I mapped human RNAseq raw data to Refseq rRNA sequences using bwa, it turned out that 20% raw reads mapped to rRNA sequences. Is it normal? But I have used a kit to remove rRNA including cytoplasmic(5s, 5.8s, 18s, 28s) and mitochondrial(12s, 16s) ribosomal RNA. I found that the most mapped reads belong to 45S pre-ribosomal RNA sequences. I have no idea about this. Someone could help me! Any suggestion is welcome!