TIMEOR: a web-based tool to uncover temporal regulatory mechanisms from multi-omics data
Poster sessions are particularly prominent at academic conferences. Posters are usually one frame of a powerpoint (or similar) presentation and are represented at full resolution to make them zoomable.
Uncovering how transcription factors (TFs) and their targets communicate at the DNA, RNA and protein levels is critical to understand signaling cascades in normal and pathogenic cells. RNA-seq has become the de factostandard to detect gene regulatory networks (GRNs) using an established set of analysis steps. Although RNA-seq analysis pipeline methods comparing two conditions have been studied, methods for interpreting ordered data (in time or space) are still in their infancy and are essential to assign cause and effect relations. Most current pipeline methods treat order as a categorical variable, thus disregarding the ordered relationship between genes, and thus not accurately reconstructing GRNs. Moreover, RNA-seq data do not provide direct evidence of interaction. Thus, methods integrating ordered RNA-seq and ChIP-seq data are urgently needed. To date there is no accessible and adaptive time course multi-omics pipeline method supporting reproducibility.
Here we present TIMEOR: Trajectory Inference and Mechanism Exploration using Omics data in R to fill this gap. This is an automatic interactive web and command line time course multi-omics pipeline method for differential gene and isoform expression (DE). It takes raw .fastq files and performs all analysis from quality control and DE to GRN reconstruction. TIMEOR has five unique features: 1) adaptive default analysis methods given an experimental design; 2) multiple method comparisons for alignment and DE (for distant and close timepoint); and 3) statistical, graphical and interactive results for data exploration. Within each cluster, 4) TIMEOR performs automated gene enrichment, pathway, network and motif analysis, and optional ChIP-seq analysis for binding and epistasis relations. Lastly, merging gene networks, time course RNA-seq and ChIP-seq data 5) TIMEOR reconstructs GRNs with directed edges by labeling the interaction type between genes and gene products.
We validated TIMEOR's GRN reconstruction with both simulated and real data. Using a previously published ten close timepoint Drosophila RNA-seq dataset, TIMEOR recapitulates past results and finds novel regulatory effects on circadian rhythm regulators after insulin stimulation. Moreover, TIMEOR reconstructs the GRN between perturbed and putative TFs by establishing high confidence interactions using ChIP-seq data for each TF. Overall, TIMEOR is the first integrative, web-based method to predict GRNs from time course RNA-seq and ChIP-seq data.
Support For National Science Foundation Graduate Fellows
Directorate for Computer & Information Science & EngineeringFind out more...