Weighted gene coexpression network analysis strategies applied to mouse weight

Differential gene network analysis for the identification of asthma-associated therapeutic targets in allergen-specific T-helper memory responses

BACKGROUND

Asthma is strongly associated with allergic sensitization, but the mechanisms that determine why only a subset of atopics develop asthma are not well understood. The aim of this study was to test the hypothesis that variations in allergen-driven CD4 T cell responses are associated with susceptibility to expression of asthma symptoms.

METHODS

The study population consisted of house dust mite (HDM) sensitized atopics with current asthma (n = 22), HDM-sensitized atopics without current asthma (n = 26), and HDM-nonsensitized controls (n = 24). Peripheral blood mononuclear cells from these groups were cultured in the presence or absence of HDM extract for 24 h. CD4 T cells were then isolated by immunomagnetic separation, and gene expression patterns were profiled on microarrays.

RESULTS

Differential network analysis of HDM-induced CD4 T cell responses in sensitized atopics with or without asthma unveiled a cohort of asthma-associated genes that escaped detection by more conventional data analysis techniques. These asthma-associated genes were enriched for targets of STAT6 signaling, and they were nested within a larger coexpression module comprising 406 genes. Upstream regulator analysis suggested that this module was driven primarily by IL-2, IL-4, and TNF signaling; reconstruction of the wiring diagram of the module revealed a series of hub genes involved in inflammation (IL-1B, NFkB, STAT1, STAT3), apoptosis (BCL2, MYC), and regulatory T cells (IL-2Ra, FoxP3). Finally, we identified several negative regulators of asthmatic CD4 T cell responses to allergens (e.g. IL-10, type I interferons, microRNAs, drugs, metabolites), and these represent logical candidates for therapeutic intervention.

CONCLUSION

Differential network analysis of allergen-induced CD4 T cell responses can unmask covert disease-associated genes and pin point novel therapeutic targets.

Different Statistical Approaches to Investigate Porcine Muscle Metabolome Profiles to Highlight New Biomarkers for Pork Quality Assessment

The aim of this study was to elucidate the underlying biochemical processes to identify potential key molecules of meat quality traits drip loss, pH of meat 1 h post-mortem (pH1), pH in meat 24 h post-mortem (pH24) and meat color. An untargeted metabolomics approach detected the profiles of 393 annotated and 1,600 unknown metabolites in 97 Duroc × Pietrain pigs. Despite obvious differences regarding the statistical approaches, the four applied methods, namely correlation analysis, principal component analysis, weighted network analysis (WNA) and random forest regression (RFR), revealed mainly concordant results. Our findings lead to the conclusion that meat quality traits pH1, pH24 and color are strongly influenced by processes of post-mortem energy metabolism like glycolysis and pentose phosphate pathway, whereas drip loss is significantly associated with metabolites of lipid metabolism. In case of drip loss, RFR was the most suitable method to identify reliable biomarkers and to predict the phenotype based on metabolites. On the other hand, WNA provides the best parameters to investigate the metabolite interactions and to clarify the complex molecular background of meat quality traits. In summary, it was possible to attain findings on the interaction of meat quality traits and their underlying biochemical processes. The detected key metabolites might be better indicators of meat quality especially of drip loss than the measured phenotype itself and potentially might be used as bio indicators.

Meta-Analysis of Tumor Stem-Like Breast Cancer Cells Using Gene Set and Network Analysis

Generally, cancer stem cells have epithelial-to-mesenchymal-transition characteristics and other aggressive properties that cause metastasis. However, there have been no confident markers for the identification of cancer stem cells and comparative methods examining adherent and sphere cells are widely used to investigate mechanism underlying cancer stem cells, because sphere cells have been known to maintain cancer stem cell characteristics. In this study, we conducted a meta-analysis that combined gene expression profiles from several studies that utilized tumorsphere technology to investigate tumor stem-like breast cancer cells. We used our own gene expression profiles along with the three different gene expression profiles from the Gene Expression Omnibus, which we combined using the ComBat method, and obtained significant gene sets using the gene set analysis of our datasets and the combined dataset. This experiment focused on four gene sets such as cytokine-cytokine receptor interaction that demonstrated significance in both datasets. Our observations demonstrated that among the genes of four significant gene sets, six genes were consistently up-regulated and satisfied the p-value of < 0.05, and our network analysis showed high connectivity in five genes. From these results, we established CXCR4, CXCL1 and HMGCS1, the intersecting genes of the datasets with high connectivity and p-value of < 0.05, as significant genes in the identification of cancer stem cells. Additional experiment using quantitative reverse transcription-polymerase chain reaction showed significant up-regulation in MCF-7 derived sphere cells and confirmed the importance of these three genes. Taken together, using meta-analysis that combines gene set and network analysis, we suggested CXCR4, CXCL1 and HMGCS1 as candidates involved in tumor stem-like breast cancer cells. Distinct from other meta-analysis, by using gene set analysis, we selected possible markers which can explain the biological mechanisms and suggested network analysis as an additional criterion for selecting candidates.

An Integrative Genomic Study Implicates the Postsynaptic Density in the Pathogenesis of Bipolar Disorder

Genome-wide association studies (GWAS) have identified several common variants associated with bipolar disorder (BD), but the biological meaning of these findings remains unclear. Integrative genomics-the integration of GWAS signals with gene expression data-may illuminate genes and gene networks that have key roles in the pathogenesis of BD. We applied weighted gene co-expression network analysis (WGCNA), which exploits patterns of co-expression among genes, to brain transcriptome data obtained by sequencing of poly-A RNA derived from postmortem dorsolateral prefrontal cortex from people with BD, along with age- and sex-matched controls. WGCNA identified 33 gene modules. Many of the modules corresponded closely to those previously reported in human cortex. Three modules were associated with BD, enriched for genes differentially expressed in BD, and also enriched for signals in prior GWAS of BD. Functional analysis of genes within these modules revealed significant enrichment of several functionally related sets of genes, especially those involved in the postsynaptic density (PSD). These results provide convergent support for the hypothesis that dysregulation of genes involved in the PSD is a key factor in the pathogenesis of BD. If replicated in larger samples, these findings could point toward new therapeutic targets for BD.

Inference for High-dimensional Differential Correlation Matrices

Motivated by differential co-expression analysis in genomics, we consider in this paper estimation and testing of high-dimensional differential correlation matrices. An adaptive thresholding procedure is introduced and theoretical guarantees are given. Minimax rate of convergence is established and the proposed estimator is shown to be adaptively rate-optimal over collections of paired correlation matrices with approximately sparse differences. Simulation results show that the procedure significantly outperforms two other natural methods that are based on separate estimation of the individual correlation matrices. The procedure is also illustrated through an analysis of a breast cancer dataset, which provides evidence at the gene co-expression level that several genes, of which a subset has been previously verified, are associated with the breast cancer. Hypothesis testing on the differential correlation matrices is also considered. A test, which is particularly well suited for testing against sparse alternatives, is introduced. In addition, other related problems, including estimation of a single sparse correlation matrix, estimation of the differential covariance matrices, and estimation of the differential cross-correlation matrices, are also discussed.

Liver transcriptomic networks reveal main biological processes associated with feed efficiency in beef cattle

BACKGROUND

The selection of beef cattle for feed efficiency (FE) traits is very important not only for productive and economic efficiency but also for reduced environmental impact of livestock. Considering that FE is multifactorial and expensive to measure, the aim of this study was to identify biological functions and regulatory genes associated with this phenotype.

