Nanopore-based native RNA sequencing provides insights into prokaryotic transcription, operon structures, rRNA maturation and modifications

Felix Grünberger; Robert Knüppel; Michael Jüttner; Martin Fenk; Andreas Borst; Robert Reichelt; Winfried Hausner; Jörg Soppa; Sebastien Ferreira-Cerca; Dina Grohmann

Before expanding the analysis to the full-length mature rRNA, we focused on the dimethylation 580 was very high for the stretch of Guanosines adjacent to position 1450 (Fig. 6a) . Mapping to single-584 nucleotide resolution is difficult as more than one base contributes to the actual electric current 585 signal in the nanopore 83 . In the next step, a comparison of a wildtype sample to a deletion mutant 586 of KsgA confirmed the absence of the dimethylation in the deletion strain and confirmed that the 587 m 6 2A modifications and not the homopolymer-G-stretch are responsible for the altered current 588 signal (Fig. 6b) . The analysis further revealed a reduced signal variability at non-modified 589 positions between the two samples in comparison to the theoretical distribution, which leads to 590 less false positives in the statistical analysis. As the de novo model has a higher error rate 591 compared to the sample-compare model, we generated two read sets by sorting full-length 16S 592 rRNA transcripts according to their 5´ leading length to mature (n: 2744) and pre-rRNAs (n: 102) 593 in the Haloferax wildtype data set (Fig. 6c) . The calculated base modification probabilities and 594 electric current signal distributions for this model are very similar to the wildtype/∆ksgA set, 595