Supplementary Figure 1. Direct RNA sequencing of the IVT pool. a, Bar plots showing the number of reads of GAPDH, ENO2, 18S rRNA, and ACTB present in bulk sequencing. b-e, Density plots showing the length of mapped reads of GAPDH (b), ACTB (c), ENO2 (d), and 18S rRNA (e) in bulk sequencing. f, Density plots showing the distribution of read lengths in the rejected reads pool when 1 or 3.5 sec decision time is used in adaptive sampling.
Supplementary Figure 2. Using adaptive sampling to deplete transcripts of interest. a, Table showing the parameters for break times (decision times) tested in the depletion mode of adaptive sampling. b, Density plots showing the distribution of read lengths of rejected reads under different decision times of adaptive sampling. c, Barplots showing the proportion of reads belonging to 18S rRNA, ACTB, GAPDH, and ENO2 in rejected reads after depleting ENO2. d, IGV plots showing the distribution of read coverage along the length of GAPDH, ACTB, 18S rRNA, and ENO2 after GAPDH and ENO2 depletion by adaptive sampling. e-g, Barcharts showing the proportion of reads from 18S rRNA, ACTB, GAPDH, and ENO2 RNAs in the rejected read pool after depleting GAPDH and ENO2 (e), GAPDH (f), 18S rRNA and ACTB (g).
Supplementary Figure 3. Features of RNA that enable enrichment of RNAs in the transcriptome. a, Barplots showing the distribution of abundance in the candida albicans transcriptome. 319 genes from the top 80-90% of abundant genes are selected for enrichment. b, Scatterplot showing the percentage of reads belonging to the depleted 4997 transcripts in bulk sequencing (X-axis) and in adaptive sampling (Y-axis). c, Scatterplots showing the percentage of reads belonging to the depleted 4997 transcripts of different lengths (200-400bp, 400-600bp, 600-1000bp, >1000bp) in bulk sequencing (X-axis) and in adaptive sampling (Y-axis). d, Scatterplots showing the percentage of reads belonging to the depleted 4997 transcripts in bulk sequencing (X-axis) and in adaptive sampling (Y-axis). The transcripts are binned according to different abundances (>500 reads, >100 and <500 reads, and <100 reads).
Supplementary Figure 4. Adaptive sampling does not impact pore health. a, Barplots showing the number of available pores over scan frequencies, single pores refer to pores that are available for sequencing. b, Detailed status of pores over the lifetime of the sequencing run. The x-axis corresponds to mux scan number (scan frequency is 1.5 hours) and the Y-axis indicates the number of pores.