tailfindr: alignment-free poly(A) length measurement for Oxford Nanopore RNA and DNA sequencing

Abstract

Polyadenylation at the 3 ′ ′ ′ ′ ′ -end is a major regulator of messenger RNA and its length is known to affect nuclear export, stability, and translation, among others. Only recently have strategies emerged that allow for genome-wide poly(A) length assessment. These methods identify genes connected to poly(A) tail measurements indirectly by short-read alignment to genetic 3 ′ ′ ′ ′ ′ -ends. Concurrently, Oxford Nanopore Technologies (ONT) established full-length isoform-specific RNA sequencing containing the entire poly(A) tail. However, assessing poly(A) length through base-calling has so far not been possible due to the inability to resolve long homopolymeric stretches in ONT sequencing. Here we present tailfindr, an R package to estimate poly(A) tail length on ONT long-read sequencing data. tailfindr operates on unaligned, base-called data. It measures poly(A) tail length from both native RNA and DNA sequencing, which makes poly(A) tail studies by full-length cDNA approaches possible for the first time. We assess tailfindr's performance across different poly(A) lengths, demonstrating that tailfindr is a versatile tool providing poly(A) tail estimates across a wide range of sequencing conditions.

So close, no matter how far: multiple paths connecting transcription to mRNA translation in eukaryotes

“…The estimation of poly(A) tail length is based on the analysis of monotonous raw signal at the beginning of the reads, while long reads enable their assignment to mRNA isoforms. More precise calculations of the tails' length can be done using dedicated algorithms, such as tailfindr (Krause et al, 2019).…”

Barcoding and demultiplexing Oxford Nanopore native RNA sequencing reads with deep residual learning

“…In 2017, the direct RNA sequencing (dRNAseq) technology appeared, making it possible for the first time to sequence native RNA molecules. Importantly, this technology could also identify chemical RNA modifications present in the native RNA molecules (4,5,8,15) , as well as estimations for their polyA-tail lengths (16,17) . However, a major caveat of dRNAseq is the amount of poly(A)-selected RNA material that is needed, i.e., typically 500ng of poly Compounding this fact, many of the recently developed DNA base-callers for nanopore signals rely on the use of DNNs, such as DeepNano (19) , DeepSignal (20) or Chiron (21) .…”

Parallel and scalable workflow for the analysis of Oxford Nanopore direct RNA sequencing datasets

“…Thus, the current outputs of the MasterOfPores workflow include: (i) base-called FAST5 files, (ii) base-called fastq file, (iii) mapping BAM file, (iv) MultiQC report, and (v) NextFlow report. In the future we plan to integrate within the MasterOfPores workflow the software for the downstream analyses of direct RNA sequencing datasets including the PolyA tail length estimation, using Nanopolish (Workman et al, 2018) and tailfindr (Krause et al, 2019)) per-transcript isoform quantification and differential expression analysis, using Flair and the analysis of RNA modifications, using Tombo (Stoiber et al, 2017) and EpiNano (Liu et al, 2019)).…”

“…Third-generation sequencing (TGS) platforms, such as the one offered by Oxford Nanopore Technologies (ONT), allow for direct measurement of both DNA and RNA molecules without prior fragmentation or amplification (Brown and Clarke, 2016), thus putting no limit on the length of DNA or RNA molecule that can be sequenced. In the past few years, ONT technology has revolutionized the fields of genomics and (epi)transcriptomics, by showing its wide range of applications in genome assembly , study of structural variations within genomes (Cretu Stancu et al, 2017), 3' poly(A) tail length estimation (Krause et al, 2019), accurate transcriptome profiling (Bolisetty et al, 2015), identification of novel isoforms (Byrne et al, 2017;Križanovic et al, 2018) and direct identification of DNA and RNA modifications (Carlsen et al, 2014;Garalde et al;Liu et al, 2019;Simpson et al, 2017). Thus, not only this technology overcomes many of the limitations of short-read sequencing, but importantly, it also can directly measure RNA and DNA modifications in their native molecules.…”