I am analyzing small RNA-seq data. I mapped the reads with BWA and used FeatureCounts to count reads mapped to individual microRNAs and found the following issues.
For microRNA genes with uniquely mapped reads (Primary @ MAPQ 25, X0=1, XT=U),
1)When I used only -M option, it gives a read count = 1 per uniquely mapped read. However, when I use --fraction -M option, it gives a read count = 0.25 or 0.33 or 0.5 per uniquely mapped read. Isn't it true that FeatureCount should only assign a fraction for a multimapping read but not a uniquely mapped read? This happened in almost all genes.
2) When I did not use -M option, it gives a read count = 0 per uniquely read. FeatureCount gives a lot of zero count per feature when there are uniquely mapped reads (Primary @ MAPQ 25, X0=1, XT=U).
I am not sure how to troubleshoot and am looking for advice from the community. Thank you.