RESULTS

Eight genes were differentially expressed between high and low feed efficient animals (HFE and LFE, respectively). Co-expression analyses identified 34 gene modules of which 4 were strongly associated with FE traits. They were mainly enriched for inflammatory response or inflammation-related terms. We also identified 463 differentially co-expressed genes which were functionally enriched for immune response and lipid metabolism. A total of 8 key regulators of gene expression profiles affecting FE were found. The LFE animals had higher feed intake and increased subcutaneous and visceral fat deposition. In addition, LFE animals showed higher levels of serum cholesterol and liver injury biomarker GGT. Histopathology of the liver showed higher percentage of periportal inflammation with mononuclear infiltrate.

CONCLUSION

Liver transcriptomic network analysis coupled with other results demonstrated that LFE animals present altered lipid metabolism and increased hepatic periportal lesions associated with an inflammatory response composed mainly by mononuclear cells. We are now focusing to identify the causes of increased liver lesions in LFE animals.

Ethical and public policy challenges for pharmacogenomics

It is timely to consider the ethical and social questions raised by progress in pharmacogenomics, based on the current importance of pharmacogenomics for avoidance of predictable side effects of drugs, and for correct choice of medications in certain cancers. It has been proposed that the entire population be genotyped for drug-metabolizing enzyme polymorphisms, as a measure that would prevent many untoward and dangerous drug reactions. Pharmacologic treatment targeting based on genomics of disease can be expected to increase greatly in the coming years. Policy and ethical issues exist on consent for large-scale genomic pharmacogenomic data collection, public vs corporate ownership of genomic research results, testing efficacy and safety of drugs used for rare genomic indications, and accessibility of treatments based on costly research that is applicable to relatively few patients. In major psychiatric disorders and intellectual deficiency, rare and de novo deletion or duplication of chromosomal segments (copy number variation), in the aggregate, are common causes of increased risk. This implies that the policy problems of pharmacogenomics will be particularly important for the psychiatric disorders.

Genes located in a chromosomal inversion are correlated with territorial song in white-throated sparrows

The genome of the white-throated sparrow (Zonotrichia albicollis) contains an inversion polymorphism on chromosome 2 that is linked to predictable variation in a suite of phenotypic traits including plumage color, aggression and parental behavior. Differences in gene expression between the two color morphs, which represent the two common inversion genotypes (ZAL2/ZAL2 and ZAL2/ZAL2(m) ), may therefore advance our understanding of the molecular underpinnings of these phenotypes. To identify genes that are differentially expressed between the two morphs and correlated with behavior, we quantified gene expression and terrirorial aggression, including song, in a population of free-living white-throated sparrows. We analyzed gene expression in two brain regions, the medial amygdala (MeA) and hypothalamus. Both regions are part of a 'social behavior network', which is rich in steroid hormone receptors and previously linked with territorial behavior. Using weighted gene co-expression network analyses, we identified modules of genes that were correlated with both morph and singing behavior. The majority of these genes were located within the inversion, showing the profound effect of the inversion on the expression of genes captured by the rearrangement. These modules were enriched with genes related to retinoic acid signaling and basic cellular functioning. In the MeA, the most prominent pathways were those related to steroid hormone receptor activity. Within these pathways, the only gene encoding such a receptor was ESR1 (estrogen receptor 1), a gene previously shown to predict song rate in this species. The set of candidate genes we identified may mediate the effects of a chromosomal inversion on territorial behavior.

Dual-trait selection for ethanol consumption and withdrawal: genetic and transcriptional network effects

BACKGROUND

Data from C57BL/6J (B6) × DBA/2J (D2) F2 intercrosses (B6xD2 F2 ), standard and recombinant inbred strains, and heterogeneous stock mice indicate that a reciprocal (or inverse) genetic relationship exists between alcohol consumption and withdrawal severity. Furthermore, some genetic studies have detected reciprocal quantitative trait loci (QTLs) for these traits. We used a novel mouse model developed by simultaneous selection for both high alcohol consumption/low withdrawal and low alcohol consumption/high withdrawal and analyzed the gene expression and genome-wide genotypic differences.

METHODS

Randomly chosen third selected generation (S3 ) mice (N = 24/sex/line), bred from a B6xD2 F2 , were genotyped using the Mouse Universal Genotyping Array, which provided 2,760 informative markers. QTL analysis used a marker-by-marker strategy with the threshold for a significant log of the odds (LOD) set at 10. Gene expression in the ventral striatum was measured using the Illumina Mouse 8.2 array. Differential gene expression and the weighted gene co-expression network analysis (WGCNA) were implemented.

RESULTS

Significant QTLs for consumption/withdrawal were detected on chromosomes (Chr) 2, 4, 9, and 12. A suggestive QTL mapped to Chr 6. Some of the QTLs overlapped with known QTLs mapped for 1 of the traits individually. One thousand seven hundred and forty-five transcripts were detected as being differentially expressed between the lines; there was some overlap with known withdrawal genes (e.g., Mpdz) located within QTL regions. WGCNA revealed several modules of co-expressed genes showing significant effects in both differential expression and intramodular connectivity; a module richly annotated with kinase-related annotations was most affected.

CONCLUSIONS

Marked effects of selection on expression and network structure were detected. QTLs overlapping with differentially expressed genes on Chr 2 (distal) and 4 suggest that these are cis-eQTLs (Chr 2: Kif3b, Kcnq2; Chr 4: Mpdz, Snapc3). Other QTLs identified were on Chr 2 (proximal), 9, and 12. Network results point to involvement of kinase-related mechanisms and outline the need for further efforts such as interrogation of noncoding RNAs.

Module Based Differential Coexpression Analysis Method for Type 2 Diabetes

More and more studies have shown that many complex diseases are contributed jointly by alterations of numerous genes. Genes often coordinate together as a functional biological pathway or network and are highly correlated. Differential coexpression analysis, as a more comprehensive technique to the differential expression analysis, was raised to research gene regulatory networks and biological pathways of phenotypic changes through measuring gene correlation changes between disease and normal conditions. In this paper, we propose a gene differential coexpression analysis algorithm in the level of gene sets and apply the algorithm to a publicly available type 2 diabetes (T2D) expression dataset. Firstly, we calculate coexpression biweight midcorrelation coefficients between all gene pairs. Then, we select informative correlation pairs using the “differential coexpression threshold” strategy. Finally, we identify the differential coexpression gene modules using maximum clique concept and k-clique algorithm. We apply the proposed differential coexpression analysis method on simulated data and T2D data. Two differential coexpression gene modules about T2D were detected, which should be useful for exploring the biological function of the related genes.

Novel structural co-expression analysis linking the NPM1-associated ribosomal biogenesis network to chronic myelogenous leukemia

Co-expression analysis reveals useful dysregulation patterns of gene cooperativeness for understanding cancer biology and identifying new targets for treatment. We developed a structural strategy to identify co-expressed gene networks that are important for chronic myelogenous leukemia (CML). This strategy compared the distributions of expressional correlations between CML and normal states, and it identified a data-driven threshold to classify strongly co-expressed networks that had the best coherence with CML. Using this strategy, we found a transcriptome-wide reduction of co-expression connectivity in CML, reflecting potentially loosened molecular regulation. Conversely, when we focused on nucleophosmin 1 (NPM1) associated networks, NPM1 established more co-expression linkages with BCR-ABL pathways and ribosomal protein networks in CML than normal. This finding implicates a new role of NPM1 in conveying tumorigenic signals from the BCR-ABL oncoprotein to ribosome biogenesis, affecting cellular growth. Transcription factors may be regulators of the differential co-expression patterns between CML and normal.

