Reconstructing differentially co-expressed gene modules and regulatory networks of soybean cells

Machine learning and bioinformatics-based insights into the potential targets of saponins in Paris polyphylla smith against non-small cell lung cancer

Background: Lung cancer has the highest mortality rate among cancers worldwide, and non-small cell lung cancer (NSCLC) is the major lethal factor. Saponins in Paris polyphylla smith exhibit antitumor activity against non-small cell lung cancer, but their targets are not fully understood.

Methods: In this study, we used differential gene analysis, lasso regression analysis and support vector machine recursive feature elimination (SVM-RFE) to screen potential key genes for NSCLC by using relevant datasets from the GEO database. The accuracy of the signature genes was verified by using ROC curves and gene expression values. Screening of potential active ingredients for the treatment of NSCLC by molecular docking of the reported active ingredients of saponins in Paris polyphylla Smith with the screened signature genes. The activity of the screened components and their effects on key genes expression were further validated by CCK-8, flow cytometry (apoptosis and cycling) and qPCR.

Results: 204 differential genes and two key genes (RHEBL1, RNPC3) stood out in the bioinformatics analysis. Overall survival (OS), First-progression survival (FP) and post-progression survival (PPS) analysis revealed that low expression of RHEBL1 and high expression of RNPC3 indicated good prognosis. In addition, Polyphyllin VI(PPVI) and Protodioscin (Prot) effectively inhibited the proliferation of non-small cell lung cancer cell line with IC50 of 4.46 μM ± 0.69 μM and 8.09 μM ± 0.67μM, respectively. The number of apoptotic cells increased significantly with increasing concentrations of PPVI and Prot. Prot induces G1/G0 phase cell cycle arrest and PPVI induces G2/M phase cell cycle arrest. After PPVI and Prot acted on this cell line for 48 h, the expression of RHEBL1 and RNPC3 was found to be consistent with the results of bioinformatics analysis.

Conclusion: This study identified two potential key genes (RHEBL1 and RNPC3) in NSCLC. Additionally, PPVI and Prot may act on RHEBL1 and RNPC3 to affect NSCLC. Our findings provide a reference for clinical treatment of NSCLC.

Role of hemoglobin alpha and hemoglobin beta in non-small-cell lung cancer based on bioinformatics analysis

The differentially expressed genes (DEGs) were identified and screened differentially in non-small-cell lung cancer (NSCLC) using information from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus databases, and the correlation of DEGs in protein interaction, function, and pathway enrichment were analyzed to search for new biomarkers and potential therapeutic targets for NSCLC. Protein-protein interaction network (PPI) analysis showed that CDK1 and GNGT1 were the most significantly upregulated hub nodes, while FPR2 was the most significantly downregulated. Gene Ontology enrichment analysis showed that upregulated DEGs were significantly enriched in protein heterodimerization activity and other functions, while downregulated DEGs were enriched in functions such as heparin-binding. Kyoto Encyclopedia of Genes and Genomes pathway analysis showed that upregulation of DEGs were significantly associated with neuroactive ligand-receptor interaction pathways, while downregulation of DEGs were significantly associated with malaria pathways. According to the analysis results, we identified hemoglobin alpha (HBA) and hemoglobin beta (HBB) as the genes of interest for further study. Through tissue level and cell level experiments, we found that the expressions of HBA and HBB in NSCLC tissues were significantly lower than those in paracancerous tissues, and downregulation of HBA and HBB could significantly affect the proliferation ability of NSCLC cells. In addition, we also found that changes in HBA and HBB may affect NSCLC cells through the p38/MAPK pathway and JNK pathway, and ultimately affect the occurrence and development of NSCLC.

GNET2: an R package for constructing gene regulatory networks from transcriptomic data

MOTIVATION

The Gene Network Estimation Tool (GNET) is designed to build gene regulatory networks (GRNs) from transcriptomic gene expression data with a probabilistic graphical model. The data preprocessing, model construction and visualization modules of the original GNET software were developed on different programming platforms, which were inconvenient for users to deploy and use.

RESULTS

Here, we present GNET2, an improved implementation of GNET as an integrated R package. GNET2 provides more flexibility for parameter initialization and regulatory module construction based on the core iterative modeling process of the original algorithm. The data exchange interface of GNET2 is handled within an R session automatically. Given the growing demand for regulatory network reconstruction from transcriptomic data, GNET2 offers a convenient option for GRN inference on large datasets.

