Fig. 4: RISER enables real-time depletion of mRNA and mtRNA.

RISER performance during live sequencing of poly(A)⁺ RNA from HEK293 cells, using a MinION Mk1B flow cell split into two conditions: RISER targeting both mRNA and mtRNA for depletion (pink), and no RISER as a control (blue). a Distribution of read lengths (y-axis, log10-scale) for each RNA class in each condition. In the box plots, the lower and upper boundaries of the box are the first and third quartiles, with the median annotated with a line inside the box. The whiskers extend to the maximum and minimum values within 1.5 times the interquartile range. Outliers were not included. The read lengths of each RNA class were compared in the control and deplete conditions using a one-tailed Wilcoxon rank sum test (H1: control > deplete). The probability of superiority (PS) is also shown above each comparison. PS is the probability that a randomly sampled read from the control condition is longer than a randomly sampled read from the deplete condition (i.e., PS close to 0.5 means the lengths are likely to be the same, whereas PS close to 1 means that the control lengths are highly likely to be larger) (mRNA: p-value p < 2.2E−308, test statistic U1 = 30493942144.5, PS = 0.87 and n = 376,048, mtRNA: p < 2.2E–308, U1 = 909552316.0, PS = 0.89 and n = 65,569, lncRNA: p = 2.3E−69, U1 = 13787658.0, PS = 0.60 and n = 9574). b Percentage of reads covering the first 1500 bases from the 3′ ends of the transcript (y-axis) for  an example mRNA (upper panel), mtRNA (middle panel) and lncRNA (lower panel). The reference positions (x-axes) are ordered from 3′ to 5′. The vertical line indicates 280nt upstream of the 3′ end, which approximately corresponds to the maximum RISER input length of 4 s. c Density distributions of the transcript fraction (x-axis) covered by the sequenced reads. d, e Distribution of the percent change in read (d) and nucleotide (nt) (e) counts (y-axis), with respect to a control run, when RISER was used to deplete mRNA and mtRNA (orange) and for a separate control run (purple). In the RISER vs control comparison, lncRNA read counts increased from 602 (control) to 885 (depletion of mRNA and mtRNA by RISER). The box plots are defined the same as in (a). Outliers were not included. For the set of transcripts in each biotype, the percent change in nt or reads using RISER was compared to the (no-RISER) control using a paired one-tailed Wilcoxon signed rank test (H1 for mRNA and mtRNA: RISER vs control < between controls, H1 for lncRNA: RISER vs control > between controls). For this comparison, lncRNAs that did not overlap with coding exons from protein-coding transcripts were used. Test parameters and statistics are provided in Suppl. Table 8. Source data are provided as a Source Data file.