Whole blood microarray analysis of pigs showing extreme phenotypes after a porcine reproductive and respiratory syndrome virus infection

Background

The presence of variability in the response of pigs to Porcine Reproductive and Respiratory Syndrome virus (PRRSv) infection, and recent demonstration of significant genetic control of such responses, leads us to believe that selection towards more disease resistant pigs could be a valid strategy to reduce its economic impact on the swine industry. To find underlying molecular differences in PRRS susceptible versus more resistant pigs, 100 animals with extremely different growth rates and viremia levels after PRRSv infection were selected from a total of 600 infected pigs. A microarray experiment was conducted on whole blood RNA samples taken at 0, 4 and 7 days post infection (dpi) from these pigs. From these data, we examined associations of gene expression with weight gain and viral load phenotypes. The single nucleotide polymorphism (SNP) marker WUR10000125 (WUR) on the porcine 60 K SNP chip was shown to be associated with viral load and weight gain after PRRSv infection, and so the effect of the WUR10000125 (WUR) genotype on expression in whole blood was also examined.

Results

Limited information was obtained through linear modeling of blood gene differential expression (DE) that contrasted pigs with extreme phenotypes, for growth or viral load or between animals with different WUR genotype. However, using network-based approaches, molecular pathway differences between extreme phenotypic classes could be identified. Several gene clusters of interest were found when Weighted Gene Co-expression Network Analysis (WGCNA) was applied to 4dpi contrasted with 0dpi data. The expression pattern of one such cluster of genes correlated with weight gain and WUR genotype, contained numerous immune response genes such as cytokines, chemokines, interferon type I stimulated genes, apoptotic genes and genes regulating complement activation. In addition, Partial Correlation and Information Theory (PCIT) identified differentially hubbed (DH) genes between the phenotypically divergent groups. GO enrichment revealed that the target genes of these DH genes are enriched in adaptive immune pathways.

Conclusion

There are molecular differences in blood RNA patterns between pigs with extreme phenotypes or with a different WUR genotype in early responses to PRRSv infection, though they can be quite subtle and more difficult to discover with conventional DE expression analyses. Co-expression analyses such as WGCNA and PCIT can be used to reveal network differences between such extreme response groups.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1741-8) contains supplementary material, which is available to authorized users.

Construction of citrus gene coexpression networks from microarray data using random matrix theory

After the sequencing of citrus genomes, gene function annotation is becoming a new challenge. Gene coexpression analysis can be employed for function annotation using publicly available microarray data sets. In this study, 230 sweet orange (Citrus sinensis) microarrays were used to construct seven coexpression networks, including one condition-independent and six condition-dependent (Citrus canker, Huanglongbing, leaves, flavedo, albedo, and flesh) networks. In total, these networks contain 37 633 edges among 6256 nodes (genes), which accounts for 52.11% measurable genes of the citrus microarray. Then, these networks were partitioned into functional modules using the Markov Cluster Algorithm. Significantly enriched Gene Ontology biological process terms and KEGG pathway terms were detected for 343 and 60 modules, respectively. Finally, independent verification of these networks was performed using another expression data of 371 genes. This study provides new targets for further functional analyses in citrus.

Identify the signature genes for diagnose of uveal melanoma by weight gene co-expression network analysis

AIM

To identify and understand the relationship between co-expression pattern and clinic traits in uveal melanoma, weighted gene co-expression network analysis (WGCNA) is applied to investigate the gene expression levels and patient clinic features. Uveal melanoma is the most common primary eye tumor in adults. Although many studies have identified some important genes and pathways that were relevant to progress of uveal melanoma, the relationship between co-expression and clinic traits in systems level of uveal melanoma is unclear yet. We employ WGCNA to investigate the relationship underlying molecular and phenotype in this study.

METHODS

Gene expression profile of uveal melanoma and patient clinic traits were collected from the Gene Expression Omnibus (GEO) database. The gene co-expression is calculated by WGCNA that is the R package software. The package is used to analyze the correlation between pairs of expression levels of genes. The function of the genes were annotated by gene ontology (GO).

RESULTS

In this study, we identified four co-expression modules significantly correlated with clinic traits. Module blue positively correlated with radiotherapy treatment. Module purple positively correlates with tumor location (sclera) and negatively correlates with patient age. Module red positively correlates with sclera and negatively correlates with thickness of tumor. Module black positively correlates with the largest tumor diameter (LTD). Additionally, we identified the hug gene (top connectivity with other genes) in each module. The hub gene RPS15A, PTGDS, CD53 and MSI2 might play a vital role in progress of uveal melanoma.

CONCLUSION

From WGCNA analysis and hub gene calculation, we identified RPS15A, PTGDS, CD53 and MSI2 might be target or diagnosis for uveal melanoma.

Gene network exploration of crosstalk between apoptosis and autophagy in chronic myelogenous leukemia

BACKGROUND

Gene expression levels change to adapt the stress, such as starvation, toxin, and radiation. The changes are signals transmitted through molecular interactions, eventually leading to two cellular fates, apoptosis and autophagy. Due to genetic variations, the signals may not be effectively transmitted to modulate apoptotic and autophagic responses. Such aberrant modulation may lead to carcinogenesis and drug resistance. The balance between apoptosis and autophagy becomes very crucial in coping with the stress. Though there have been evidences illustrating the apoptosis-autophagy interplay, the underlying mechanism and the participation of the regulators including transcription factors (TFs) and microRNAs (miRNAs) remain unclear.

RESULTS

Gene network is a graphical illustration for exploring the functional linkages and the potential coordinate regulations of genes. Microarray dataset for the study of chronic myeloid leukemia was obtained from Gene Expression Omnibus. The expression profiles of those genes related to apoptosis and autophagy, including MCL1, BCL2, ATG, beclin-1, BAX, BAK, E2F, cMYC, PI3K, AKT, BAD, and LC3, were extracted from the dataset to construct the gene networks.

CONCLUSION

The network analysis of these genes explored the underlying mechanisms and the roles of TFs and miRNAs for the crosstalk between apoptosis and autophagy.

Comparative network analysis via differential graphlet communities

While current protein interaction data provides a rich resource for molecular biology, it mostly lacks condition-specific details. Abundance of mRNA data for most diseases provides potential to model condition-specific transcriptional changes. Transcriptional data enables modeling disease mechanisms, and in turn provide potential treatments. While approaches to compare networks constructed from healthy and disease samples have been developed, they do not provide the complete comparison, evaluations are performed on very small networks, or no systematic network analyses are performed on differential network structures. We propose a novel method for efficiently exploiting network structure information in the comparison between any graphs, and validate results in non-small cell lung cancer. We introduce the notion of differential graphlet community to detect deregulated subgraphs between any graphs such that the network structure information is exploited. The differential graphlet community approach systematically captures network structure differences between any graphs. Instead of using connectivity of each protein or each edge, we used shortest path distributions on differential graphlet communities in order to exploit network structure information on identified deregulated subgraphs. We validated the method by analyzing three non-small cell lung cancer datasets and validated results on four independent datasets. We observed that the shortest path lengths are significantly longer for normal graphs than for tumor graphs between genes that are in differential graphlet communities, suggesting that tumor cells create "shortcuts" between biological processes that may not be present in normal conditions.

Gene differential coexpression analysis based on biweight correlation and maximum clique

Differential coexpression analysis usually requires the definition of 'distance' or 'similarity' between measured datasets. Until now, the most common choice is Pearson correlation coefficient. However, Pearson correlation coefficient is sensitive to outliers. Biweight midcorrelation is considered to be a good alternative to Pearson correlation since it is more robust to outliers. In this paper, we introduce to use Biweight Midcorrelation to measure 'similarity' between gene expression profiles, and provide a new approach for gene differential coexpression analysis. Firstly, we calculate the biweight midcorrelation coefficients between all gene pairs. Then, we filter out non-informative correlation pairs using the 'half-thresholding' strategy and calculate the differential coexpression value of gene, The experimental results on simulated data show that the new approach performed better than three previously published differential coexpression analysis (DCEA) methods. Moreover, we use the maximum clique analysis to gene subset included genes identified by our approach and previously reported T2D-related genes, many additional discoveries can be found through our method.

