samtools stats - percentage of target genome covered is greater than 100

[miguelpmachado]

Dear Samtools team,
I recently run samtools stats with --target-regions and noticed that the "percentage of target genome with coverage > 0 (%)" is greater than 100 (103.04).
This sounds odd to me, but I couldn't find any extra information that would clarify me why is this happening.
I attached the samtools stats report
If you could help me, I would appreciate.
Thank you in advance.
Best regards,
Miguel

I'm using the v1.9 with a Docker image (miguelpmachado/samtools:1.9)

[valeriuo]

I am struggling to reproduce this issue. Can we have an anonymized subset of your data which triggers this behaviour?

[miguelpmachado]

Hi @valeriuo ,
Thank you for being checking this.
I downsampled the original files and produced a similar result. It's not 103.04%, but it is grater than 100% (100.36%).
I uploaded the files to WeTransfer. You can find them here: https://we.tl/t-J46ZTUSiKc
There you can find the bam, the target regions file and the samtools stats report that I obtained.
I hope you can help me trying to understand what is going on.
Thank you once more.
Best regards,
Miguel

[valeriuo]

Thank you for the data! I've managed to identify the problem. A fix will follow shortly.