed to construct a gene co-expression network in TCGA glioblastoma samples. Gene ontology (GO) and pathway-enrichment analysis were used to identify significance of gene modules. Cox proportional hazards regression model was used to predict outcome of glioblastoma patients.

Results

We performed weighted gene co-expression network analysis (WGCNA) and identified a gene module (yellow module) related to the survival time of TCGA glioblastoma samples. Then, 228 hub genes were calculated based on gene significance (GS) and module significance (MS). Four genes (OSMR + SOX21 + MED10 + PTPRN) were selected to construct a Cox proportional hazards regression model with high accuracy (AUC = 0.905). The prognostic value of the Cox proportional hazards regression model was also confirmed in GSE16011 dataset (GBM: n = 156).

Conclusion

We developed a promising mRNA signature for estimating overall survival in glioblastoma patients.

Electronic supplementary material

The online version of this article (10.1186/s12920-018-0407-1) contains supplementary material, which is available to authorized users.

Candidate Biomarkers and Molecular Mechanism Investigation for Glioblastoma Multiforme Utilizing WGCNA

To reveal the potential molecular mechanism of glioblastoma multiforme (GBM) and provide the candidate biomarkers for GBM gene therapy. Microarray dataset GSE50161 was obtained from GEO database. The differentially expressed genes (DEGs) were identified between GBM samples and control samples, followed by the module partition analysis based on WGCNA. Then, the pathway and functional enrichment analyses of DEGs were performed. The hub genes were further investigated, followed by the survival analysis and data validation. A total of 1913 DEGs were investigated between two groups, followed by analysis of 5 modules using WGCNA. These DEGs were mainly enriched in functions like inflammatory response. The hub genes including upregulated N-Myc and STAT Interactor (NMI), Capping Actin Protein-Gelsolin Like (CAPG), and Proteasome Subunit Beta 8 (PSMB8) were revealed as potential liquid biopsy molecules for GBM diagnose. Moreover, Nucleolar and Spindle Associated Protein 1 (NUSAP1) and G Protein-Coupled Receptor 65 (GPR65) were outstanding genes in survival analysis. Our results suggested that CPNE6, HAPLN2, CMTM3, NMI, CAPG, and PSMB8 might be used as potential molecules for liquid biopsy of GBM. NUSAP1 and GPR65 might be novel prognostic targets for GBM gene therapy. Furthermore, the upregulated NMI might play an important role in GBM progression via inflammatory response.

Application of weighted gene co-expression network analysis to identify key modules and hub genes in oral squamous cell carcinoma tumorigenesis

Purpose

Oral squamous cell carcinoma (OSCC) is one of the most common malignant diseases worldwide, yet its molecular mechanisms are largely unknown. We aimed to construct gene co-expression networks to identify key modules and hub genes involved in the pathogenesis of OSCC.

Patients and methods

We used dataset GSE30784 to construct co-expression networks by weighted gene co-expression network analysis (WGCNA). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed by Database for Annotation, Visualization and Integrated Discovery (DAVID). Hub genes were screened and validated by other datasets.

Results

Turquoise and brown modules were found to be the most significantly related to tumorigenesis. Functional enrichment analysis showed that the turquoise module was associated with cell–cell adhesion, extracellular matrix and collagen catabolic process. A total of 10 hub genes (MMP1, TNFRSF12A, PLAU, FSCN1, PDPN, KRT78, EVPL, GGT6, SMIM5 and CYSRT1) were identified and validated at transcriptional and translational levels. Their genetic alteration and survival analysis were also revealed.

Conclusion

We identified two modules and 10 hub genes, which were associated with the tumorigenesis of OSCC. The two modules provided references that will advance the understanding of mechanisms of tumorigenesis in OSCC. Moreover, the hub genes may serve as biomarkers and therapeutic targets for precise diagnosis and treatment of OSCC in the future.