Transcriptional profiling and co-expression network analysis identifies potential biomarkers to differentiate chronic hepatitis B and the caused cirrhosis

Liver cirrhosis is one of the most common non-neoplastic causes of mortality worldwide. Chronic hepatitis B (CHB) is a major cause of liver cirrhosis in China. To find biomarkers for the diagnosis of CHB caused cirrhosis (HBC), we examined the transcriptional profiling of CHB and HBC. The leukocyte samples of CHB (n = 5) and HBC (n = 5) were analyzed by microarray. The results showed that 2128 mapped genes were differentially expressed between CHB and HBC (fold change ≥ 2.0, p < 0.05). Gene ontology (GO) analysis indicated that these 2128 differentially expressed genes (DEGs) were enriched for immune response and cell formation functions mostly. Moreover, co-expression networks using the k-core algorithm were established to determine the core genes, which may play important roles in the progression of CHB to HBC. There were markedly different gene co-expression patterns in CHB and HBC. We validated the five core genes, CASP1, TGFBI, IFI30, HLA-DMA and PAG1 in CHB (n = 60) and HBC (n = 60) by quantitative RT-PCR. The expression of the five genes were consistent with microarray, and there were statistically significant co-expression patterns of TGFβ1, PAG1 and HLA-DMA mRNA (Pearson correlation coefficient >0.6). Furthermore, we constructed an mRNA panel of TGFBI, IFI30, HLA-DMA and PAG1 (TIPH HBCtest) by means of a logistic regression model, and evaluated the TIPH HBCtest for HBC diagnosis by area under the receiver operating characteristic curve (AUC) analysis, which showed a higher accuracy (AUC = 0.903). This study suggested that there are particular transcriptional profiles, gene co-expression patterns and core genes in CHB and HBC. The TIPH HBC test may be useful in the diagnosis of HBC from CHB.

Identification of co-expression gene networks, regulatory genes and pathways for obesity based on adipose tissue RNA Sequencing in a porcine model

BACKGROUND

Obesity is a complex metabolic condition in strong association with various diseases, like type 2 diabetes, resulting in major public health and economic implications. Obesity is the result of environmental and genetic factors and their interactions, including genome-wide genetic interactions. Identification of co-expressed and regulatory genes in RNA extracted from relevant tissues representing lean and obese individuals provides an entry point for the identification of genes and pathways of importance to the development of obesity. The pig, an omnivorous animal, is an excellent model for human obesity, offering the possibility to study in-depth organ-level transcriptomic regulations of obesity, unfeasible in humans. Our aim was to reveal adipose tissue co-expression networks, pathways and transcriptional regulations of obesity using RNA Sequencing based systems biology approaches in a porcine model.

METHODS

We selected 36 animals for RNA Sequencing from a previously created F2 pig population representing three extreme groups based on their predicted genetic risks for obesity. We applied Weighted Gene Co-expression Network Analysis (WGCNA) to detect clusters of highly co-expressed genes (modules). Additionally, regulator genes were detected using Lemon-Tree algorithms.

RESULTS

WGCNA revealed five modules which were strongly correlated with at least one obesity-related phenotype (correlations ranging from -0.54 to 0.72, P < 0.001). Functional annotation identified pathways enlightening the association between obesity and other diseases, like osteoporosis (osteoclast differentiation, P = 1.4E-7), and immune-related complications (e.g. Natural killer cell mediated cytotoxity, P = 3.8E-5; B cell receptor signaling pathway, P = 7.2E-5). Lemon-Tree identified three potential regulator genes, using confident scores, for the WGCNA module which was associated with osteoclast differentiation: CCR1, MSR1 and SI1 (probability scores respectively 95.30, 62.28, and 34.58). Moreover, detection of differentially connected genes identified various genes previously identified to be associated with obesity in humans and rodents, e.g. CSF1R and MARC2.

CONCLUSIONS

To our knowledge, this is the first study to apply systems biology approaches using porcine adipose tissue RNA-Sequencing data in a genetically characterized porcine model for obesity. We revealed complex networks, pathways, candidate and regulatory genes related to obesity, confirming the complexity of obesity and its association with immune-related disorders and osteoporosis.

Co-expression analysis of fetal weight-related genes in ovine skeletal muscle during mid and late fetal development stages

BACKGROUND

Muscle development and lipid metabolism play important roles during fetal development stages. The commercial Texel sheep are more muscular than the indigenous Ujumqin sheep.

RESULTS

We performed serial transcriptomics assays and systems biology analyses to investigate the dynamics of gene expression changes associated with fetal longissimus muscles during different fetal stages in two sheep breeds. Totally, we identified 1472 differentially expressed genes during various fetal stages using time-series expression analysis. A systems biology approach, weighted gene co-expression network analysis (WGCNA), was used to detect modules of correlated genes among these 1472 genes. Dramatically different gene modules were identified in four merged datasets, corresponding to the mid fetal stage in Texel and Ujumqin sheep, the late fetal stage in Texel and Ujumqin sheep, respectively. We further detected gene modules significantly correlated with fetal weight, and constructed networks and pathways using genes with high significances. In these gene modules, we identified genes like TADA3, LMNB1, TGF-β3, EEF1A2, FGFR1, MYOZ1, and FBP2 correlated with fetal weight.

CONCLUSION

Our study revealed the complex network characteristics involved in muscle development and lipid metabolism during fetal development stages. Diverse patterns of the network connections observed between breeds and fetal stages could involve some hub genes, which play central roles in fetal development, correlating with fetal weight. Our findings could provide potential valuable biomarkers for selection of body weight-related traits in sheep and other livestock.

Gene Co-expression Analysis to Characterize Genes Related to Marbling Trait in Hanwoo (Korean) Cattle

Marbling (intramuscular fat) is an important trait that affects meat quality and is a casual factor determining the price of beef in the Korean beef market. It is a complex trait and has many biological pathways related to muscle and fat. There is a need to identify functional modules or genes related to marbling traits and investigate their relationships through a weighted gene co-expression network analysis based on the system level. Therefore, we investigated the co-expression relationships of genes related to the 'marbling score' trait and systemically analyzed the network topology in Hanwoo (Korean cattle). As a result, we determined 3 modules (gene groups) that showed statistically significant results for marbling score. In particular, one module (denoted as red) has a statistically significant result for marbling score (p = 0.008) and intramuscular fat (p = 0.02) and water capacity (p = 0.006). From functional enrichment and relationship analysis of the red module, the pathway hub genes (IL6, CHRNE, RB1, INHBA and NPPA) have a direct interaction relationship and share the biological functions related to fat or muscle, such as adipogenesis or muscle growth. This is the first gene network study with m.logissimus in Hanwoo to observe co-expression patterns in divergent marbling phenotypes. It may provide insights into the functional mechanisms of the marbling trait.

Gene co-expression network analysis identifies porcine genes associated with variation in Salmonella shedding

Background

Salmonella enterica serovar Typhimurium is a gram-negative bacterium that can colonise the gut of humans and several species of food producing farm animals to cause enteric or septicaemic salmonellosis. While many studies have looked into the host genetic response to Salmonella infection, relatively few have used correlation of shedding traits with gene expression patterns to identify genes whose variable expression among different individuals may be associated with differences in Salmonella clearance and resistance. Here, we aimed to identify porcine genes and gene co-expression networks that differentiate distinct responses to Salmonella challenge with respect to faecal Salmonella shedding.

