3 Person Mixture 4 Or is it 2

3 Person Mixture #4 (Or is it 2? ) The Hardest Mixture I Know

Bed Sheet Stain 1

Bed Sheet Stain 2

Suspect Underwear

Data Table • Is Victim included?

The Hardest Mixture I Know • We have to deal with family sexual assault cases • These by definition include biologically related people • Sometimes step-parent situation • Sometimes biological parent situation • But they all live together so no “random man”

The Hardest Mixture I Know • Evidence commonly shows alleles of both Mom and Dad (or Mom and Step-Dad) • Some advantages: – Not “unrelated” people (well, step-dad) – Not “random” people • Therefore, no statistics apply – “The probability of a random, unrelated person contributing to the mixture is…” • If someone knows a relevant stat, let us know

The Hardest Mixture I Know • OK - so we don’t worry about a stat • What do we need to worry about? – Hp vs. Hd – (But I don’t know a relevant LR to use) • But we are asked to answer a question: – Is the victim contributing to that semen stain found on the parents bedding and Dad’s boxers or not?

The Hardest Mixture I Know • Is she “there” because it’s mixture of her biological parents • OR • Is she “there” because she has been sexually assaulted but her biological father • These types of cases are the most difficult mixtures I’ve ever seen (obvious reasons)

The Hardest Mixture I Know • So it’s what I call a “real time” LR • It has nothing to do with any LR stat that I know • Instead I assume each hypothesis is true and I test them • Then I decide if one hypothesis is supported better than the other

The Hardest Mixture I Know • How do I do that? • Start with Defense hypothesis: – The evidence is consistent with a 2 person mixture of Mom and Dad only – Use the 2 person calculator with only Mom and Dad applied • Does it “graph” well? • Is there a “valid” combination at each locus? – In other words, does our validated PHR and P calculator show that the data “fits”

The Hardest Mixture I Know • Then the Prosecution hypothesis: – The evidence is consistent with a three person mixture of Mom, Dad, AND Victim (bio daughter) – Use the 3 person calculator with all three reference profiles applied • Does it “graph” well • Is there a “valid” combination at each locus? – In other words, does our validated PHR and P calculator show that the data “fits”

The Hardest Mixture I Know • A note about testing the three person hypothesis • Bio Mom, Bio Dad, and Bio Child is worst case scenario for “not excluded” options – The “double” or “circular” sharing types – We will not get much in terms of P and PHR • The “graph” will be pretty lousy • BUT we can still see if there is a “valid” contributor option in our 3 person calculator

The Hardest Mixture I Know • The three person graph for this scenario at the time this handout was prepared is basically worthless • At all the loci where a “not excluded” combination is viable, we see empty space in the graph – all we get is the name of the locus on the X-axis • At all loci where two contributors share a type, we don’t even get an X-axis label

The Hardest Mixture I Know • So I just will give a “count” for the number of loci that show a valid three person option • Our programmers are working feverishly to get this working better by the workshop – (Buy stock in Mt. Dew) • Show P for the one person that doesn’t share with a designation that the other 2 share a type • Show a designation for “not excluded” loci • Ultimately solve the “not excluded” dilemma(? )

The Hardest Mixture I Know • We made a bunch of validation samples from three volunteers I know very well – I had to bribe my daughter – She felt she deserved a new Barbie in exchange for her “Buckle DNA” • We made it “worst case” by keeping Mom and Dad equal contributors • When either Mom or Dad is a minor, it makes it easier to determine if Daughter is “really there”

The Hardest Mixture I Know • Because there are so many “not excluded” options, we have to do a further test than just “viable” at our normal validated PHR of 50% • We will instead test at 80% PHR • In real life, you would do this in iterative steps and watch what happens for each hypothesis maybe all the way from 50 -90% • For this hand out, it’ll just be 80% and 90%

The Hardest Mixture I Know • First we test to see if all three people have relevant alleles present to check if Mom or Dad is low level or dropping out but daughter still all there • All concerns about <300 etc. are thrown out • This is testing what fits, not worrying about a potential missing allele from a random, unrelated person – these are all related people living together

Sheet Stain 1 • Mom and Dad at 80% PHR • Support at 11 loci (not counting Amel) But Amel can be useful Suspect varies 43 -69% 43 -59% if you ignore D 16 (16% window)

Sheet Stain 1 • Mom and Dad and Daughter at 80% PHR • Support at 9 loci Lots of “not excludeds” and “ 2 people sharing types” loci – won’t graph

Sheet Stain 1 Answer • • • Remember this is a mixture I made in the lab Hp – 11 of 15 loci show support (not bad) Hd – 9 of 15 loci show support (still OK) I’m going “inconclusive” as to V’s status Not much difference • The mixture is really only Mom and Dad

Sheet Stain 2 • Mom and Dad at 80% PHR • Support at 9 loci Proportions not very consistent (Amel flips) Suspect varies 51 -71% (20% window)

Sheet Stain 2 • Mom and Dad and Daughter at 80% PHR • Support at 10 loci

Sheet Stain 2 • Mom and Dad at 90% PHR • Support at 4 loci (plus Amel flips Major/minor)

Sheet Stain 2 • Mom and Dad and Daughter at 90% PHR • Still support at 10 loci

Sheet Stain 2 Answer • Hp – – 11 of 15 loci show support at 80% PHR – 10 of 15 loci show support at 90% PHR • Hd – – 9 of 15 loci show support at 80% PHR – 4 of 15 loci show support at 90% PHR • I’m going with the Daughter “really there” • The mixture is Mom and Dad and Daughter with Daughter present at 10 ish% (10: 2)

Suspect Underwear • Mom and Dad at 80% PHR • Support at 6 loci Major/minor flips all over the place!! Suspect varies 40 -68% (28% window)

Suspect Underwear • Mom and Dad and Daughter at 80% PHR • Support at 14 loci Do we even need to bother checking at 90%? ?

Suspect Underwear • Mom and Dad at 90% PHR • Support at 4 loci Look at that cool flip-flop! But, it’s not in the “correct for degradation” order

Suspect Underwear • Mom and Dad and Daughter at 90% PHR • Support at 12 loci

Suspect Underwear Answer • Hp – – 14 of 15 loci show support at 80% PHR – 12 of 15 loci show support at 90% PHR • Hd – – 6 of 15 loci show support at 80% PHR – 4 of 15 loci show support at 90% PHR • Is there really a question? • The mixture is Mom and Dad and Daughter with Daughter present at 33 ish% (10: 10)