Signal Analysis Of Nanopore Rna Sequencing To Interrogate Poly(A) Tails And Post-Transcriptional Modifications

Although development of high throughput RNA sequencing technologies has allowed significant advances in our understanding of transcriptome complexity and regulation, certain questions remain intractable with conventional short read cDNA sequencing. Native RNA sequencing technology can address these difficult questions; including identity and frequency of splice variants, full-length poly(A) tail assessment, base modifications and transcript haplotyping. To accomplish this, we leveraged our extensive consortia-generated GM12878 native RNA dataset to showcase this exciting technology and the promise it holds for the direct and simultaneous interrogation of these RNA features. We have applied tools to directly interrogate the raw electrical current data from nanopore sequencing to examine these questions. Specifically, we have begun to empirically establish how different modifications in variable sequence contexts will modulate the nanopore current. To accomplish this we have generated training sets to characterize N6-methyladenosine, focusing on the METTL3 motif (GGm6ACU). We identified variable current signatures in regions with known modifications, even identifying isoform specific modification patterns. Examining poly(A) tail lengths, we have applied software (nanopolish) to call poly-A tails from our native RNA sequencing data. Using this software we measured gene specific and even isoform specific poly(A) tail lengths. As expected, we measured a clear difference between mitochondrial transcripts with a shorter poly(A) tail length than nuclear transcripts. Specifically we found different isoforms of the same gene with different poly(A) tail lengths, with isoforms with retained introns typically having a higher average tail length than isoforms where the intron was spliced out.