AVAILABILITY AND IMPLEMENTATION

The source code of GNET2 is available at https://github.com/jianlin-cheng/GNET2.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Identification of potential diagnostic and prognostic biomarkers for LUAD based on TCGA and GEO databases

Accumulating evidence has demonstrated that gene alterations play a crucial role in LUAD development, progression, and prognosis. The present study aimed to identify the hub genes associated with LUAD. In the present study, we used TCGA database to screen the hub genes. Then, we validated the results by GEO datasets. Finally, we used cBioPortal, UALCAN, qRT-PCR, HPA database, TCGA database, and Kaplan–Meier plotter database to estimate the gene mutation, gene transcription, protein expression, clinical features of hub genes in patients with LUAD. A total of 5930 DEGs were screened out in TCGA database. Enrichment analysis revealed that DEGs were involved in the transcriptional misregulation in cancer, viral carcinogenesis, cAMP signaling pathway, calcium signaling pathway, and ECM–receptor interaction. The combining results of MCODE and CytoHubba showed that ADCY8, ADRB2, CALCA, GCG, GNGT1, and NPSR1 were hub genes. Then, we verified the above results by GSE118370, GSE136043, and GSE140797 datasets. Compared with normal lung tissues, the expression levels of ADCY8 and ADRB2 were lower in LUAD tissues, but the expression levels of CALCA, GCG, GNGT1, and NPSR1 were higher. In the prognosis analyses, the low expression of ADCY8 and ADRB2 and the high expression of CALCA, GCG, GNGT1, and NPSR1 were correlated with poor OS and poor PFS. The significant differences in the relationship of the expression of 6 hub genes and clinical features were observed. In conclusion, 6 hub genes will not only contribute to elucidating the pathogenesis of LUAD and may be potential therapeutic targets for LUAD.

The renoprotective effects of soy protein in the aging rat kidney

Aging is a risk factor for chronic kidney disease (CKD) and is itself associated with alterations in renal structure and function. There are no specific interventions to attenuate age-dependent renal dysfunction and the mechanism(s) responsible for these deficits have not been fully elucidated. In this study, male Fischer 344 rats, which develop age-dependent nephropathy, were feed a casein- or soy protein diet beginning at 16 mon (late life intervention) and renal structure and function was assessed at 20 mon. The soy diet did not significantly affect body weight, but was renoprotective as assessed by decreased proteinuria, increased glomerular filtration rate (GFR) and decreased urinary kidney injury molecule-1 (Kim-1). Renal fibrosis, as assessed by hydroxyproline content, was decreased by the soy diet, as were several indicators of inflammation. RNA sequencing identified several candidates for the renoprotective effects of soy, including decreased expression of Twist2, a basic helix-loop-helix transcription factor that network analysis suggest may regulate the expression of several genes associated with renal dysfunction. Twist2 expression is upregulated in the aging kidney and the unilateral ureteral obstruction of fibrosis; the expression is limited to distal tubules of mice. Taken together, these data demonstrate the renoprotective potential of soy protein, putatively by reducing inflammation and fibrosis, and identify Twist2 as a novel mediator of renal dysfunction that is targeted by soy.

Learning Differential Module Networks Across Multiple Experimental Conditions

Module network inference is a statistical method to reconstruct gene regulatory networks, which uses probabilistic graphical models to learn modules of coregulated genes and their upstream regulatory programs from genome-wide gene expression and other omics data. Here, we review the basic theory of module network inference, present protocols for common gene regulatory network reconstruction scenarios based on the Lemon-Tree software, and show, using human gene expression data, how the software can also be applied to learn differential module networks across multiple experimental conditions.

Genome-Wide Transcriptomic Analysis Reveals Insights into the Response to Citrus bark cracking viroid (CBCVd) in Hop (Humulus lupulus L.)

