Abstract
Many cellular mRNAs contain the modified base m6A, and recent studies have suggested that various stimuli can lead to changes in m6A. The most common method to map m6A and to predict changes in m6A between conditions is methylated RNA immunoprecipitation sequencing (MeRIP-seq), through which methylated regions are detected as peaks in transcript coverage from immunoprecipitated RNA relative to input RNA. Here, we generated replicate controls and reanalyzed published MeRIP-seq data to estimate reproducibility across experiments. We found that m6A peak overlap in mRNAs varies from ~30 to 60% between studies, even in the same cell type. We then assessed statistical methods to detect changes in m6A peaks as distinct from changes in gene expression. However, from these published data sets, we detected few changes under most conditions and were unable to detect consistent changes across studies of similar stimuli. Overall, our work identifies limits to MeRIP-seq reproducibility in the detection both of peaks and of peak changes and proposes improved approaches for analysis of peak changes.
Highlights
Many cellular mRNAs contain the modified base m6A, and recent studies have suggested that various stimuli can lead to changes in m6A
We found that of these peaks, 4/4 from MACS2, 5/5 from MeTPeak, and 4/5 from MeTDiff showed decreased m6A(m) enrichment following METTL3/METTL14 depletion, suggesting that these are true m6A sites
For our comparison of m6A(m) peak changes in these data sets, we identified probable changes in peaks based on statistical significance using Quad-negative binomial (QNB) or the generalized linear models (GLMs) with log[2] fold change difference between peaks and genes of ≥1
Summary
Many cellular mRNAs contain the modified base m6A, and recent studies have suggested that various stimuli can lead to changes in m6A. While studies have suggested that m6A shows widespread changes in response to diverse stimuli, they have applied inconsistent analysis methods to detect changes in m6A and often don’t control for differences in RNA expression between conditions or typical variability in peak heights between replicates. In some cases, these studies have reported m6A changes based on simple differences in peak count[24,26,27,32]. Only one MeRIP-seq study has used more than three replicates per condition[34], while ten have used only one[17,20,32,33,38,39,40,41,42,43], suggesting that most studies may not have enough power to detect changes in m6A(m)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
