Large-scale gene co-expression network as a source of functional annotation for cattle genes

Transcriptomic Insight into the Pollen Tube Growth of Olea europaea L. subsp. europaea Reveals Reprogramming and Pollen-Specific Genes Including New Transcription Factors

The pollen tube is a key innovation of land plants that is essential for successful fertilisation. Its development and growth have been profusely studied in model organisms, but in spite of the economic impact of olive trees, little is known regarding the genome-wide events underlying pollen hydration and growth in this species. To fill this gap, triplicate mRNA samples at 0, 1, 3, and 6 h of in vitro germination of olive cultivar Picual pollen were analysed by RNA-seq. A bioinformatics R workflow called RSeqFlow was developed contemplating the best practices described in the literature, covering from expression data filtering to differential expression and clustering, to finally propose hub genes. The resulting olive pollen transcriptome consisted of 22,418 reliable transcripts, where 5364 were differentially expressed, out of which 173 have no orthologue in plants and up to 3 of them might be pollen-specific transcription factors. Functional enrichment revealed a deep transcriptional reprogramming in mature olive pollen that is also dependent on protein stability and turnover to allow pollen tube emergence, with many hub genes related to heat shock proteins and F-box-containing proteins. Reprogramming extends to the first 3 h of growth, including processes consistent with studies performed in other plant species, such as global down-regulation of biosynthetic processes, vesicle/organelle trafficking and cytoskeleton remodelling. In the last stages, growth should be maintained from persistent transcripts. Mature pollen is equipped with transcripts to successfully cope with adverse environments, even though the in vitro growth seems to induce several stress responses. Finally, pollen-specific transcription factors were proposed as probable drivers of pollen germination in olive trees, which also shows an overall increased number of pollen-specific gene isoforms relative to other plants.

Network analysis reveals different hub genes and molecular pathways for pig in vitro fertilized early embryos and parthenogenotes

Maternal-to-zygotic transition (MZT) occurs when maternal transcripts decay and zygotic genome is activated gradually at early stage of embryo development. Previously, single cell RNA-seq (scRNA-seq) has helped us to uncover the MZT-associated mRNA dynamics of in vitro produced pig early embryos. Here, to further investigate functional modules and hub genes associated with MZT process, the weighted gene-coexpression network analysis (WGCNA) was performed on our previously generated 45 scRNA-seq datasets. For the in vitro fertilized embryo (IVF) group, 5 significant modules were identified (midnightblue/black/red and blue/brown modules, positively correlated with 1-cell (IVF1) and 8-cell (IVF8), respectively), containing genes mainly enriched in signaling pathways such as Wnt, regulation of RNA transcription, fatty acid metabolic process, poly(A) RNA binding and lysosome. For the parthenogenetically activated embryo (PA) group, 9 significant modules were identified (black/purple/red, brown/turquoise/yellow, and magenta/blue/green modules, positively correlated with MII oocytes, 1-cell (PA1), and 8-cell (PA8), respectively), mainly enriched in extracellular exosome, poly(A) RNA binding, mitochondrion, transcription factor activity. Moreover, some of identified hub genes within 3 IVF and 9 PA significant modules, including ADCY2, DHX34, KDM4A, GDF10, ABCC10, PAFAH2, HEXIM2, COQ9, DCAF11, SGK1, ESRRB etc., have been reported to play vital roles in different biological processes. Our findings provide information and resources for subsequent in-depth study on the regulation and function of MZT in pig embryos.

Gene Co-Expression Network Tools and Databases for Crop Improvement

Transcriptomics has significantly grown as a functional genomics tool for understanding the expression of biological systems. The generated transcriptomics data can be utilised to produce a gene co-expression network that is one of the essential downstream omics data analyses. To date, several gene co-expression network databases that store correlation values, expression profiles, gene names and gene descriptions have been developed. Although these resources remain scattered across the Internet, such databases complement each other and support efficient growth in the functional genomics area. This review presents the features and the most recent gene co-expression network databases in crops and summarises the present status of the tools that are widely used for constructing the gene co-expression network. The highlights of gene co-expression network databases and the tools presented here will pave the way for a robust interpretation of biologically relevant information. With this effort, the researcher would be able to explore and utilise gene co-expression network databases for crops improvement.

