I want to build a table of counts using 'summarizeOverlaps()' from a PE RNA-seq data set generated with a strand-specific library preparation protocol from NewEngland Biolabs called "NEBNext Ultra Directional RNA Library Prep Kit for Illumina". The web page from that protocol says that it "utilizes dUTP-based methodology". The help page from the 'GAlignmentPairs' class explains that when PE data is generated using a stranded protocol such as 'dUTP' one should build such an object using the non-default argument 'strandMode=2'. I want to use 'summarizeOverlaps()' over the set of BAM files, so doing something like:
bfl <- BamFileList(bamfiles, asMates=TRUE) se <- summarizeOverlaps(features, bfl, mode="Union", singleEnd=FALSE, ignore.strand=FALSE, fragments=TRUE)
I've tried to set 'strandMode=2' in the call to 'BamFileList()' and 'summarizeOverlaps()' and within 'ScanBamParam()' through the 'param' argument to 'summarizeOverlaps()', but the argument 'strandMode' does not belong to any those methods. I've googled about it without luck, so my question is, how can I ensure that 'strandMode=2' is being used under the hood through this approach to ensure I'm informing the counting algorithm about the correct library preparation protocol?