VQSR - overlapping training sets
when I have multiple training sets for VQSR,
what happens with overlapping variants between these two sets?
Especially when the two sets have different priors attached.
Example from Best Practices:
I have HapMap with prior Q15 and
1000G with prior Q10.
How will a variant influence the model when being in both sets?
I would expect that only the highest prior will be used.