Target mean coverage : Picard Versus Qualimap, why so different?

Question

[pegeot.henri]

Hello,

I would like to know if any of you know why the target mean coverage is so different when calculated by Picard or Qualimap.

For thesame bam and bed files :

Exome A ; Qualimap = 127.8 ; Picard = 88.2

Exome B ; Qualimap = 151.9 ; Picard = 104.9

Exome A ; Qualimap = 113.6 ; Picard = 77.7

How do you explain the differences ? I know Picard excludes duplicates, however, I cannot find the information for qualimap.

Additionally, do you recommend a method or command line to calculate mean target coverage ?

Thank you