A computational evaluation of over-representation of regulatory motifs in the promoter regions of differentially expressed genes

MORA and EnsembleTFpredictor: An ensemble approach to reveal functional transcription factor regulatory networks

MOTIVATION

Our study aimed to identify biologically relevant transcription factors (TFs) that control the expression of a set of co-expressed or co-regulated genes.

RESULTS

We developed a fully automated pipeline, Motif Over Representation Analysis (MORA), to detect enrichment of known TF binding motifs in any query sequences. MORA performed better than or comparable to five other TF-prediction tools as evaluated using hundreds of differentially expressed gene sets and ChIP-seq datasets derived from known TFs. Additionally, we developed EnsembleTFpredictor to harness the power of multiple TF-prediction tools to provide a list of functional TFs ranked by prediction confidence. When applied to the test datasets, EnsembleTFpredictor not only identified the target TF but also revealed many TFs known to cooperate with the target TF in the corresponding biological systems. MORA and EnsembleTFpredictor have been used in two publications, demonstrating their power in guiding experimental design and in revealing novel biological insights.

FL-DTD: an integrated pipeline to predict the drug interacting targets by feedback loop-based network analysis

Drug target discovery is an essential step to reveal the mechanism of action (MoA) underlying drug therapeutic effects and/or side effects. Most of the approaches are usually labor-intensive while unable to identify the tissue-specific interacting targets, especially the targets with weaker drug binding affinity. In this work, we proposed an integrated pipeline, FL-DTD, to predict the drug interacting targets of novel compounds in a tissue-specific manner. This method was built based on a hypothesis that cells under a status of homeostasis would take responses to drug perturbation by activating feedback loops. Therefore, the drug interacting targets can be predicted by analyzing the network responses after drug perturbation. We evaluated this method using the expression data of estrogen stimulation, gene manipulation and drug perturbation and validated its good performance to identify the annotated drug targets. Using STAT3 as a target protein, we applied this method to drug perturbation data of 500 natural compounds and predicted five compounds with STAT3 interacting activities. Experimental assay validated the STAT3-interacting activities of four compounds. Overall, our evaluation suggests that FL-DTD predicts the drug interacting targets with good accuracy and can be used for drug target discovery.

Dissection of the module network implementation "LemonTree": enhancements towards applications in metagenomics and translation in autoimmune maladies

Under the current deluge of omics, module networks distinctively emerge as methods capable of not only identifying inherently coherent groups (modules), thus reducing dimensionality, but also hypothesizing cause-effect relationships between modules and their regulators. Module networks were first designed in the transcriptomic era and further exploited in the multi-omic context to assess (for example) miRNA regulation of gene expression. Despite a number of available implementations, expansion of module networks to other omics is constrained by a limited characterization of the solutions' (modules plus regulators) accuracy and stability - an immediate need for the better characterization of molecular biology complexity in silico. We hence carefully assessed for LemonTree - a popular and open source module network implementation - the dependency of the software performances (sensitivity, specificity, false discovery rate, solutions' stability) on the input parameters and on the data quality (sample size, expression noise) based on synthetic and real data. In the process, we uncovered and fixed an issue in the code for the regulator assignment procedure. We concluded this evaluation with a table of recommended parameter settings. Finally, we applied these recommended settings to gut-intestinal metagenomic data from rheumatoid arthritis patients, to characterize the evolution of the gut-intestinal microbiome under different pharmaceutical regimens (methotrexate and prednisone) and we inferred innovative clinical recommendations with therapeutic potential, based on the computed module network.

A Systematic Investigation into Aging Related Genes in Brain and Their Relationship with Alzheimer's Disease

Aging, as a complex biological process, is accompanied by the accumulation of functional loses at different levels, which makes age to be the biggest risk factor to many neurological diseases. Even following decades of investigation, the process of aging is still far from being fully understood, especially at a systematic level. In this study, we identified aging related genes in brain by collecting the ones with sustained and consistent gene expression or DNA methylation changes in the aging process. Functional analysis with Gene Ontology to these genes suggested transcriptional regulators to be the most affected genes in the aging process. Transcription regulation analysis found some transcription factors, especially Specificity Protein 1 (SP1), to play important roles in regulating aging related gene expression. Module-based functional analysis indicated these genes to be associated with many well-known aging related pathways, supporting the validity of our approach to select aging related genes. Finally, we investigated the roles of aging related genes on Alzheimer's Disease (AD). We found that aging and AD related genes both involved some common pathways, which provided a possible explanation why aging made the brain more vulnerable to Alzheimer's Disease.