Viroids are smallest known pathogen that consist of non-capsidated, single-stranded non-coding RNA replicons and they exploits host factors for their replication and propagation. The severe stunting disease caused by Citrus bark cracking viroid (CBCVd) is a serious threat, which spreads rapidly within hop gardens. In this study, we employed comprehensive transcriptome analyses to dissect host-viroid interactions and identify gene expression changes that are associated with disease development in hop. Our analysis revealed that CBCVd-infection resulted in the massive modulation of activity of over 2000 genes. Expression of genes associated with plant immune responses (protein kinase and mitogen-activated protein kinase), hypersensitive responses, phytohormone signaling pathways, photosynthesis, pigment metabolism, protein metabolism, sugar metabolism, and modification, and others were altered, which could be attributed to systemic symptom development upon CBCVd-infection in hop. In addition, genes encoding RNA-dependent RNA polymerase, pathogenesis-related protein, chitinase, as well as those related to basal defense responses were up-regulated. The expression levels of several genes identified from RNA sequencing analysis were confirmed by qRT-PCR. Our systematic comprehensive CBCVd-responsive transcriptome analysis provides a better understanding and insights into complex viroid-hop plant interaction. This information will assist further in the development of future measures for the prevention of CBCVd spread in hop fields.

Hub genes and key pathways of non-small lung cancer identified using bioinformatics

Non-small cell lung cancer (NSCLC) is the most common type of lung cancer, accounting for ~80% of all lung cancer cases. The aim of the present study was to identify key genes and pathways in NSCLC, in order to improve understanding of the mechanism of lung cancer. The GSE33532 gene expression dataset, containing 20 normal and 80 NSCLC samples, was used. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to obtain the enrichment data of differently expressed genes (DEGs). Disease modules within NSCLC were constructed by Cytoscape, using protein-protein interaction (PPI) from the Search Tool for the Retrieval of Interacting Genes database. In addition, the Kaplan Meier plotter KMplot was used to assess the top hub genes in the PPI network. As a result, 1,795 genes were identified in NSCLC; 729 were upregulated and 1,066 were downregulated. The results of the GO analysis indicated that the upregulated DEGs were significantly enriched in 'biological processes' (BP), including 'cell cycle and nuclear division'; the downregulated DEGs were also significantly enriched in BP, including 'response to wounding', 'anatomical structure morphogenesis' and 'response to stimulus'. Upregulated DEGs were also enriched in 'cell cycle', 'DNA replication' and the 'tumor protein 53 signaling pathway', while the downregulated DEGs were also enriched in 'complement and coagulation cascades', 'malaria' and 'cell adhesion molecules'. The top 9 hub genes were cyclin-dependent kinase 9 (CDK1), polo-like kinase 1, aurora kinase B, cell division cycle 20, baculoviral initiator of apoptosis repeat containing 5, mitotic checkpoint serine/threonine kinase B, proliferating cell nuclear antigen (PCNA), centromere protein A and MAD2 mitotic arrest deficient-like 1, and the KMplot results revealed that the high expression levels of these genes resulted in significantly low survival rates, compared with low expression samples (P<0.05), with the exception of PCNA and CDK1. In the pathway crosstalk analysis, 26 nodes and 41 interactions were divided into two groups: One module of the two groups primarily included 'metabolism of amino acid' and the other primarily contained 'tumor necrosis signaling' pathways. In conclusion, the present study assisted in improving the understanding of the molecular mechanisms underlying NSCLC development, and the results may help the understanding of the biological mechanism of NSCLC.

Genetic insights into migraine and glutamate: a protagonist driving the headache

Migraine is a complex polygenic disorder that continues to be a great source of morbidity in the developed world with a prevalence of 12% in the Caucasian population. Genetic and pharmacological studies have implicated the glutamate pathway in migraine pathophysiology. Glutamate profoundly impacts brain circuits that regulate core symptom domains in a range of neuropsychiatric conditions and thus remains a "hot" target for drug discovery. Glutamate has been implicated in cortical spreading depression (CSD), the phenomenon responsible for migraine with aura and in animal models carrying FHM mutations. Genotyping case-control studies have shown an association between glutamate receptor genes, namely, GRIA1 and GRIA3 with migraine with indirect supporting evidence from GWAS. New evidence localizes PRRT2 at glutamatergic synapses and shows it affects glutamate signalling and glutamate receptor activity via interactions with GRIA1. Glutamate-system defects have also been recently implicated in a novel FHM2 ATP1A2 disease-mutation mouse model. Adding to the growing evidence neurophysiological findings support a role for glutamate in cortical excitability. In addition to the existence of multiple genes to choreograph the functions of fast-signalling glutamatergic neurons, glutamate receptor diversity and regulation is further increased by the post-translational mechanisms of RNA editing and miRNAs. Ongoing genetic studies, GWAS and meta-analysis implicate neurogenic mechanisms in migraine pathology and the first genome-wide associated locus for migraine on chromosome X. Finally, in addition to glutamate modulating therapies, the kynurenine pathway has emerged as a candidate for involvement in migraine pathophysiology. In this review we discuss recent genetic evidence and glutamate modulating therapies that bear on the hypothesis that a glutamatergic mechanism may be involved in migraine susceptibility.