Results

Peripheral blood transcriptome profiles from 16 pigs belonging to extremes of the trait of faecal Salmonella shedding counts recorded up to 20 days post-inoculation (low shedders (LS), n = 8; persistent shedders (PS), n = 8) were generated using RNA-sequencing from samples collected just before (day 0) and two days after (day 2) Salmonella inoculation. Weighted gene co-expression network analysis (WGCNA) of day 0 samples identified four modules of co-expressed genes significantly correlated with Salmonella shedding counts upon future challenge. Two of those modules consisted largely of innate immunity related genes, many of which were significantly up-regulated at day 2 post-inoculation. The connectivity at both days and the mean gene-wise expression levels at day 0 of the genes within these modules were higher in networks constructed using LS samples alone than those using PS alone. Genes within these modules include those previously reported to be involved in Salmonella resistance such as SLC11A1 (formerly NRAMP1), TLR4, CD14 and CCR1 and those for which an association with Salmonella is novel, for example, SIGLEC5, IGSF6 and TNFSF13B.

Conclusions

Our analysis integrates gene co-expression network analysis, gene-trait correlations and differential expression to provide new candidate regulators of Salmonella shedding in pigs. The comparatively higher expression (also confirmed in an independent dataset) and the significantly higher connectivity of genes within the Salmonella shedding associated modules in LS compared to PS even before Salmonella challenge may be factors that contribute to the decreased faecal Salmonella shedding observed in LS following challenge.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-452) contains supplementary material, which is available to authorized users.

dna: An R package for differential network analysis

Differential network analysis provides a framework for examining if there is sufficient statistical evidence to conclude that the structure of a network differs under two experimental conditions or if the structures of two networks are different. The R package dna provides tools and procedures for differential network analysis of genomic data. The focus of this package is on gene-gene networks, but the methods are easily adaptable for more general biological processes. This package includes preprocessing tools for simultaneously preparing a pair of networks for analysis, procedures for computing connectivity scores between pairs of genes based on many available statistical techniques, and tools for handling modules of genes based on these scores. Also, procedures are provided for performing permutation tests based on these scores to determine if the connectivity of a gene differs between the two networks, to determine if the connectivity of a particular set of important genes differs between the two networks, and to determine if the overall module structure differs between the two networks. Several built-in options are available for the types of scores and distances used in the testing procedures, and additionally, the procedures provide flexible methods that allow the user to define custom scores and distances.

Weighted Interaction SNP Hub (WISH) network method for building genetic networks for complex diseases and traits using whole genome genotype data

Background

High-throughput genotype (HTG) data has been used primarily in genome-wide association (GWA) studies; however, GWA results explain only a limited part of the complete genetic variation of traits. In systems genetics, network approaches have been shown to be able to identify pathways and their underlying causal genes to unravel the biological and genetic background of complex diseases and traits, e.g., the Weighted Gene Co-expression Network Analysis (WGCNA) method based on microarray gene expression data. The main objective of this study was to develop a scale-free weighted genetic interaction network method using whole genome HTG data in order to detect biologically relevant pathways and potential genetic biomarkers for complex diseases and traits.

Results

We developed the Weighted Interaction SNP Hub (WISH) network method that uses HTG data to detect genome-wide interactions between single nucleotide polymorphism (SNPs) and its relationship with complex traits. Data dimensionality reduction was achieved by selecting SNPs based on its: 1) degree of genome-wide significance and 2) degree of genetic variation in a population. Network construction was based on pairwise Pearson's correlation between SNP genotypes or the epistatic interaction effect between SNP pairs. To identify modules the Topological Overlap Measure (TOM) was calculated, reflecting the degree of overlap in shared neighbours between SNP pairs. Modules, clusters of highly interconnected SNPs, were defined using a tree-cutting algorithm on the SNP dendrogram created from the dissimilarity TOM (1-TOM). Modules were selected for functional annotation based on their association with the trait of interest, defined by the Genome-wide Module Association Test (GMAT). We successfully tested the established WISH network method using simulated and real SNP interaction data and GWA study results for carcass weight in a pig resource population; this resulted in detecting modules and key functional and biological pathways related to carcass weight.

Conclusions

We developed the WISH network method which is a novel 'systems genetics' approach to study genetic networks underlying complex trait variation. The WISH network method reduces data dimensionality and statistical complexity in associating genotypes with phenotypes in GWA studies and enables researchers to identify biologically relevant pathways and potential genetic biomarkers for any complex trait of interest.

Spatio-temporal analysis of type 2 diabetes mellitus based on differential expression networks

T2DM is complex in its dynamical dependence on multiple tissues, disease states, and factors' interactions. However, most existing work devoted to characterizing its pathophysiology from one static tissue, individual factors, or single state. Here we perform a spatio-temporal analysis on T2DM by developing a new form of molecular network, i.e. 'differential expression network' (DEN), which can reflect phenotype differences at network level. Static DENs show that three tissues (white adipose, skeletal muscle, and liver) all suffer from severe inflammation and perturbed metabolism, among which metabolic functions are seriously affected in liver. Dynamical analysis on DENs reveals metabolic function changes in adipose and liver are consistent with insulin resistance (IR) deterioration. Close investigation on IR pathway identifies 'disease interactions', revealing that IR deterioration is earlier than that on SlC2A4 in adipose and muscle. Our analysis also provides evidence that rising of insulin secretion is the root cause of IR in diabetes.

Machine learning-based differential network analysis: a study of stress-responsive transcriptomes in Arabidopsis

Machine learning (ML) is an intelligent data mining technique that builds a prediction model based on the learning of prior knowledge to recognize patterns in large-scale data sets. We present an ML-based methodology for transcriptome analysis via comparison of gene coexpression networks, implemented as an R package called machine learning-based differential network analysis (mlDNA) and apply this method to reanalyze a set of abiotic stress expression data in Arabidopsis thaliana. The mlDNA first used a ML-based filtering process to remove nonexpressed, constitutively expressed, or non-stress-responsive "noninformative" genes prior to network construction, through learning the patterns of 32 expression characteristics of known stress-related genes. The retained "informative" genes were subsequently analyzed by ML-based network comparison to predict candidate stress-related genes showing expression and network differences between control and stress networks, based on 33 network topological characteristics. Comparative evaluation of the network-centric and gene-centric analytic methods showed that mlDNA substantially outperformed traditional statistical testing-based differential expression analysis at identifying stress-related genes, with markedly improved prediction accuracy. To experimentally validate the mlDNA predictions, we selected 89 candidates out of the 1784 predicted salt stress-related genes with available SALK T-DNA mutagenesis lines for phenotypic screening and identified two previously unreported genes, mutants of which showed salt-sensitive phenotypes.

The genetics of gene expression in complex mouse crosses as a tool to study the molecular underpinnings of behavior traits

Complex Mus musculus crosses provide increased resolution to examine the relationships between gene expression and behavior. While the advantages are clear, there are numerous analytical and technological concerns that arise from the increased genetic complexity that must be considered. Each of these issues is discussed, providing an initial framework for complex cross study design and planning.

Analysis of time-resolved gene expression measurements across individuals

Genetic and environmental determinants of altered cellular function, disease state, and drug response are increasingly studied using time-resolved transcriptomic profiles. While it is widely acknowledged that the rate of biological processes may vary between individuals, data analysis approaches that go beyond evaluating differential expression of single genes have so far not taken this variability into account. To this end, we introduce here a robust multi-gene data analysis approach and evaluate it in a biomarker discovery scenario across four publicly available datasets. In our evaluation, existing methods perform surprisingly poorly on time-resolved data; only the approach taking the variability into account yields reproducible and biologically plausible results. Our results indicate the need to capture gene expression between potentially heterogeneous individuals at multiple time points, and highlight the importance of robust data analysis in the presence of heterogeneous gene expression responses.