Condition-specific target prediction from motifs and expression

MOTIVATION

It is commonplace to predict targets of transcription factors (TFs) by sequence matching with their binding motifs. However, this ignores the particular condition of the cells. Gene expression data can provide condition-specific information, as is, e.g. exploited in Motif Enrichment Analysis.

RESULTS

Here, we introduce a novel tool named condition-specific target prediction (CSTP) to predict condition-specific targets for TFs from expression data measured by either microarray or RNA-seq. Based on the philosophy of guilt by association, CSTP infers the regulators of each studied gene by recovering the regulators of its co-expressed genes. In contrast to the currently used methods, CSTP does not insist on binding sites of TFs in the promoter of the target genes. CSTP was applied to three independent biological processes for evaluation purposes. By analyzing the predictions for the same TF in three biological processes, we confirm that predictions with CSTP are condition-specific. Predictions were further compared with true TF binding sites as determined by ChIP-seq/chip. We find that CSTP predictions overlap with true binding sites to a degree comparable with motif-based predictions, although the two target sets do not coincide.

AVAILABILITY AND IMPLEMENTATION

CSTP is available via a web-based interface at http://cstp.molgen.mpg.de.

oPOSSUM-3: advanced analysis of regulatory motif over-representation across genes or ChIP-Seq datasets

oPOSSUM-3 is a web-accessible software system for identification of over-represented transcription factor binding sites (TFBS) and TFBS families in either DNA sequences of co-expressed genes or sequences generated from high-throughput methods, such as ChIP-Seq. Validation of the system with known sets of co-regulated genes and published ChIP-Seq data demonstrates the capacity for oPOSSUM-3 to identify mediating transcription factors (TF) for co-regulated genes or co-recovered sequences. oPOSSUM-3 is available at http://opossum.cisreg.ca.

The gsdf gene locus harbors evolutionary conserved and clustered genes preferentially expressed in fish previtellogenic oocytes

The gonadal soma-derived factor (GSDF) belongs to the transforming growth factor-β superfamily and is conserved in teleostean fish species. Gsdf is specifically expressed in the gonads, and gene expression is restricted to the granulosa and Sertoli cells in trout and medaka. The gsdf gene expression is correlated to early testis differentiation in medaka and was shown to stimulate primordial germ cell and spermatogonia proliferation in trout. In the present study, we show that the gsdf gene localizes to a syntenic chromosomal fragment conserved among vertebrates although no gsdf-related gene is detected on the corresponding genomic region in tetrapods. We demonstrate using quantitative RT-PCR that most of the genes localized in the synteny are specifically expressed in medaka gonads. Gsdf is the only gene of the synteny with a much higher expression in the testis compared to the ovary. In contrast, gene expression pattern analysis of the gsdf surrounding genes (nup54, aff1, klhl8, sdad1, and ptpn13) indicates that these genes are preferentially expressed in the female gonads. The tissue distribution of these genes is highly similar in medaka and zebrafish, two teleostean species that have diverged more than 110 million years ago. The cellular localization of these genes was determined in medaka gonads using the whole-mount in situ hybridization technique. We confirm that gsdf gene expression is restricted to Sertoli and granulosa cells in contact with the premeiotic and meiotic cells. The nup54 gene is expressed in spermatocytes and previtellogenic oocytes. Transcripts corresponding to the ovary-specific genes (aff1, klhl8, and sdad1) are detected only in previtellogenic oocytes. No expression was detected in the gonocytes in 10 dpf embryos. In conclusion, we show that the gsdf gene localizes to a syntenic chromosomal fragment harboring evolutionary conserved genes in vertebrates. These genes are preferentially expressed in previtelloogenic oocytes, and thus, they display a different cellular localization compared to that of the gsdf gene indicating that the later gene is not co-regulated. Interestingly, our study identifies new clustered genes that are specifically expressed in previtellogenic oocytes (nup54, aff1, klhl8, sdad1).