I am wondering whether mapping reads to a very short reference like the reference is only 1kb or so but reads are 150 bp. I do notice that bowtie2 and bbmap2 will be problematic because they have a significantly less number of reads mapped to those 1kb compare to when those 1kb are part of a much longer sequences (we can alway extract reads mapped to the 1kb according to its position in the longer sequence). I assume it is because essentially all the mappers are based on local alignment, for reads with alignment ratio less than a certain threshold, like 0.4 or so, for example (only part of the reads mapped to 1kb), they will be filtered. Should I worry if I mapped reads to such short references and I want to calculate reference coverage or total number of reads mapped to it.
Dear bwa team,
I am wondering whether mapping reads to a very short reference like the reference is only 1kb or so but reads are 150 bp. I do notice that bowtie2 and bbmap2 will be problematic because they have a significantly less number of reads mapped to those 1kb compare to when those 1kb are part of a much longer sequences (we can alway extract reads mapped to the 1kb according to its position in the longer sequence). I assume it is because essentially all the mappers are based on local alignment, for reads with alignment ratio less than a certain threshold, like 0.4 or so, for example (only part of the reads mapped to 1kb), they will be filtered. Should I worry if I mapped reads to such short references and I want to calculate reference coverage or total number of reads mapped to it.
Thanks,
Jianshu