Extended Data Fig. 10: RNA m5C BS-seq performance under different conditions, sanger validation and RNA pairing analysis.
From: Systematic calibration of epitranscriptomic maps using a synthetic modification-free RNA library
a) The workflow to conduct BS-seq under different conditions of bisulfite treatment. b) Comparison of the overall conversion rates estimated by ERCC mixes under different conversion conditions. Data are presented as mean values + /- SD (n = 4). c) Boxplot for conversion rates in each ERCC oligos under different conditions (n = 39,372, 4 replicates). Reads depth ≥10 is used to filter the low-abundance oligos. Box boundaries represent 25th and 75th percentiles; center line represents the median; whiskers indicate ±1.5× interquartile range (IQR). d) Counts of non-converted sites before and after calibration using IVT RNA control. Three datasets under different conditions are compared to show the effectiveness of calibration. e) Putative methylation rates (non-conversion rates) of five candidate m5C sites validated by BS-converted Sanger sequencing. The reported methylation rates from previous literatures are also included. The PCR primers for sequencing are listed in Supplementary Table 29. f) Correlation of methylation rates for five sites determined in BS-seq and Sanger sequencing. g) Pairing probabilities of upstream and downstream 25 bp flanking sequencing are calculated in respect to the high confidence m5C sites, false-positive sites and random sampling sites. Coordinate 0 indicates the location of interested sites. Pairing bias = log2(Pair_probfalse (i,j)/Pair_probrandom (i,j)), where Pair_prob(i,j) is the mean pairing probability of bases between position i and position j. Red color indicates the RNA fragments are more likely to form pairing than random sites.