Inducing hair follicle neogenesis with secreted proteins enriched in embryonic skin

Organ development is a sophisticated process of self-organization. However, despite growing understanding of the developmental mechanisms, little is known about how to reactivate them postnatally for regeneration. We found that treatment of adult non-hair fibroblasts with cell-free extract from embryonic skin conferred upon them the competency to regenerate hair follicles. Proteomics analysis identified three secreted proteins enriched in the embryonic skin, apolipoprotein-A1, galectin-1 and lumican that together were essential and sufficient to induce new hair follicles. These 3 proteins show a stage-specific co-enrichment in the perifolliculogenetic embryonic dermis. Mechanistically, exposure to embryonic skin extract or to the combination of the 3 proteins altered the gene expression to an inductive hair follicle dermal papilla fibroblast-like profile and activated Igf and Wnt signaling, which are crucial for the regeneration process. Therefore, a cocktail of organ-specific extracellular proteins from the embryonic environment can render adult cells competent to re-engage in developmental interactions for organ neogenesis. Identification of factors that recreate the extracellular context of respective developing tissues can become an important strategy to promote regeneration in adult organs.

Dramatic dysbalancing of the Wnt pathway in breast cancers

Wnt signaling is important for breast development and remodeling during pregnancy and lactation. Epigenetic modifications change expression levels of components of the Wnt pathway, underlying oncogenic transformation. However, no clear Wnt component increasing expression universally across breast cancer (BC) or its most Wnt-dependent triple-negative BC (TNBC) subgroup has been identified, delaying development of targeted therapies. Here we perform network correlation analysis of expression of >100 Wnt pathway components in hundreds of healthy and cancerous breast tissues. Varying in expression levels among people, Wnt components remarkably coordinate their production; this coordination is dramatically decreased in BC. Clusters with coordinated gene expression exist within the healthy cohort, highlighting Wnt signaling subtypes. Different BC subgroups are identified, characterized by different remaining Wnt signaling signatures, providing the rational for patient stratification for personalizing the therapeutic applications. Key pairwise interactions within the Wnt pathway (some inherited and some established de novo) emerge as targets for future drug discovery against BC.

Analysis of Human Nuclear Protein Complexes by Quantitative Mass Spectrometry Profiling

Analysis of protein complexes provides insights into how the ensemble of expressed proteome is organized into functional units. While there have been advances in techniques for proteome-wide profiling of cytoplasmic protein complexes, information about human nuclear protein complexes are very limited. To close this gap, we combined native size exclusion chromatography (SEC) with label-free quantitative MS profiling to characterize hundreds of nuclear protein complexes isolated from human glioblastoma multiforme T98G cells. We identified 1794 proteins that overlapped between two biological replicates of which 1244 proteins were characterized as existing within stably associated putative complexes. co-IP experiments confirmed the interaction of PARP1 with Ku70/Ku80 proteins and HDAC1 (histone deacetylase complex 1) and CHD4. HDAC1/2 also co-migrated with various SIN3A and nucleosome remodeling and deacetylase components in SEC fractionation including SIN3A, SAP30, RBBP4, RBBP7, and NCOR1. Co-elution of HDAC1/2/3 with both the KDM1A and RCOR1 further confirmed that these proteins are integral components of human deacetylase complexes. Our approach also demonstrated the ability to identify potential moonlighting complexes and novel complexes containing uncharacterized proteins. Overall, the results demonstrated the utility of SEC fractionation and LC-MS analysis for system-wide profiling of proteins to predict the existence of distinct forms of nuclear protein complexes.

DNA methylation biomarkers for the occurrence of lung adenocarcinoma from TCGA data mining

The development of lung cancer is a combination of multifactor, multistage, and multiple genetic alterations processes. DNA methylation is an important factor. Currently, the study on the genome-scale epigenetic modification for studying the pathogenesis of lung cancer is still lacking. Here, we aimed to identify the epigenetic modifications of lung cancer, thus to provide scientific basis for the personalized medicine, and research of classification screening for lung adenocarcinoma patients. The DNA methylation data, and the corresponding clinical information of lung adenocarcinoma samples were extracted from the Cancer Genome Atlas (TCGA) database. We explored the association of DNA methylation and gene transcription expression of lung adenocarcinoma by identifying the differentially expressed genes, DNA methylated locis, functional gene clusters, and the relevant genes associated with the survival. We identified 17 differentially expressed genes which had differentially methylated locis, 4 functional gene clusters regulated by methylation, and 522 genes, which were relevant to the survival time of patients. Our study suggested that methylation controlled the gene expression in a variety of ways, which had high/low expression and hyper-/hypo-methylation. Genes of different methylation status showed the different survival curve. The genes and methylated locis identified in this study could be potential biomarkers and therapeutic targets for lung adenocarcinoma.