Cross-population selection signatures in Canchim composite beef cattle

Analyses of livestock genomes have been used to detect selection signatures, which are genomic regions associated with traits under selection leading to a change in allele frequency. The objective of the present study was to characterize selection signatures in Canchim composite beef cattle using cross-population analyses with the founder Nelore and Charolais breeds. High-density single nucleotide polymorphism genotypes were available on 395 Canchim representing the target population, along with genotypes from 809 Nelore and 897 Charolais animals representing the reference populations. Most of the selection signatures were co-located with genes whose functions agree with the expectations of the breeding programs; these genes have previously been reported to associate with meat quality, as well as reproductive traits. Identified genes were related to immunity, adaptation, morphology, as well as behavior, could give new perspectives for understanding the genetic architecture of Canchim. Some selection signatures identified genes that were recently introduced in Canchim, such as the loci related to the polled trait.

Temporal analysis of the bovine lymph node transcriptome during cattle tick (Rhipicephalus microplus) infestation

Livestock production is a fundamental source of revenue and nutrition, wherein cattle-farming constitutes one of the major agricultural industries. Vectors and vector-borne diseases constitute one of the major factors that decrease the livelihood of all farming communities, more so in resource-poor communities and developing countries. Understanding the immunological responses during tick infestation in cattle is instrumental in the development of novel and improved tick control strategies, such as vaccines. In this study, gene expression patterns were compared within the lymph nodes of three cattle breeds at different life stages of the cattle tick, Rhipicephalus microplus. For Bonsmara (5/8Bos taurus indicus × 3/8B. t. taurus) cattle specifically, some 183 genes were found to be differentially expressed within the lymph nodes during larval and adult tick feeding, relative to uninfested cattle. Overall, the data provides evidence for a transcriptional regulatory network that is activated during immature tick infestation, but is down-regulated towards basal transcriptional levels when adult ticks are feeding. Specific processes in the lymph nodes of Bonsmara cattle were found to be differentially regulated on a transcriptional level. These include: (1) Leukocyte recruitment to the lymph node via chemokines and chemotaxis, (2) Trans-endothelial and intranodal movement on the reticular network, (3) Active regulation of cellular transcription and translation in the lymph node (including leukocyte associated cellular regulatory networks) and (4) Chemokine receptors regulating the movement of cells out of the lymph node. This work provides a first transcriptome analysis of bovine lymph node responses in tick-infested cattle. Findings show a dynamic immune response to tick infestation for the Bonsmara cattle breed, and that suppression of the maturation of the cattle hosts' immunity is especially evident during the larval feeding stages.

Construction, analysis and validation of co-expression network to understand stress adaptation in Deinococcus radiodurans R1

Systems biology based approaches have been effectively utilized to mine high throughput data. In the current study, we have performed system-level analysis for Deinococcus radiodurans R1 by constructing a gene co-expression network based on several microarray datasets available in the public domain. This condition-independent network was constructed by Weighted Gene Co-expression Network Analysis (WGCNA) with 61 microarray samples from 9 different experimental conditions. We identified 13 co-expressed modules, of which, 11 showed functional enrichments of one or more pathway/s or biological process. Comparative analysis of differentially expressed genes and proteins from radiation and desiccation stress studies with our co-expressed modules revealed the association of cyan with radiation response. Interestingly, two modules viz darkgreen and tan was associated with radiation as well as desiccation stress responses. The functional analysis of these modules showed enrichment of pathways important for adaptation of radiation or desiccation stress. To decipher the regulatory roles of these stress responsive modules, we identified transcription factors (TFs) and then calculated a Biweight mid correlation between modules hub gene and the identified TFs. We obtained 7 TFs for radiation and desiccation responsive modules. The expressions of 3 TFs were validated in response to gamma radiation using qRT-PCR. Along with the TFs, selected close neighbor genes of two important TFs, viz., DR_0997 (CRP) and DR_2287 (AsnC family transcriptional regulator) in the darkgreen module were also validated. In our network, among 13 hub genes associated with 13 modules, the functionality of 5 hub genes which are annotated as hypothetical proteins (hypothetical hub genes) in D. radiodurans genome has been revealed. Overall the study provided a better insight of pathways and regulators associated with relevant DNA damaging stress response in D. radiodurans.

