I'm processing (peak merging, differential peak finding, drawing Venn diagram and PCA plot) my m6A-seq data recently using DiffBind. However, as the program is developed originally for analyzing ChIP-seq data, I'm hesitant to apply default settings to my data which is actually from RNA based immunoprecipitation sequencing.

Using default settings of the program originally developed for ChIP-seq data to process m6A-seq data may not be very suitable. For example, I previously called peaks using MACS2 and let it built models to predict the fragment size automatically, but aberrant results appeared as pre-mRNA splicing events will make aligner generate large gaps between alignments. So maybe some parameters of DiffBind should also be tweaked manually. I'm wondering which parameter should I particularly be aware of? Such as fragment size, summits, DBA$config$singleEnd and etc. in dba.count?

Any advice would be much appreciated.

Regards, Zhiyi