Overlapping functional modules detection in PPI network with pair-wise constrained non-negative matrix tri-factorisation

A large amount of available protein–protein interaction (PPI) data has been generated by high‐throughput experimental techniques. Uncovering functional modules from PPI networks will help us better understand the underlying mechanisms of cellular functions. Numerous computational algorithms have been designed to identify functional modules automatically in the past decades. However, most community detection methods (non‐overlapping or overlapping types) are unsupervised models, which cannot incorporate the well‐known protein complexes as a priori. The authors propose a novel semi‐supervised model named pairwise constrains nonnegative matrix tri‐factorisation (PCNMTF), which takes full advantage of the well‐known protein complexes to find overlapping functional modules based on protein module indicator matrix and module correlation matrix simultaneously from PPI networks. PCNMTF determinately models and learns the mixed module memberships of each protein by considering the correlation among modules simultaneously based on the non‐negative matrix tri‐factorisation. The experiment results on both synthetic and real‐world biological networks demonstrate that PCNMTF gains more precise functional modules than that of state‐of‐the‐art methods.

Application of a co‑expression network for the analysis of aggressive and non‑aggressive breast cancer cell lines to predict the clinical outcome of patients

Breast cancer metastasis is a demanding problem in clinical treatment of patients with breast cancer. It is necessary to examine the mechanisms of metastasis for developing therapies. Classification of the aggressiveness of breast cancer is an important issue in biological study and for clinical decisions. Although aggressive and non‑aggressive breast cancer cells can be easily distinguished among different cell lines, it is very difficult to distinguish in clinical practice. The aim of the current study was to use the gene expression analysis from breast cancer cell lines to predict clinical outcomes of patients with breast cancer. Weighted gene co‑expression network analysis (WGCNA) is a powerful method to account for correlations between genes and extract co‑expressed modules of genes from large expression datasets. Therefore, WGCNA was applied to explore the differences in sub‑networks between aggressive and non‑aggressive breast cancer cell lines. The greatest difference topological overlap networks in both groups include potential information to understand the mechanisms of aggressiveness. The results show that the blue and red modules were significantly associated with the biological processes of aggressiveness. The sub‑network, which consisted of TMEM47, GJC1, ANXA3, TWIST1 and C19orf33 in the blue module, was associated with an aggressive phenotype. The sub‑network of LOC100653217, CXCL12, SULF1, DOK5 and DKK3 in the red module was associated with a non‑aggressive phenotype. In order to validate the hazard ratio of these genes, the prognostic index was constructed to integrate them and examined using data from the Cancer Genomic Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Patients with breast cancer from TCGA in the high‑risk group had a significantly shorter overall survival time compared with patients in the low‑risk group (hazard ratio=1.231, 95% confidence interval=1.058‑1.433, P=0.0071, by the Wald test). A similar result was produced from the GEO database. The findings may provide a novel strategy for measuring cancer aggressiveness in patients with breast cancer.

A matrix rank based concordance index for evaluating and detecting conditional specific co-expressed gene modules

BACKGROUND

Gene co-expression network analysis (GCNA) is widely adopted in bioinformatics and biomedical research with applications such as gene function prediction, protein-protein interaction inference, disease markers identification, and copy number variance discovery. Currently there is a lack of rigorous analysis on the mathematical condition for which the co-expressed gene module should satisfy.

METHODS

In this paper, we present a linear algebraic based Centralized Concordance Index (CCI) for evaluating the concordance of co-expressed gene modules from gene co-expression network analysis. The CCI can be used to evaluate the performance for co-expression network analysis algorithms as well as for detecting condition specific co-expression modules. We applied CCI in detecting lung tumor specific gene modules.

RESULTS AND DISCUSSION

Simulation showed that CCI is a robust indicator for evaluating the concordance of a group of co-expressed genes. The application to lung cancer datasets revealed interesting potential tumor specific genetic alterations including CNVs and even hints for gene-fusion. Deeper analysis required for understanding the molecular mechanisms of all such condition specific co-expression relationships.

CONCLUSION

The CCI can be used to evaluate the performance for co-expression network analysis algorithms as well as for detecting condition specific co-expression modules. It is shown to be more robust to outliers and interfering modules than density based on Pearson correlation coefficients.

