Chapter 3 Fluent genomics with plyranges and tximeta

We construct a simple workflow for fluent genomics data analysis using the R/Bioconductor ecosystem. This involves three core steps: import the data into an appropriate abstraction, model the data with respect to the biological questions of interest, and integrate the results with respect to their underlying genomic coordinates. Here we show how to implement these steps to integrate published RNA-seq and ATAC-seq experiments on macrophage cell lines. Using tximeta, we import RNA-seq transcript quantifications into an analysis-ready data structure, called the SummarizedExperiment, that contains the ranges of the reference transcripts and metadata on their provenance. Using SummarizedExperiments to represent the ATAC-seq and RNA-seq data, we model differentially accessible (DA) chromatin peaks and differentially expressed (DE) genes with existing Bioconductor packages. Using plyranges we then integrate the results to see if there is an enrichment of DA peaks near DE genes by finding overlaps and aggregating over log-fold change thresholds. The combination of these packages and their integration with the Bioconductor ecosystem provide a coherent framework for analysts to iteratively and reproducibly explore their biological data.