Identification of Gene Modules Associated with Low Temperatures Response in Bambara Groundnut by Network-Based Analysis

Bambara groundnut (Vigna subterranea (L.) Verdc.) is an African legume and is a promising underutilized crop with good seed nutritional values. Low temperature stress in a number of African countries at night, such as Botswana, can effect the growth and development of bambara groundnut, leading to losses in potential crop yield. Therefore, in this study we developed a computational pipeline to identify and analyze the genes and gene modules associated with low temperature stress responses in bambara groundnut using the cross-species microarray technique (as bambara groundnut has no microarray chip) coupled with network-based analysis. Analyses of the bambara groundnut transcriptome using cross-species gene expression data resulted in the identification of 375 and 659 differentially expressed genes (p<0.01) under the sub-optimal (23°C) and very sub-optimal (18°C) temperatures, respectively, of which 110 genes are commonly shared between the two stress conditions. The construction of a Highest Reciprocal Rank-based gene co-expression network, followed by its partition using a Heuristic Cluster Chiseling Algorithm resulted in 6 and 7 gene modules in sub-optimal and very sub-optimal temperature stresses being identified, respectively. Modules of sub-optimal temperature stress are principally enriched with carbohydrate and lipid metabolic processes, while most of the modules of very sub-optimal temperature stress are significantly enriched with responses to stimuli and various metabolic processes. Several transcription factors (from MYB, NAC, WRKY, WHIRLY & GATA classes) that may regulate the downstream genes involved in response to stimulus in order for the plant to withstand very sub-optimal temperature stress were highlighted. The identified gene modules could be useful in breeding for low-temperature stress tolerant bambara groundnut varieties.

A gene expression microarray for Nicotiana benthamiana based on de novo transcriptome sequence assembly

BACKGROUND

Nicotiana benthamiana has been widely used in laboratories around the world for studying plant-pathogen interactions and posttranscriptional gene expression silencing. Yet the exploration of its transcriptome has lagged behind due to the lack of both adequate sequence information and genome-wide analysis tools, such as DNA microarrays. Despite the increasing use of high-throughput sequencing technologies, the DNA microarrays still remain a popular gene expression tool, because they are cheaper and less demanding regarding bioinformatics skills and computational effort.

RESULTS

We designed a gene expression microarray with 103,747 60-mer probes, based on two recently published versions of N. benthamiana transcriptome (v.3 and v.5). Both versions were reconstructed from RNA-Seq data of non-strand-specific pooled-tissue libraries, so we defined the sense strand of the contigs prior to designing the probe. To accomplish this, we combined a homology search against Arabidopsis thaliana proteins and hybridization to a test 244k microarray containing pairs of probes, which represented individual contigs. We identified the sense strand in 106,684 transcriptome contigs and used this information to design an Nb-105k microarray on an Agilent eArray platform. Following hybridization of RNA samples from N. benthamiana roots and leaves we demonstrated that the new microarray had high specificity and sensitivity for detection of differentially expressed transcripts. We also showed that the data generated with the Nb-105k microarray may be used to identify incorrectly assembled contigs in the v.5 transcriptome, by detecting inconsistency in the gene expression profiles, which is indicated using multiple microarray probes that match the same v.5 primary transcripts.

CONCLUSIONS

We provided a complete design of an oligonucleotide microarray that may be applied to the research of N. benthamiana transcriptome. This, in turn, will allow the N. benthamiana research community to take full advantage of microarray capabilities for studying gene expression in this plant. Additionally, by defining the sense orientation of over 106,000 contigs, we substantially improved the functional information on the N. benthamiana transcriptome. The simple hybridization-based approach for detecting the sense orientation of computationally assembled sequences can be used for updating the transcriptomes of other non-model organisms, including cases where no significant homology to known proteins exists.