Normalized lmQCM: An Algorithm for Detecting Weak Quasi-Cliques in Weighted Graph with Applications in Gene Co-Expression Module Discovery in Cancers

In this paper, we present a new approach for mining weighted networks to identify densely connected modules such as quasi-cliques. Quasi-cliques are densely connected subnetworks in a network. Detecting quasi-cliques is an important topic in data mining, with applications such as social network study and biomedicine. Our approach has two major improvements upon previous work. The first is the use of local maximum edges to initialize the search in order to avoid excessive overlaps among the modules, thereby greatly reducing the computing time. The second is the inclusion of a weight normalization procedure to enable discovery of "subtle" modules with more balanced sizes. We carried out careful tests on multiple parameters and settings using two large cancer datasets. This approach allowed us to identify a large number of gene modules enriched in both biological functions and chromosomal bands in cancer data, suggesting potential roles of copy number variations (CNVs) involved in the cancer development. We then tested the genes in selected modules with enriched chromosomal bands using The Cancer Genome Atlas data, and the results strongly support our hypothesis that the coexpression in these modules are associated with CNVs. While gene coexpression network analyses have been widely adopted in disease studies, most of them focus on the functional relationships of coexpressed genes. The relationship between coexpression gene modules and CNVs are much less investigated despite the potential advantage that we can infer from such relationship without genotyping data. Our new approach thus provides a means to carry out deep mining of the gene coexpression network to obtain both functional and genetic information from the expression data.

Gene Co-Expression Analysis Predicts Genetic Variants Associated with Drug Responsiveness in Lung Cancer

Responsiveness to drugs is an important concern in designing personalized treatment for cancer patients. Currently genetic markers are often used to guide targeted therapy. However, deeper understanding of the molecular basis for drug responses and discovery of new predictive biomarkers for drug sensitivity are much needed. In this paper, we present a workflow for identifying condition-specific gene co-expression networks associated with responses to the tyrosine kinase inhibitor, Erlotinib, in lung adenocarcinoma cell lines using data from the Cancer Cell Line Encyclopedia by combining network mining and statistical analysis. Particularly, we have identified multiple gene modules specifically co-expressed in the drug responsive cell lines but not in the unresponsive group. Interestingly, most of these modules are enriched on specific cytobands, suggesting potential copy number variation events on these loci. Our results therefore imply that there are multiple genetic loci with copy number variations associated with the Erlotinib responses. The existence of CNVs in these loci is also confirmed in lung cancer tissue samples using the TCGA data. Since these structural variations are inferred from functional genomics data, these CNVs are functional variations. These results suggest the condition specific gene co- expression network mining approach is an effective approach in predicting candidate biomarkers for drug responses.

Dynamic tracking of functional gene modules in treated juvenile idiopathic arthritis

BACKGROUND

We have previously shown that childhood-onset rheumatic diseases show aberrant patterns of gene expression that reflect pathology-associated co-expression networks. In this study, we used novel computational approaches to examine how disease-associated networks are altered in one of the most common rheumatic diseases of childhood, juvenile idiopathic arthritis (JIA).

METHODS

Using whole blood gene expression profiles derived from children in a pediatric rheumatology clinical trial, we used a network approach to understanding the impact of therapy and the underlying biology of response/non-response to therapy.

RESULTS

We demonstrate that therapy for JIA is associated with extensive re-ordering of gene expression networks, even in children who respond inadequately to therapy. Furthermore, we observe distinct differences in the evolution of specific network properties when we compare children who have been treated successfully with those who have inadequate treatment response.

CONCLUSIONS

Despite the inherent noisiness of whole blood gene expression data, our findings demonstrate how therapeutic response might be mapped and understood in pathologically informative cells in a broad range of human inflammatory diseases.

Genomic and transcriptome analysis revealing an oncogenic functional module in meningiomas

OBJECT

Meningiomas are among the most common primary adult brain tumors. Although typically benign, roughly 2%-5% display malignant pathological features. The key molecular pathways involved in malignant transformation remain to be determined.

