I aligned my single-end illumina reads with bwa aln to the human mitochondrial reference.
Then took only aligned reads and called consensus with:
samtools mpileup -uf mt_dna.fa mt_sorted.bam | bcftools view -cg - | perl vcfutils.pl vcf2fq -d 2 > cns.fq
The realigned my mapped reads to the called consensus, however now only 23% of them mapped.
Why is that?