Tweet
Hi,
I am trying to count reads from BAM files in specific BED features using bedtools multicov.
If I use the standard parameters (i.e. just -bams x.bam y.bam -bed x.bed), it works as expected.
However, I have problems with the parameters for overlap, -f and for strandedness, -s/-S.
With the settings to -s (enforcing same strand), I get 0 counts. With the inverted -S, I get counts. Since I am using datasets that have been generated with dUTP first strand method, is this expected? Cufflinks assembles the correct orientations of my transcripts with --fr-firststrand.
For the overlap, since my bed files contain exons, I want to exclude reads from pre-mRNA that span intron-exon, therefore I set the overlap to 100% with -f 1. However, this will result in 0 counted reads. Do I misunderstand the -f setting? I only want to count reads that are 100% within the BED features of interest. I have paired-end data, so I use the -split option.
I am trying to count reads from BAM files in specific BED features using bedtools multicov.
If I use the standard parameters (i.e. just -bams x.bam y.bam -bed x.bed), it works as expected.
However, I have problems with the parameters for overlap, -f and for strandedness, -s/-S.
With the settings to -s (enforcing same strand), I get 0 counts. With the inverted -S, I get counts. Since I am using datasets that have been generated with dUTP first strand method, is this expected? Cufflinks assembles the correct orientations of my transcripts with --fr-firststrand.
For the overlap, since my bed files contain exons, I want to exclude reads from pre-mRNA that span intron-exon, therefore I set the overlap to 100% with -f 1. However, this will result in 0 counted reads. Do I misunderstand the -f setting? I only want to count reads that are 100% within the BED features of interest. I have paired-end data, so I use the -split option.