Two gene co-expression modules differentiate psychotics and controls

Schizophrenia (SCZ) and bipolar disorder (BD) are highly heritable psychiatric disorders. Associated genetic and gene expression changes have been identified, but many have not been replicated and have unknown functions. We identified groups of genes whose expressions varied together, that is co-expression modules, then tested them for association with SCZ. Using weighted gene co-expression network analysis, we show that two modules were differentially expressed in patients versus controls. One, upregulated in cerebral cortex, was enriched with neuron differentiation and neuron development genes, as well as disease genome-wide association study genetic signals; the second, altered in cerebral cortex and cerebellum, was enriched with genes involved in neuron protection functions. The findings were preserved in five expression data sets, including sets from three brain regions, from a different microarray platform, and from BD patients. From those observations, we propose neuron differentiation and development pathways may be involved in etiologies of both SCZ and BD, and neuron protection function participates in pathological process of the diseases.

Gene Network and Proteomic Analyses of Cardiac Responses to Pathological and Physiological Stress

Background—

The molecular mechanisms underlying similarities and differences between physiological and pathological left ventricular hypertrophy (LVH) are of intense interest. Most previous work involved targeted analysis of individual signaling pathways or screening of transcriptomic profiles. We developed a network biology approach using genomic and proteomic data to study the molecular patterns that distinguish pathological and physiological LVH.

Methods and Results—

A network-based analysis using graph theory methods was undertaken on 127 genome-wide expression arrays of in vivo murine LVH. This revealed phenotype-specific pathological and physiological gene coexpression networks. Despite >1650 common genes in the 2 networks, network structure is significantly different. This is largely because of rewiring of genes that are differentially coexpressed in the 2 networks; this novel concept of differential wiring was further validated experimentally. Functional analysis of the rewired network revealed several distinct cellular pathways and gene sets. Deeper exploration was undertaken by targeted proteomic analysis of mitochondrial, myofilament, and extracellular subproteomes in pathological LVH. A notable finding was that mRNA–protein correlation was greater at the cellular pathway level than for individual loci.

Conclusions—

This first combined gene network and proteomic analysis of LVH reveals novel insights into the integrated pathomechanisms that distinguish pathological versus physiological phenotypes. In particular, we identify differential gene wiring as a major distinguishing feature of these phenotypes. This approach provides a platform for the investigation of potentially novel pathways in LVH and offers a freely accessible protocol ( http://sites.google.com/site/cardionetworks ) for similar analyses in other cardiovascular diseases.

Deciphering ascorbic acid regulatory pathways in ripening tomato fruit using a weighted gene correlation network analysis approach

Genotype is generally determined by the co-expression of diverse genes and multiple regulatory pathways in plants. Gene co-expression analysis combining with physiological trait data provides very important information about the gene function and regulatory mechanism. L-Ascorbic acid (AsA), which is an essential nutrient component for human health and plant metabolism, plays key roles in diverse biological processes such as cell cycle, cell expansion, stress resistance, hormone synthesis, and signaling. Here, we applied a weighted gene correlation network analysis approach based on gene expression values and AsA content data in ripening tomato (Solanum lycopersicum L.) fruit with different AsA content levels, which leads to identification of AsA relevant modules and vital genes in AsA regulatory pathways. Twenty-four modules were compartmentalized according to gene expression profiling. Among these modules, one negatively related module containing genes involved in redox processes and one positively related module enriched with genes involved in AsA biosynthetic and recycling pathways were further analyzed. The present work herein indicates that redox pathways as well as hormone-signal pathways are closely correlated with AsA accumulation in ripening tomato fruit, and allowed us to prioritize candidate genes for follow-up studies to dissect this interplay at the biochemical and molecular level.

The ontogeny of the human connectome: development and dynamic changes of brain connectivity across the life span

The human brain comprises distributed cortical regions that are structurally and functionally connected into a network that is known as the human connectome. Elaborate developmental processes starting in utero herald connectome genesis, with dynamic changes in its architecture continuing throughout life. Connectome changes during development, maturation, and aging may be governed by a set of biological rules or algorithms, forming and shaping the macroscopic architecture of the brain's wiring network. To explore the presence of developmental patterns indicative of such rules, this review considers insights from studies on the cellular and the systems level into macroscopic connectome genesis and dynamics across the life span. We observe that in parallel with synaptogenesis, macroscopic connectome formation and transformation is characterized by an initial overgrowth and subsequent elimination of cortico-cortical axonal projections. Furthermore, dynamic changes in connectome organization throughout the life span are suggested to follow an inverted U-shaped pattern, with an increasingly integrated topology during development, a plateau lasting for the majority of adulthood and an increasingly localized topology in late life. Elucidating developmental patterns in brain connectivity is crucial for our understanding of the human connectome and how it may give rise to brain function, including the occurrence of brain network disorders across the life span.

Leveraging gene co-expression networks to pinpoint the regulation of complex traits and disease, with a focus on cardiovascular traits

Over the past decade, the number of genome-scale transcriptional datasets in publicly available databases has climbed to nearly one million, providing an unprecedented opportunity for extensive analyses of gene co-expression networks. In systems-genetic studies of complex diseases researchers increasingly focus on groups of highly interconnected genes within complex transcriptional networks (referred to as clusters, modules or subnetworks) to uncover specific molecular processes that can inform functional disease mechanisms and pathological pathways. Here, we outline the basic paradigms underlying gene co-expression network analysis and critically review the most commonly used computational methods. Finally, we discuss specific applications of network-based approaches to the study of cardiovascular traits, which highlight the power of integrated analyses of networks, genetic and gene-regulation data to elucidate the complex mechanisms underlying cardiovascular disease.

Systems genetics of environmental response in the mature wheat embryo

Quantitative phenotypic traits are influenced by genetic and environmental variables as well as the interaction between the two. Underlying genetic × environment interaction is the influence that the surrounding environment exerts on gene expression. Perturbation of gene expression by environmental factors manifests itself in alterations to gene co-expression networks and ultimately in phenotypic plasticity. Comparative gene co-expression networks have been used to uncover biological mechanisms that differentiate tissues or other biological factors. In this study, we have extended consensus and differential Weighted Gene Co-Expression Network Analysis to compare the influence of different growing environments on gene co-expression in the mature wheat (Triticum aestivum) embryo. This network approach was combined with mapping of individual gene expression QTL to examine the genetic control of environmentally static and variable gene expression. The approach is useful for gene expression experiments containing multiple environments and allowed for the identification of specific gene co-expression modules responsive to environmental factors. This procedure identified conserved coregulation of gene expression between environments related to basic developmental and cellular functions, including protein localization and catabolism, vesicle composition/trafficking, Golgi transport, and polysaccharide metabolism among others. Environmentally unique modules were found to contain genes with predicted functions in responding to abiotic and biotic environmental variables. These findings represent the first report using consensus and differential Weighted Gene Co-expression Network Analysis to characterize the influence of environment on coordinated transcriptional regulation.

When is hub gene selection better than standard meta-analysis?

