Confusing Picard HsMetrics ON_BAIT_VS_SELECTED for near=250 and near=0
0
0
Entering edit mode
16 months ago
Floydian_slip ▴ 150

Hi, for an assay developed by my vendor, I ran latest Picard HsMetrics (2.21.8) with two different options for the same set of BAM files: one with a NEAR_DISTANCE=250 and one with NEAR_DISTANCE=0. The bait and Target interval file was the same. With the default value of NEAR_DISTANCE=250, I got ON_BAIT_VS_SELECTED value around 0.45 for my 12 samples and with NEAR_DISTANCE=0 I got it around 0.5.

ON_BAIT_VS_SELECTED is defined as the fraction of bases on or near baits that are covered by baits, ON_BAIT_BASES/(ON_BAIT_BASES + NEAR_BAIT_BASES). In the 2nd case, since my near_distance = 0 it means that all the bases are either on the bait or outside the near distance so near_bait_bases = 0 which means that I should see ON_BAIT_VS_SELECTED = 1 but I see it around 0.5. Am I misunderstanding anything?

The other question I have is: if both my Target and Bait files are the same, why is my mean target coverage around 2000X where as mean bait coverage is 3000X.

sequencing next-gen picard HsMetrics • 508 views