R2Dtool: Integration and visualization of isoform-resolved RNA features

Abstract

Long-read sequencing enables isoform-resolved detection of functionally important RNA elements, such as RNA chemical modifications, RNA secondary structure, or RNA-protein interaction sites. Importantly, the functional impact of these elements can relate to their positions relative to isoform-specific transcript features, such as start and stop codons, open reading frames, exon-exon junctions and transcript termini. Relating transcriptomic and genomic features during sequencing data analysis is challenged by the flexibility of RNA biogenesis and the diversity of alternative mRNA transcripts. To address these challenges and streamline the mapping between transcriptome-mapped data and genome-mapped data, we developed R2Dtool. We illustrate R2Dtool capability to process long-read transcriptomic information and potential for interpretation in the context of transcript and genomic annotation features using epitranscriptomics data. R2Dtool embraces and expedites analysis of RNA complexity, and we anticipate it will empower multiple transcriptomic studies for interpretation and discovery. R2Dtool is freely available under the MIT license from https://github.com/comprna/R2Dtool.

Prediction of m6A and m5C at single-molecule resolution reveals a cooccurrence of RNA modifications across the transcriptome

“…We used our R2Dtool 65 (https://github.com/comprna/R2Dtool) to perform positional annotation of the CHEUI Model 2 RNA methylation calls, and to transpose the methylation predictions from transcriptomic to genomic coordinates. First, the R2Dtool script cheui_to_bed .…”

“…We used our R2Dtool 72 (https://github.com/comprna/R2Dtool) to perform positional annotation of the CHEUI Model 2 RNA methylation calls, and to transpose the methylation predictions from transcriptomic to genomic coordinates. First, the R2Dtool script cheui_to_bed.sh was run with default parameters to convert the CHEUI methylation calls to a bed-like format (i.e., tab-delimited, where column 1 represents the reference sequence, column 2 represents interval start, and column 3 represents interval end).…”

“…The absolute distance (in nucleotides) and relative metagene position (as a fraction of the overall UTR or CDS length) of each methylation site with respect to the reference transcript isoform were calculated using R2Dtool 72 . The relative meta-transcript coordinates were derived as previously described 73 , placing the modifications along three equal-sized segments of length L. Position 0 represents the transcript start site (TSS), position L represents the CDS start, position 2L represents the CDS end, and position 3L represents the polyadenylation site (PAS).…”

Integrative analysis of nanopore direct RNA sequencing data reveals a role of PUS7-dependent pseudouridylation in regulation of m6A and m5C modifications

“…The limited detection of Ψ sites in the 5′ UTR could be due to a lower level of coverage of the 5′ UTR compared to CDS and 3’UTR as nanopore sequencing moves in the 3’ to 5’ direction. To address this issue, we generated a metatranscript plot using R2D tools 20 to compare the proportion of significant Ψ sites along mRNA transcripts, which showed comparable Ψ density distribution in the 5’UTR, CDS and 3’UTR (Figure 2H).…”

Prediction of m6A and m5C at single-molecule resolution reveals a transcriptome-wide co-occurrence of RNA modifications

“…The absolute distance (in nucleotides) and relative metagene position (as a fraction of the overall UTR or CDS length) of each methylation site with respect to the reference transcript isoform were calculated using R2Dtool 68 . The relative meta-transcript coordinates were derived as previously described 69 , placing the modifications along three equalsized segments of length L. Position 0 represents the transcript start site (TSS), position L represents the CDS start, position 2 L represents the CDS end, and position 3 L represents the polyadenylation site (PAS).…”