Abstract
Oxford Nanopore Technologies (ONT) offers an efficient and effective solution for studying DNA methylation. Both R9.4.1 and R10.4.1 ONT chemistries have been widely used to generate numerous DNA methylation data, making it inevitable to conduct cross-ONT-chemistry methylation analysis. However, the two ONT chemistries have unique designs and may cause potential bias in methylation detection, complicating cross-chemistry methylation investigation. In this study, we sequenced two pairs of wild-type and their knockout samples using R9.4.1 and R10.4.1 chemistries and investigated the concordance and bias of the two pairs of ONT methylation data. Although we confirmed high concordances of methylation data generated by the two ONT chemistries and the improvement of R10 chemistry in repeat regions, we found that both chemistries possess detection bias for methylation. Thus, cross-ONT-chemistry methylation studies identified hundreds of thousands of differential methylation sites caused by chemistry variabilities. We also explored different methods to calculate coverage and methylation percentages. Our evaluations provide valuable recommendations for cross-ONT-chemistry methylation analysis and suggest better practices for robust methylation investigation.