I've used Bismark v0.14.1 (http://www.bioinformatics.babraham.ac.uk/projects/bismark/) to align bs-seq reads and call methylated C's. Now I'd like to perform a differential methylation analysis using bsseq v1.6.0.
I've generated the coverage files using
bismark_methylation_extractor but I don't understand whether I should use the 1-based files (.cov) or the 0-based ones (.zero.cov).
bsseq::read.bismark() help says:
• The genomic co-ordinates of the Bismark output file may be zero-based or one-based depending on whether the ‘--zero_based’ argument is used. Furthermore, the default co-ordinate system varies by version of Bismark. bsseq makes no assumptions about the basis of the genomic co-ordinates and it is left to the user to ensure that the appropriate basis is used in the analysis of their data.
which I find a bit ambiguous.
Should I use 1-based or 0-based coverage files from Bismark to perform a differential methylation analysis in bsseq?