METHODS

Illumina expression microarrays were used to assess gene expression levels, and Illumina single-nucleotide polymorphism arrays were used to identify copy number variants in benign, atypical, and malignant meningiomas (19 tumors, including 4 malignant ones). The authors also reanalyzed 2 expression data sets generated on Affymetrix microarrays (n = 68, including 6 malignant ones; n = 56, including 3 malignant ones). A weighted gene coexpression network approach was used to identify coexpression modules associated with malignancy.

RESULTS

At the genomic level, malignant meningiomas had more chromosomal losses than atypical and benign meningiomas, with average length of 528, 203, and 34 megabases, respectively. Monosomic loss of chromosome 22 was confirmed to be one of the primary chromosomal level abnormalities in all subtypes of meningiomas. At the transcriptome level, the authors identified 23 coexpression modules from the weighted gene coexpression network. Gene functional enrichment analysis highlighted a module with 356 genes that was highly related to tumorigenesis. Four intramodular hubs within the module (GAB2, KLF2, ID1, and CTF1) were oncogenic in other cancers such as leukemia. A putative meningioma tumor suppressor MN1 was also identified in this module with differential expression between malignant and benign meningiomas.

CONCLUSIONS

The authors' genomic and transcriptome analysis of meningiomas provides novel insights into the molecular pathways involved in malignant transformation of meningiomas, with implications for molecular heterogeneity of the disease.

Mining the tissue-tissue gene co-expression network for tumor microenvironment study and biomarker prediction

BACKGROUND

Recent discovery in tumor development indicates that the tumor microenvironment (mostly stroma cells) plays an important role in cancer development. To understand how the tumor microenvironment (TME) interacts with the tumor, we explore the correlation of the gene expressions between tumor and stroma. The tumor and stroma gene expression data are modeled as a weighted bipartite network (tumor-stroma coexpression network) where the weight of an edge indicates the correlation between the expression profiles of the corresponding tumor gene and stroma gene. In order to efficiently mine this weighted bipartite network, we developed the Bipartite subnetwork Component Mining algorithm (BCM), and we show that the BCM algorithm can efficiently mine weighted bipartite networks for dense Bipartite sub-Networks (BiNets) with density guarantees.

RESULTS

We applied BCM to the tumor-stroma coexpression network and find 372 BiNets that demonstrate statistical significance in survival tests. A good number of these BiNets demonstrate strong prognosis powers on at least one breast cancer patient cohort, which suggests that these BiNets are potential biomarkers for breast cancer prognosis. Further study on these 372 BiNets by the network merging approach reveals that they form 10 macro bipartite networks which show orchestrated key biological processes in both tumor and stroma. In addition, by further examining the BiNets that are significant in ER-negative breast cancer patient prognosis, we discovered a ubiquitin C (UBC) gene network that demonstrates strong prognosis power in nearly all types of breast cancer subtypes we used in this study.

CONCLUSIONS

The results support our hypothesis that the UBC gene network plays an important role in breast cancer prognosis and therapy and it is a potential prognostic biomarker for multiple breast cancer subtypes.

Feature selection and survival modeling in The Cancer Genome Atlas

Purpose

Personalized medicine is predicated on the concept of identifying subgroups of a common disease for better treatment. Identifying biomarkers that predict disease subtypes has been a major focus of biomedical science. In the era of genome-wide profiling, there is controversy as to the optimal number of genes as an input of a feature selection algorithm for survival modeling.

Patients and methods

The expression profiles and outcomes of 544 patients were retrieved from The Cancer Genome Atlas. We compared four different survival prediction methods: (1) 1-nearest neighbor (1-NN) survival prediction method; (2) random patient selection method and a Cox-based regression method with nested cross-validation; (3) least absolute shrinkage and selection operator (LASSO) optimization using whole-genome gene expression profiles; or (4) gene expression profiles of cancer pathway genes.

Results

The 1-NN method performed better than the random patient selection method in terms of survival predictions, although it does not include a feature selection step. The Cox-based regression method with LASSO optimization using whole-genome gene expression data demonstrated higher survival prediction power than the 1-NN method, but was outperformed by the same method when using gene expression profiles of cancer pathway genes alone.

Conclusion

The 1-NN survival prediction method may require more patients for better performance, even when omitting censored data. Using preexisting biological knowledge for survival prediction is reasonable as a means to understand the biological system of a cancer, unless the analysis goal is to identify completely unknown genes relevant to cancer biology.