I have four RNA-seq data sets that I've been comparing. For a large number of genes there are only a handful of reads (say 1-10) that map back to that gene. Others are in the range of 11-20, and so on.
What is a reasonable cutoff for the minimum number of reads. If there is only 1 read for a gene in sample A and sample B, but none in C and D that is so low that one really can't draw any reasonable conclusions from that. So where do you make that cutoff. Does anyone know of a statistical test that can be used to decide this?
Thanks
What is a reasonable cutoff for the minimum number of reads. If there is only 1 read for a gene in sample A and sample B, but none in C and D that is so low that one really can't draw any reasonable conclusions from that. So where do you make that cutoff. Does anyone know of a statistical test that can be used to decide this?
Thanks
Comment