Soybean Roots Grown under Heat Stress Show Global Changes in Their Transcriptional and Proteomic Profiles

Heat stress is likely to be a key factor in the negative impact of climate change on crop production. Heat stress significantly influences the functions of roots, which provide support, water, and nutrients to other plant organs. Likewise, roots play an important role in the establishment of symbiotic associations with different microorganisms. Despite the physiological relevance of roots, few studies have examined their response to heat stress. In this study, we performed genome-wide transcriptomic and proteomic analyses on isolated root hairs, which are a single, epidermal cell type, and compared their response to stripped roots. On average, we identified 1849 and 3091 genes differentially regulated in root hairs and stripped roots, respectively, in response to heat stress. Our gene regulatory module analysis identified 10 key modules that might control the majority of the transcriptional response to heat stress. We also conducted proteomic analysis on membrane fractions isolated from root hairs and compared these responses to stripped roots. These experiments identified a variety of proteins whose expression changed within 3 h of application of heat stress. Most of these proteins were predicted to play a significant role in thermo-tolerance, as well as in chromatin remodeling and post-transcriptional regulation. The data presented represent an in-depth analysis of the heat stress response of a single cell type in soybean.

Identification of key transcription factors in caerulein-induced pancreatitis through expression profiling data

The current study aimed to isolate key transcription factors (TFs) in caerulein-induced pancreatitis, and to identify the difference between wild type and Mist1 knockout (KO) mice, in order to elucidate the contribution of Mist1 to pancreatitis. The gene profile of GSE3644 was downloaded from the Gene Expression Omnibus database then analyzed using the t-test. The isolated differentially expressed genes (DEGs) were mapped into a transcriptional regulatory network derived from the Integrated Transcription Factor Platform database and in the network, the interaction pairs involving at least one DEG were screened. Fisher’s exact test was used to analyze the functional enrichment of the target genes. A total of 1,555 and 3,057 DEGs were identified in the wild type and Mist1KO mice treated with caerulein, respectively. DEGs screened in Mist1KO mice were predominantly enriched in apoptosis, mitogen-activated protein kinase signaling and other cancer-associated pathways. A total of 188 and 51 TFs associated with pathopoiesis were isolated in Mist1KO and wild type mice, respectively. Out of the top 10 TFs (ranked by P-value), 7 TFs, including S-phase kinase-associated protein 2 (Skp2); minichromosome maintenance complex component 3 (Mcm3); cell division cycle 6 (Cdc6); cyclin B1 (Ccnb1); mutS homolog 6 (Msh6); cyclin A2 (Ccna2); and cyclin B2 (Ccnb2), were expressed in the two types of mouse. These TFs were predominantly involved in phosphorylation, DNA replication, cell division and DNA mismatch repair. In addition, specific TFs, including minichromosome maintenance complex component 7 (Mcm7); lymphoid-specific helicase (Hells); and minichromosome maintenance complex component 6 (Mcm6), that function in the unwinding of DNA were identified to participate in Mist1KO pancreatitis. The DEGs, including Cdc6, Mcm6, Msh6 and Wdr1 are closely associated with the regulation of caerulein-induced pancreatitis. Furthermore, other identified TFs were also involved in this type of regulation.

From Gigabyte to Kilobyte: A Bioinformatics Protocol for Mining Large RNA-Seq Transcriptomics Data

RNA-Seq techniques generate hundreds of millions of short RNA reads using next-generation sequencing (NGS). These RNA reads can be mapped to reference genomes to investigate changes of gene expression but improved procedures for mining large RNA-Seq datasets to extract valuable biological knowledge are needed. RNAMiner--a multi-level bioinformatics protocol and pipeline--has been developed for such datasets. It includes five steps: Mapping RNA-Seq reads to a reference genome, calculating gene expression values, identifying differentially expressed genes, predicting gene functions, and constructing gene regulatory networks. To demonstrate its utility, we applied RNAMiner to datasets generated from Human, Mouse, Arabidopsis thaliana, and Drosophila melanogaster cells, and successfully identified differentially expressed genes, clustered them into cohesive functional groups, and constructed novel gene regulatory networks. The RNAMiner web service is available at http://calla.rnet.missouri.edu/rnaminer/index.html.