Common Gene Modules Identified for Chicken Adiposity by Network Construction and Comparison

Excessive fat deposition can cause chicken health problem, and affect production efficiency by causing great economic losses to the industry. However, the molecular underpinnings of the complex adiposity trait remain elusive. In the current study, we constructed and compared the gene co-expression networks on four transcriptome profiling datasets, from two chicken lines under divergent selection for abdominal fat contents, in an attempt to dissect network compositions underlying adipose tissue growth and development. After functional enrichment analysis, nine network modules important to adipogenesis were discovered to be involved in lipid metabolism, PPAR and insulin signaling pathways, and contained hub genes related to adipogenesis, cell cycle, inflammation, and protein synthesis. Moreover, after additional functional annotation and network module comparisons, common sub-modules of similar functionality for chicken fat deposition were identified for different chicken lines, apart from modules specific to each chicken line. We further validated the lysosome pathway, and found TFEB and its downstream target genes showed similar expression patterns along with chicken preadipocyte differentiation. Our findings could provide novel insights into the genetic basis of complex adiposity traits, as well as human obesity and related metabolic diseases.

FunCoup 4: new species, data, and visualization

This release of the FunCoup database (http://funcoup.sbc.su.se) is the fourth generation of one of the most comprehensive databases for genome-wide functional association networks. These functional associations are inferred via integrating various data types using a naive Bayesian algorithm and orthology based information transfer across different species. This approach provides high coverage of the included genomes as well as high quality of inferred interactions. In this update of FunCoup we introduce four new eukaryotic species: Schizosaccharomyces pombe, Plasmodium falciparum, Bos taurus, Oryza sativa and open the database to the prokaryotic domain by including networks for Escherichia coli and Bacillus subtilis. The latter allows us to also introduce a new class of functional association between genes - co-occurrence in the same operon. We also supplemented the existing classes of functional association: metabolic, signaling, complex and physical protein interaction with up-to-date information. In this release we switched to InParanoid v8 as the source of orthology and base for calculation of phylogenetic profiles. While populating all other evidence types with new data we introduce a new evidence type based on quantitative mass spectrometry data. Finally, the new JavaScript based network viewer provides the user an intuitive and responsive platform to further evaluate the results.

Improved annotation of the domestic pig genome through integration of Iso-Seq and RNA-seq data

BACKGROUND

Our understanding of the pig transcriptome is limited. RNA transcript diversity among nine tissues was assessed using poly(A) selected single-molecule long-read isoform sequencing (Iso-seq) and Illumina RNA sequencing (RNA-seq) from a single White cross-bred pig.

RESULTS

Across tissues, a total of 67,746 unique transcripts were observed, including 60.5% predicted protein-coding, 36.2% long non-coding RNA and 3.3% nonsense-mediated decay transcripts. On average, 90% of the splice junctions were supported by RNA-seq within tissue. A large proportion (80%) represented novel transcripts, mostly produced by known protein-coding genes (70%), while 17% corresponded to novel genes. On average, four transcripts per known gene (tpg) were identified; an increase over current EBI (1.9 tpg) and NCBI (2.9 tpg) annotations and closer to the number reported in human genome (4.2 tpg). Our new pig genome annotation extended more than 6000 known gene borders (5' end extension, 3' end extension, or both) compared to EBI or NCBI annotations. We validated a large proportion of these extensions by independent pig poly(A) selected 3'-RNA-seq data, or human FANTOM5 Cap Analysis of Gene Expression data. Further, we detected 10,465 novel genes (81% non-coding) not reported in current pig genome annotations. More than 80% of these novel genes had transcripts detected in > 1 tissue. In addition, more than 80% of novel intergenic genes with at least one transcript detected in liver tissue had H3K4me3 or H3K36me3 peaks mapping to their promoter and gene body, respectively, in independent liver chromatin immunoprecipitation data.

CONCLUSIONS

These validated results show significant improvement over current pig genome annotations.