Since hub nodes have been found to play important roles in many networks, highly connected hub genes are expected to play an important role in biology as well. However, the empirical evidence remains ambiguous. An open question is whether (or when) hub gene selection leads to more meaningful gene lists than a standard statistical analysis based on significance testing when analyzing genomic data sets (e.g., gene expression or DNA methylation data). Here we address this question for the special case when multiple genomic data sets are available. This is of great practical importance since for many research questions multiple data sets are publicly available. In this case, the data analyst can decide between a standard statistical approach (e.g., based on meta-analysis) and a co-expression network analysis approach that selects intramodular hubs in consensus modules. We assess the performance of these two types of approaches according to two criteria. The first criterion evaluates the biological insights gained and is relevant in basic research. The second criterion evaluates the validation success (reproducibility) in independent data sets and often applies in clinical diagnostic or prognostic applications. We compare meta-analysis with consensus network analysis based on weighted correlation network analysis (WGCNA) in three comprehensive and unbiased empirical studies: (1) Finding genes predictive of lung cancer survival, (2) finding methylation markers related to age, and (3) finding mouse genes related to total cholesterol. The results demonstrate that intramodular hub gene status with respect to consensus modules is more useful than a meta-analysis p-value when identifying biologically meaningful gene lists (reflecting criterion 1). However, standard meta-analysis methods perform as good as (if not better than) a consensus network approach in terms of validation success (criterion 2). The article also reports a comparison of meta-analysis techniques applied to gene expression data and presents novel R functions for carrying out consensus network analysis, network based screening, and meta analysis.

Genome-wide associations for feed utilisation complex in primiparous Holstein-Friesian dairy cows from experimental research herds in four European countries

Genome-wide association studies for difficult-to-measure traits are generally limited by the sample size with accurate phenotypic data. The objective of this study was to utilise data on primiparous Holstein–Friesian cows from experimental farms in Ireland, the United Kingdom, the Netherlands and Sweden to identify genomic regions associated with the feed utilisation complex: fat and protein corrected milk yield (FPCM), dry matter intake (DMI), body condition score (BCS) and live-weight (LW). Phenotypic data and 37 590 single nucleotide polymorphisms (SNPs) were available on up to 1629 animals. Genetic parameters of the traits were estimated using a linear animal model with pedigree information, and univariate genome-wide association analyses were undertaken using Bayesian stochastic search variable selection performed using Gibbs sampling. The variation in the phenotypes explained by the SNPs on each chromosome was related to the size of the chromosome and was relatively consistent for each trait with the possible exceptions of BTA4 for BCS, BTA7, BTA13, BTA14, BTA18 for LW and BTA27 for DMI. For LW, BCS, DMI and FPCM, 266, 178, 206 and 254 SNPs had a Bayes factor .3, respectively. Olfactory genes and genes involved in the sensory smell process were overrepresented in a 500 kbp window around the significant SNPs. Potential candidate genes were involved with functions linked to insulin, epidermal growth factor and tryptophan.

Network-based approach identified cell cycle genes as predictor of overall survival in lung adenocarcinoma patients

Lung adenocarcinoma is the most common type of primary lung cancer. The purpose of this study was to delineate gene expression patterns for survival prediction in lung adenocarcinoma. Gene expression profiles of 82 (discovery set) and 442 (validation set 1) lung adenocarcinoma tumor tissues were analyzed using a systems biology-based network approach. We also examined the expression profiles of 78 adjacent normal lung tissues from 82 patients. We found a significant correlation of an expression module with overall survival (adjusted hazard ratio or HR=1.71; 95% CI=1.06-2.74 in discovery set; adjusted HR=1.26; 95% CI=1.08-1.49 in validation set 1). This expression module contained genes enriched in the biological process of the cell cycle. Interestingly, the cell cycle gene module and overall survival association were also significant in normal lung tissues (adjusted HR=1.91; 95% CI, 1.32-2.75). From these survival-related modules, we further defined three hub genes (UBE2C, TPX2, and MELK) whose expression-based risk indices were more strongly associated with poor 5-year survival (HR=3.85, 95% CI=1.34-11.05 in discovery set; HR=1.72, 95% CI=1.21-2.46 in validation set 1; and HR=3.35, 95% CI=1.08-10.04 in normal lung set). The 3-gene prognostic result was further validated using 92 adenocarcinoma tumor samples (validation set 2); patients with a high-risk gene signature have a 1.52-fold increased risk (95% CI, 1.02-2.24) of death than patients with a low-risk gene signature. These results suggest that a network-based approach may facilitate discovery of key genes that are closely linked to survival in patients with lung adenocarcinoma.

Gene coexpression networks reveal key drivers of phenotypic divergence in lake whitefish

A functional understanding of processes involved in adaptive divergence is one of the awaiting opportunities afforded by high-throughput transcriptomic technologies. Functional analysis of coexpressed genes has succeeded in the biomedical field in identifying key drivers of disease pathways. However, in ecology and evolutionary biology, functional interpretation of transcriptomic data is still limited. Here, we used Weighted Gene Co-Expression Network Analysis (WGCNA) to identify modules of coexpressed genes in muscle and brain tissue of a lake whitefish backcross progeny. Modules were connected to gradients of known adaptive traits involved in the ecological speciation process between benthic and limnetic ecotypes. Key drivers, that is, hub genes of functional modules related to reproduction, growth, and behavior were identified, and module preservation was assessed in natural populations. Using this approach, we identified modules of coexpressed genes involved in phenotypic divergence and their key drivers, and further identified a module part specifically rewired in the backcross progeny. Functional analysis of transcriptomic data can significantly contribute to the understanding of the mechanisms underlying ecological speciation. Our findings point to bone morphogenetic protein and calcium signaling as common pathways involved in coordinated evolution of trophic behavior, trophic morphology (gill rakers), and reproduction. Results also point to pathways implicating hemoglobins and constitutive stress response (HSP70) governing growth in lake whitefish.

An inferential framework for biological network hypothesis tests

BACKGROUND

Networks are ubiquitous in modern cell biology and physiology. A large literature exists for inferring/proposing biological pathways/networks using statistical or machine learning algorithms. Despite these advances a formal testing procedure for analyzing network-level observations is in need of further development. Comparing the behaviour of a pharmacologically altered pathway to its canonical form is an example of a salient one-sample comparison. Locating which pathways differentiate disease from no-disease phenotype may be recast as a two-sample network inference problem.

RESULTS

We outline an inferential method for performing one- and two-sample hypothesis tests where the sampling unit is a network and the hypotheses are stated via network model(s). We propose a dissimilarity measure that incorporates nearby neighbour information to contrast one or more networks in a statistical test. We demonstrate and explore the utility of our approach with both simulated and microarray data; random graphs and weighted (partial) correlation networks are used to form network models. Using both a well-known diabetes dataset and an ovarian cancer dataset, the methods outlined here could better elucidate co-regulation changes for one or more pathways between two clinically relevant phenotypes.

CONCLUSIONS

Formal hypothesis tests for gene- or protein-based networks are a logical progression from existing gene-based and gene-set tests for differential expression. Commensurate with the growing appreciation and development of systems biology, the dissimilarity-based testing methods presented here may allow us to improve our understanding of pathways and other complex regulatory systems. The benefit of our method was illustrated under select scenarios.

Cluster and propensity based approximation of a network

Background

The models in this article generalize current models for both correlation networks and multigraph networks. Correlation networks are widely applied in genomics research. In contrast to general networks, it is straightforward to test the statistical significance of an edge in a correlation network. It is also easy to decompose the underlying correlation matrix and generate informative network statistics such as the module eigenvector. However, correlation networks only capture the connections between numeric variables. An open question is whether one can find suitable decompositions of the similarity measures employed in constructing general networks. Multigraph networks are attractive because they support likelihood based inference. Unfortunately, it is unclear how to adjust current statistical methods to detect the clusters inherent in many data sets.

