Hello, (hopefully you will see this comment I know it's been like 8 years now since your answer was posted) but I wanted to ask you how did you set up your bwa-mem reference file so that the contigs (chromosomes in your case) had different lengths?
I'm running into a weird situation where I have a reference file with various CDS sequences all different lengths, yet when I run the samtools idxstats sp1.bam | head -n 5 command, the contigs are all the same length and always the length is for the longest CDS sequence in the file. Any ideas??