Using Machine Learning to Measure Relatedness Between Genes: A Multi-Features Model

Measuring conditional relatedness between a pair of genes is a fundamental technique and still a significant challenge in computational biology. Such relatedness can be assessed by gene expression similarities while suffering high false discovery rates. Meanwhile, other types of features, e.g., prior-knowledge based similarities, is only viable for measuring global relatedness. In this paper, we propose a novel machine learning model, named Multi-Features Relatedness (MFR), for accurately measuring conditional relatedness between a pair of genes by incorporating expression similarities with prior-knowledge based similarities in an assessment criterion. MFR is used to predict gene-gene interactions extracted from the COXPRESdb, KEGG, HPRD, and TRRUST databases by the 10-fold cross validation and test verification, and to identify gene-gene interactions collected from the GeneFriends and DIP databases for further verification. The results show that MFR achieves the highest area under curve (AUC) values for identifying gene-gene interactions in the development, test, and DIP datasets. Specifically, it obtains an improvement of 1.1% on average of precision for detecting gene pairs with both high expression similarities and high prior-knowledge based similarities in all datasets, comparing to other linear models and coexpression analysis methods. Regarding cancer gene networks construction and gene function prediction, MFR also obtains the results with more biological significances and higher average prediction accuracy, than other compared models and methods. A website of the MFR model and relevant datasets can be accessed from http://bmbl.sdstate.edu/MFR.

Co-Expression Network Analysis Identifies miRNA⁻mRNA Networks Potentially Regulating Milk Traits and Blood Metabolites

MicroRNAs (miRNA) regulate mRNA networks to coordinate cellular functions. In this study, we constructed gene co-expression networks to detect miRNA modules (clusters of miRNAs with similar expression patterns) and miRNA–mRNA pairs associated with blood (triacylglyceride and nonesterified fatty acids) and milk (milk yield, fat, protein, and lactose) components and milk fatty acid traits following dietary supplementation of cows’ diets with 5% linseed oil (LSO) (n = 6 cows) or 5% safflower oil (SFO) (n = 6 cows) for 28 days. Using miRNA transcriptome data from mammary tissues of cows for co-expression network analysis, we identified three consensus modules: blue, brown, and turquoise, composed of 70, 34, and 86 miRNA members, respectively. The hub miRNAs (miRNAs with the most connections with other miRNAs) were miR-30d, miR-484 and miR-16b for blue, brown, and turquoise modules, respectively. Cell cycle arrest, and p53 signaling and transforming growth factor–beta (TGF-β) signaling pathways were the common gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways enriched for target genes of the three modules. Protein percent (p = 0.03) correlated with the turquoise module in LSO treatment while protein yield (p = 0.003) and milk yield (p = 7 × 10⁻⁰⁴) correlated with the turquoise model, protein and milk yields and lactose percent (p < 0.05) correlated with the blue module and fat percent (p = 0.04) correlated with the brown module in SFO treatment. Several fatty acids correlated (p < 0.05) with the blue (CLA:9,11) and brown (C4:0, C12:0, C22:0, C18:1n9c and CLA:10,12) modules in LSO treatment and with the turquoise (C14:0, C18:3n3 and CLA:9,11), blue (C14:0 and C23:0) and brown (C6:0, C16:0, C22:0, C22:6n3 and CLA:10,12) modules in SFO treatment. Correlation of miRNA and mRNA data from the same animals identified the following miRNA–mRNA pairs: miR-183/RHBDD2 (p = 0.003), miR-484/EIF1AD (p = 0.011) and miR-130a/SBSPON (p = 0.004) with lowest p-values for the blue, brown, and turquoise modules, respectively. Milk yield, protein yield, and protein percentage correlated (p < 0.05) with 28, 31 and 5 miRNA–mRNA pairs, respectively. Our results suggest that, the blue, brown, and turquoise modules miRNAs, hub miRNAs, miRNA–mRNA networks, cell cycle arrest GO term, p53 signaling and TGF-β signaling pathways have considerable influence on milk and blood phenotypes following dietary supplementation of dairy cows’ diets with 5% LSO or 5% SFO.

Analysis of protein-protein interaction network based on transcriptome profiling of ovine granulosa cells identifies candidate genes in cyclic recruitment of ovarian follicles

After pubertal, cohort of small antral follicles enters to gonadotrophin-sensitive development, called recruited follicles. This study was aimed to identify candidate genes in follicular cyclic recruitment via analysis of protein-protein interaction (PPI) network. Differentially expressed genes (DEGs) in ovine granulosa cells of small antral follicles between follicular and luteal phases were accumulated among gene/protein symbols of the Ensembl annotation. Following directed graphs, PTPN6 and FYN have the highest indegree and outdegree, respectively. Since, these hubs being up-regulated in ovine granulosa cells of small antral follicles during the follicular phase, it represents an accumulation of blood immune cells in follicular phase in comparison with luteal phase. By contrast, the up-regulated hubs in the luteal phase including CDK1, INSRR and TOP2A which stimulated DNA replication and proliferation of granulosa cells, they known as candidate genes of the cyclic recruitment.