Integrative multi-omics module network inference with Lemon-Tree

Module network inference is an established statistical method to reconstruct co-expression modules and their upstream regulatory programs from integrated multi-omics datasets measuring the activity levels of various cellular components across different individuals, experimental conditions or time points of a dynamic process. We have developed Lemon-Tree, an open-source, platform-independent, modular, extensible software package implementing state-of-the-art ensemble methods for module network inference. We benchmarked Lemon-Tree using large-scale tumor datasets and showed that Lemon-Tree algorithms compare favorably with state-of-the-art module network inference software. We also analyzed a large dataset of somatic copy-number alterations and gene expression levels measured in glioblastoma samples from The Cancer Genome Atlas and found that Lemon-Tree correctly identifies known glioblastoma oncogenes and tumor suppressors as master regulators in the inferred module network. Novel candidate driver genes predicted by Lemon-Tree were validated using tumor pathway and survival analyses. Lemon-Tree is available from http://lemon-tree.googlecode.com under the GNU General Public License version 2.0.

Transcriptomic analysis identifies gene networks regulated by estrogen receptor α (ERα) and ERβ that control distinct effects of different botanical estrogens

The estrogen receptors (ERs) ERα and ERβ mediate the actions of endogenous estrogens as well as those of botanical estrogens (BEs) present in plants. BEs are ingested in the diet and also widely consumed by postmenopausal women as dietary supplements, often as a substitute for the loss of endogenous estrogens at menopause. However, their activities and efficacies, and similarities and differences in gene expression programs with respect to endogenous estrogens such as estradiol (E2) are not fully understood. Because gene expression patterns underlie and control the broad physiological effects of estrogens, we have investigated and compared the gene networks that are regulated by different BEs and by E2. Our aim was to determine if the soy and licorice BEs control similar or different gene expression programs and to compare their gene regulations with that of E2. Gene expression was examined by RNA-Seq in human breast cancer (MCF7) cells treated with control vehicle, BE or E2. These cells contained three different complements of ERs, ERα only, ERα+ERβ, or ERβ only, reflecting the different ratios of these two receptors in different human breast cancers and in different estrogen target cells. Using principal component, hierarchical clustering, and gene ontology and interactome analyses, we found that BEs regulated many of the same genes as did E2. The genes regulated by each BE, however, were somewhat different from one another, with some genes being regulated uniquely by each compound. The overlap with E2 in regulated genes was greatest for the soy isoflavones genistein and S-equol, while the greatest difference from E2 in gene expression pattern was observed for the licorice root BE liquiritigenin. The gene expression pattern of each ligand depended greatly on the cell background of ERs present. Despite similarities in gene expression pattern with E2, the BEs were generally less stimulatory of genes promoting proliferation and were more pro-apoptotic in their gene regulations than E2. The distinctive patterns of gene regulation by the individual BEs and E2 may underlie differences in the activities of these soy and licorice-derived BEs in estrogen target cells containing different levels of the two ERs.

Predicting gene regulatory networks of soybean nodulation from RNA-Seq transcriptome data

BACKGROUND

High-throughput RNA sequencing (RNA-Seq) is a revolutionary technique to study the transcriptome of a cell under various conditions at a systems level. Despite the wide application of RNA-Seq techniques to generate experimental data in the last few years, few computational methods are available to analyze this huge amount of transcription data. The computational methods for constructing gene regulatory networks from RNA-Seq expression data of hundreds or even thousands of genes are particularly lacking and urgently needed.

RESULTS

We developed an automated bioinformatics method to predict gene regulatory networks from the quantitative expression values of differentially expressed genes based on RNA-Seq transcriptome data of a cell in different stages and conditions, integrating transcriptional, genomic and gene function data. We applied the method to the RNA-Seq transcriptome data generated for soybean root hair cells in three different development stages of nodulation after rhizobium infection. The method predicted a soybean nodulation-related gene regulatory network consisting of 10 regulatory modules common for all three stages, and 24, 49 and 70 modules separately for the first, second and third stage, each containing both a group of co-expressed genes and several transcription factors collaboratively controlling their expression under different conditions. 8 of 10 common regulatory modules were validated by at least two kinds of validations, such as independent DNA binding motif analysis, gene function enrichment test, and previous experimental data in the literature.

CONCLUSIONS

We developed a computational method to reliably reconstruct gene regulatory networks from RNA-Seq transcriptome data. The method can generate valuable hypotheses for interpreting biological data and designing biological experiments such as ChIP-Seq, RNA interference, and yeast two hybrid experiments.