I am attempting to use Bismark to align my bisulfite sequencing reads to the reference sequence. In my experiment I used a patch PCR method to enrich for a ~40kb region of the genome so the genome_prep performed in Bismark was done with this sequence. Trimming of the reads was done using cutadapt. I was having trouble removing the adapter sequence in skewer because of SNP's in the adapter so I switched over to cutadapt and chopped of 25 bases at the 5' end and quality trimmed the 3' end. When I then attempt to align these reads in Bismark using the --bowtie2 and --non_directional options I see that zero of around 14k reads are aligning. Even when reducing the quality threshold to its lowest in every category still zero reads align. Has anyone else had this issue before? I am hoping that this points more towards an issue with the program and not my sequence data since I would expect even a part of a few reads to align at low quality thresholds. Is there any way to tell if my data is junk when it has been bisulfite treated?
This is unlikely to be a bismark problem and likely has more to do with how you prepared the genome. Try aligning against the whole genome and see what happens. It may be that you simply aligned against the wrong region.