Identification of high-confidence human poly(A) RNA isoform scaffolds using nanopore sequencing

Abstract

Nanopore sequencing devices read individual RNA strands directly. This facilitates identification of exon linkages and nucleotide modifications; however, using conventional methods the 5′ and 3′ ends of poly(A) RNA cannot be identified unambiguously. This is due in part to the architecture of the nanopore/enzyme-motor complex, and in part to RNA degradation in vivo and in vitro that can obscure transcription start and end sites. In this study, we aimed to identify individual full-length human RNA isoform scaffolds among ~4 million nanopore poly(A)-selected RNA reads. First, to identify RNA strands bearing 5′ m7G caps, we exchanged the biological cap for a modified cap attached to a 45-nucleotide oligomer. This oligomer adaptation method improved 5′ end sequencing and ensured correct identification of the 5′ m7G capped ends. Second, among these 5′-capped nanopore reads, we screened for ionic current signatures consistent with a 3′ polyadenylation site. Combining these two steps, we identified 294,107 individual high-confidence full-length RNA scaffolds, most of which (257,721) aligned to protein-coding genes. Of these, 4,876 scaffolds indicated unannotated isoforms that were often internal to longer, previously identified RNA isoforms. Orthogonal data confirmed the validity of these high-confidence RNA scaffolds.

Advances in nanopore direct RNA sequencing

“…To verify these results, we reanalyzed human GM12878 poly(A) RNA data using Guppy v. 6.3.2 (default quality score cutoff = 7). For our group's data in the Workman study (~2.6 million reads) 6 and for the Mulroney et al study (~3.8 million reads) 10 , we found median accuracies of 90.6% and 89.8%, respectively, in agreement with the published results.…”

“…This obscures the true identity of that strand end. In our hands, even when we demonstrably sequenced poly(A) RNA from the 3′-poly(A) tail to the 5′-m7G cap, we could not resolve the final six nucleotides 10 . Consequently, it is important that investigators either achieve truly full-length reads or that they define what is meant by the term.…”

“…Nanopore DRS facilitates discovery of previously unknown mRNA isoforms. These preliminary discoveries require validation using orthogonal data such as reverse transcription (RT)-PCR, and transcription start site markers such as DNase-seq, polymerase II chromatin immunoprecipitation (ChIP)-seq, SPI1 ChIP-seq and CAGE (cap analysis of gene expression) 10 . For mRNA, unambiguous validation requires documentation of an associated protein 34 .…”

Quantitative analysis of tRNA abundance and modifications by nanopore RNA sequencing

“…Although several works have shown that this technology can also be used to study short RNA molecules, such as snoRNAs and snR-NAs 96,98 , DRS is inefficient at capturing RNA molecules shorter than 200 nucleotides (nt) and is generally considered unable to capture sequences shorter than ~100 nt 96,97 , limiting its applicability to study short RNA populations, such as tRNAs. In addition, the first ~15 nt at the 5′ end of RNA molecules are typically lost in DRS runs 93,99 , as this portion cannot be adequately basecalled due to the increase in the RNA translocation speed when the 5′ end of the molecule exits the helicase 99 . To overcome these limitations, we reasoned that extension of the 5′ and 3′ ends of the tRNA would lead to improved sequencing of tRNA molecules, as these would now be beyond the ~100-nt threshold, in addition to capturing the sequence and modification information of 5′ tRNA ends.…”

Context-Aware Transcript Quantification from Long Read RNA-Seq data with Bambu

“…However, the classification of reads as full-length in Bambu is influenced by both sequencing and alignment errors as well as RNA degradation, with non-unique full-length reads still being ambiguous. Approaches have been developed to enrich full-length reads experimentally 38 , or selectively analyse full-length reads during quantification 39 . These approaches are more effective in identifying which transcripts are truly expressed, however, as they significantly reduce the number of reads used for quantification, the overall abundance estimates are likely to be less accurate.…”