Very wide drop out range in Sensitivity Analysis

[Tatjana Curovic]

Hello,

If one gets a drop-out range in SA of 0.04 to 0.94 for the suspect (Hp are suspect plus 2 unknowns, or 3 unknowns) and within this range the log LR ranges from -1 to 5, can such a mixture even be interpreted at all? 

Do we still apply the conservative rule of choosing the drop out rate that yields the lowest LR? Indeed in my case a very high drop out rate is expected due to circumstances of the case ( fired bullet) and a low drop out is masked by the presence of a full major component and its questionably called stutters. The major's alleles and his stutters account for 21 allelic numbers of the POI minor. The mixture is 11:1 where 1 accounts for two minor contributors.  The total quantity of biological material in the mixture was less than 0.01 ng ( 100 picograms) .   Obviously it is a very complex mixture,, due to high stutters, expected degradation of minors, indistingishable stutter positions from minors, LCN, and  very high degree of allele sharing between POI and the major component and its stutters.  Personally I think this is one of those mixtures that simply cannot be interpreted and wonder if the wide range of the drop out rate in LRMix SA gives further evidence of its non interpretability? Or, if it can ber interpreted how does one first  interpret such a wide drop out range?   

Thank you very much for any insights or further direction on this. 

Tanya