Results

Here we present an intuitive and parsimonious parametrization of a general similarity measure such as a network adjacency matrix. The cluster and propensity based approximation (CPBA) of a network not only generalizes correlation network methods but also multigraph methods. In particular, it gives rise to a novel and more realistic multigraph model that accounts for clustering and provides likelihood based tests for assessing the significance of an edge after controlling for clustering. We present a novel Majorization-Minimization (MM) algorithm for estimating the parameters of the CPBA. To illustrate the practical utility of the CPBA of a network, we apply it to gene expression data and to a bi-partite network model for diseases and disease genes from the Online Mendelian Inheritance in Man (OMIM).

Conclusions

The CPBA of a network is theoretically appealing since a) it generalizes correlation and multigraph network methods, b) it improves likelihood based significance tests for edge counts, c) it directly models higher-order relationships between clusters, and d) it suggests novel clustering algorithms. The CPBA of a network is implemented in Fortran 95 and bundled in the freely available R package PropClust.

Dissection of regulatory networks that are altered in disease via differential co-expression

Comparing the gene-expression profiles of sick and healthy individuals can help in understanding disease. Such differential expression analysis is a well-established way to find gene sets whose expression is altered in the disease. Recent approaches to gene-expression analysis go a step further and seek differential co-expression patterns, wherein the level of co-expression of a set of genes differs markedly between disease and control samples. Such patterns can arise from a disease-related change in the regulatory mechanism governing that set of genes, and pinpoint dysfunctional regulatory networks.

Here we present DICER, a new method for detecting differentially co-expressed gene sets using a novel probabilistic score for differential correlation. DICER goes beyond standard differential co-expression and detects pairs of modules showing differential co-expression. The expression profiles of genes within each module of the pair are correlated across all samples. The correlation between the two modules, however, differs markedly between the disease and normal samples.

We show that DICER outperforms the state of the art in terms of significance and interpretability of the detected gene sets. Moreover, the gene sets discovered by DICER manifest regulation by disease-specific microRNA families. In a case study on Alzheimer's disease, DICER dissected biological processes and protein complexes into functional subunits that are differentially co-expressed, thereby revealing inner structures in disease regulatory networks.

The most fundamental and popular gene-expression experiments measure genome-wide transcription levels in two populations: perturbed and wild type, or cases and controls. The genes that show significantly different expression between the two populations (the differentially expressed genes) are useful for understanding the biology underlying the phenotype difference, and can sometimes also serve as biomarkers for classification. In contrast, genes that have similar expression to each other across all profiles (co-expressed genes) can yield clues about the functional commonality of the two populations. Differential co-expression has recently been proposed as a way to combine the benefits of these two approaches: it seeks gene groups that are co-expressed in one phenotype much more than in the other. Here we develop a new method for detecting differential co-expression and test it on case-control expression profiles of several diseases. Our algorithm improves upon the state of the art in the strength of the detected patterns and in agreement with current biological knowledge. We show that our method can predict gene regulators that are associated with the disease of interest and demonstrate that it can dissect known biological pathways into subcomponents that are not detected using standard analyses.

Modeling host genetic regulation of influenza pathogenesis in the collaborative cross

Genetic variation contributes to host responses and outcomes following infection by influenza A virus or other viral infections. Yet narrow windows of disease symptoms and confounding environmental factors have made it difficult to identify polymorphic genes that contribute to differential disease outcomes in human populations. Therefore, to control for these confounding environmental variables in a system that models the levels of genetic diversity found in outbred populations such as humans, we used incipient lines of the highly genetically diverse Collaborative Cross (CC) recombinant inbred (RI) panel (the pre-CC population) to study how genetic variation impacts influenza associated disease across a genetically diverse population. A wide range of variation in influenza disease related phenotypes including virus replication, virus-induced inflammation, and weight loss was observed. Many of the disease associated phenotypes were correlated, with viral replication and virus-induced inflammation being predictors of virus-induced weight loss. Despite these correlations, pre-CC mice with unique and novel disease phenotype combinations were observed. We also identified sets of transcripts (modules) that were correlated with aspects of disease. In order to identify how host genetic polymorphisms contribute to the observed variation in disease, we conducted quantitative trait loci (QTL) mapping. We identified several QTL contributing to specific aspects of the host response including virus-induced weight loss, titer, pulmonary edema, neutrophil recruitment to the airways, and transcriptional expression. Existing whole-genome sequence data was applied to identify high priority candidate genes within QTL regions. A key host response QTL was located at the site of the known anti-influenza Mx1 gene. We sequenced the coding regions of Mx1 in the eight CC founder strains, and identified a novel Mx1 allele that showed reduced ability to inhibit viral replication, while maintaining protection from weight loss.

Host responses to an infectious agent are highly variable across the human population, however, it is not entirely clear how various factors such as pathogen dose, demography, environment and host genetic polymorphisms contribute to variable host responses and infectious outcomes. In this study, a new in vivo experimental model was used that recapitulates many of the genetic characteristics of an outbred population, such as humans. By controlling viral dose, environment and demographic variables, we were able to focus on the role that host genetic variation plays in influenza virus infection. Both the range of disease phenotypes and the combinations of sets of disease phenotypes at 4 days post infection across this population exhibited a large amount of diversity, reminiscent of the variation seen across the human population. Multiple host genome regions were identified that contributed to different aspects of the host response to influenza infection. Taken together, these results emphasize the critical role of host genetics in the response to infectious diseases. Given the breadth of host responses seen within this population, several new models for unique host responses to infection were identified.

Systems analysis of human brain gene expression: mechanisms for HIV-associated neurocognitive impairment and common pathways with Alzheimer's disease

Background

Human Immunodeficiency Virus-1 (HIV) infection frequently results in neurocognitive impairment. While the cause remains unclear, recent gene expression studies have identified genes whose transcription is dysregulated in individuals with HIV-association neurocognitive disorder (HAND). However, the methods for interpretation of such data have lagged behind the technical advances allowing the decoding genetic material. Here, we employ systems biology methods novel to the field of NeuroAIDS to further interrogate extant transcriptome data derived from brains of HIV + patients in order to further elucidate the neuropathogenesis of HAND. Additionally, we compare these data to those derived from brains of individuals with Alzheimer’s disease (AD) in order to identify common pathways of neuropathogenesis.

Methods

In Study 1, using data from three brain regions in 6 HIV-seronegative and 15 HIV + cases, we first employed weighted gene co-expression network analysis (WGCNA) to further explore transcriptome networks specific to HAND with HIV-encephalitis (HIVE) and HAND without HIVE. We then used a symptomatic approach, employing standard expression analysis and WGCNA to identify networks associated with neurocognitive impairment (NCI), regardless of HIVE or HAND diagnosis. Finally, we examined the association between the CNS penetration effectiveness (CPE) of antiretroviral regimens and brain transcriptome. In Study 2, we identified common gene networks associated with NCI in both HIV and AD by correlating gene expression with pre-mortem neurocognitive functioning.

Results

Study 1: WGCNA largely corroborated findings from standard differential gene expression analyses, but also identified possible meta-networks composed of multiple gene ontology categories and oligodendrocyte dysfunction. Differential expression analysis identified hub genes highly correlated with NCI, including genes implicated in gliosis, inflammation, and dopaminergic tone. Enrichment analysis identified gene ontology categories that varied across the three brain regions, the most notable being downregulation of genes involved in mitochondrial functioning. Finally, WGCNA identified dysregulated networks associated with NCI, including oligodendrocyte and mitochondrial functioning. Study 2: Common gene networks dysregulated in relation to NCI in AD and HIV included mitochondrial genes, whereas upregulation of various cancer-related genes was found.

Conclusions

While under-powered, this study identified possible biologically-relevant networks correlated with NCI in HIV, and common networks shared with AD, opening new avenues for inquiry in the investigation of HAND neuropathogenesis. These results suggest that further interrogation of existing transcriptome data using systems biology methods can yield important information.