ggDNAvis: a ggplot2-based R package for creating high-quality DNA sequence and modification visualisations

Abstract

In the genomics age, enormous volumes of DNA sequencing information are continuously produced and analysed. Pipelines for processing DNA information are widespread and mature. However, tools for rendering sequence and methylation information can be limited, often resulting in authors taking varied, manual approaches to visualising their DNA sequences. ggDNAvis is designed to easily produce high-quality renders of DNA sequence and methylation information, with extensive customisation options. The three core features are visualisation of a single DNA sequence, multiple sequences at once, or methylation of multiple DNA molecules at once. Additionally, ggDNAvis has tools for reading, writing, processing, and organising genetic data from file formats such as FASTQ to extract suitable inputs to the main visualisation functions. Single-sequence visualisation accepts any DNA or RNA sequence of any length from any source, and is thus extremely versatile and useful in a wide range of biological contexts. Multiple-sequence visualisation was conceived in the context of visualising many long-read sequencing (e.g. Nanopore) reads over causative genes of short tandem repeat (STR) expansion diseases such as neuronal intranuclear inclusion disease (NIID) caused by NOTCH2NLC, Huntingtons disease (HD) caused by HTT, and fragile X syndrome (FXS) caused by FMR1, but is generally applicable for visualising any set of multiple sequences. Methylation visualisation requires read-level methylation data, which is most commonly obtained through modification-capable basecalling of signal-level Nanopore data. ggDNAvis functionality is available as a Shiny web app and an R package on CRAN, with outputs from the latter supporting extension and annotation via the diverse ggplot2 ecosystem.