ORTI: An Open-Access Repository of Transcriptional Interactions for Interrogating Mammalian Gene Expression Data

Abstract

Transcription factors (TFs) play a fundamental role in coordinating biological processes in response to stimuli. Consequently, we often seek to determine the key TFs and their regulated target genes (TGs) amidst gene expression data. This requires a knowledge-base of TF-TG interactions, which would enable us to determine the topology of the transcriptional network and predict novel regulatory interactions. To address this, we generated an Open-access Repository of Transcriptional Interactions, ORTI, by integrating available TF-TG interaction databases. These databases rely on different types of experimental evidence, including low-throughput assays, high-throughput screens, and bioinformatics predictions. We have subsequently categorised TF-TG interactions in ORTI according to the quality of this evidence. To demonstrate its capabilities, we applied ORTI to gene expression data and identified modulated TFs using an enrichment analysis. Combining this with pairwise TF-TG interactions enabled us to visualise temporal regulation of a transcriptional network. Additionally, ORTI enables the prediction of novel TF-TG interactions, based on how well candidate genes co-express with known TGs of the target TF. By filtering out known TF-TG interactions that are unlikely to occur within the experimental context, this analysis predicts context-specific TF-TG interactions. We show that this can be applied to experimental designs of varying complexities. In conclusion, ORTI is a rich and publicly available database of experimentally validated mammalian transcriptional interactions which is accompanied with tools that can identify and predict transcriptional interactions, serving as a useful resource for unravelling the topology of transcriptional networks.

The transcriptional response to oxidative stress is part of, but not sufficient for, insulin resistance in adipocytes

“…This has the potential to be augmented in a biologically-meaningful way using an approach previously applied to phosphoproteomics data 17 , whereby prior knowledge of kinase-substrate interactions determined the optimal clustering of genes. This could be applied to transcriptional data using known (experimentally-validated) TF-TG interactions from public repositories (e.g., ORTI database 18 ), which can then be used to identify enriched TFs within the clusters 17 . This relies on the assumption that the TGs for a single TF will be co-regulated and thus have similar expression patterns 17 , 18 .…”

“…This could be applied to transcriptional data using known (experimentally-validated) TF-TG interactions from public repositories (e.g., ORTI database 18 ), which can then be used to identify enriched TFs within the clusters 17 . This relies on the assumption that the TGs for a single TF will be co-regulated and thus have similar expression patterns 17 , 18 . Overall, this would enable time-series data to be used to generate transcriptional cascades from context-specific TF-TG interactions 18 .…”

Functional prediction of long non-coding RNAs in ovarian cancer-associated fibroblasts indicate a potential role in metastasis

“…The names of the interacting TFs were disambiguated using NCBI to obtain their official symbols provided by the HGNC (HUGO Gene Nomenclature Committee). These TFs were used to search ORTI 77 , a recently compiled repository of mammalian transcriptional interactions, to obtain their experimentally-validated (Rank 1) target genes (TGs). The compiled TF-TG interactions are sourced from a range of experimental conditions.…”

“…The compiled TF-TG interactions are sourced from a range of experimental conditions. In order to identify TFs modulated under the specific context of the study, we used ORTI application 1 77 , which performs TF enrichment analysis on a list of genes differentially expressed in CAFs versus NOFs ( p -value < 0.05 and |log2(fold-change)| > 1, using a similar moderated t-test employed to derive differentially expressed lncRNAs). TFs with enrichment p- values < 0.05 were then considered as ‘active’ in the context of the study.…”

The Challenges and Opportunities of LncRNAs in Ovarian Cancer Research and Clinical Use

“…For instance, RNAInter [167] and StarBase-ENCORI [168] contain a vast, multi-omic-based amount of information about whole RNA interactomes, whereas miRTarBase [169] is only valid for miRNA interactions. In other databases, information about how lncRNAs are regulated can be retrieved, as in the case of ORTI [170], which contains information relating to TFs and their regulated genes, including lncRNAs. One reason why gene expression can be dysregulated is methylation status.…”