Cattle infection response network and its functional modules

BACKGROUND

Weighted Gene Co-expression Network analysis, a powerful technique used to extract co-expressed gene pattern from mRNA expression data, was constructed to infer common immune strategies used by cattle in response to five different bacterial species (Escherichia coli, Mycobacterium avium, Mycobacterium bovis, Salmonella and Staphylococcus aureus) and a protozoa (Trypanosoma Congolense) using 604 publicly available gene expression microarrays from 12 cattle infection experiments.

RESULTS

A total of 14,999 transcripts that were differentially expressed (DE) in at least three different infection experiments were consolidated into 15 modules that contained between 43 and 4441 transcripts. The high number of shared DE transcripts between the different types of infections indicated that there were potentially common immune strategies used in response to these infections. The number of transcripts in the identified modules varied in response to different infections. Fourteen modules showed a strong functional enrichment for specific GO/pathway terms related to "immune system process" (71%), "metabolic process" (71%), "growth and developmental process" (64%) and "signaling pathways" (50%), which demonstrated the close interconnection between these biological pathways in response to different infections. The largest module in the network had several over-represented GO/pathway terms related to different aspects of lipid metabolism and genes in this module were down-regulated for the most part during various infections. Significant negative correlations between this module's eigengene values, three immune related modules in the network, and close interconnection between their hub genes, might indicate the potential co-regulation of these modules during different infections in bovine. In addition, the potential function of 93 genes with no functional annotation was inferred based on neighbor analysis and functional uniformity among associated genes. Several hypothetical genes were differentially expressed during experimental infections, which might indicate their important role in cattle response to different infections.

CONCLUSIONS

We identified several biological pathways involved in immune response to different infections in cattle. These findings provide rich information for experimental biologists to design experiments, interpret experimental results, and develop novel hypothesis on immune response to different infections in cattle.

Constructing a comprehensive gene co-expression based interactome in Bos taurus

Integrating genomic information into cattle breeding is an important approach to exploring genotype-phenotype relationships for complex traits related to diary and meat production. To assist with genomic-based selection, a reference map of interactome is needed to fully understand and identify the functional relevant genes. To this end, we constructed a co-expression analysis of 92 tissues and this represents the systematic exploration of gene-gene relationship in Bos taurus. By using robust WGCNA (Weighted Gene Correlation Network Analysis), we described the gene co-expression network of 5,000 protein-coding genes with majority variations in expression across 92 tissues. Further module identifications found 55 highly organized functional clusters representing diverse cellular activities. To demonstrate the re-use of our interaction for functional genomics analysis, we extracted a sub-network associated with DNA binding genes in Bos taurus. The subnetwork was enriched within regulation of transcription from RNA polymerase II promoter representing central cellular functions. In addition, we identified 28 novel linker genes associated with more than 100 DNA binding genes. Our WGCNA-based co-expression network reconstruction will be a valuable resource for exploring the molecular mechanisms of incompletely characterized proteins and for elucidating larger-scale patterns of functional modulization in the Bos taurus genome.

Discovering Condition-Specific Gene Co-Expression Patterns Using Gaussian Mixture Models: A Cancer Case Study

A gene co-expression network (GCN) describes associations between genes and points to genetic coordination of biochemical pathways. However, genetic correlations in a GCN are only detectable if they are present in the sampled conditions. With the increasing quantity of gene expression samples available in public repositories, there is greater potential for discovery of genetic correlations from a variety of biologically interesting conditions. However, even if gene correlations are present, their discovery can be masked by noise. Noise is introduced from natural variation (intrinsic and extrinsic), systematic variation (caused by sample measurement protocols and instruments), and algorithmic and statistical variation created by selection of data processing tools. A variety of published studies, approaches and methods attempt to address each of these contributions of variation to reduce noise. Here we describe an approach using Gaussian Mixture Models (GMMs) to address natural extrinsic (condition-specific) variation during network construction from mixed input conditions. To demonstrate utility, we build and analyze a condition-annotated GCN from a compendium of 2,016 mixed gene expression data sets from five tumor subtypes obtained from The Cancer Genome Atlas. Our results show that GMMs help discover tumor subtype specific gene co-expression patterns (modules) that are significantly enriched for clinical attributes.