myers~i·~/il{ - epic › state-policy › foia › dna-software › epic...2002/02/16 · 2.87...

STRmix VI.0.7.49 and V2.0 TrueAllele Casework

Interpretation of Complex DNA Mixtures PART II: 3-Person Mixtures

December 30, 20 14

o Unexpected instances of TrueAllele giving identical LRs occurred as well. In these cases, contributors and/or mixtures that assigned weight to mUltiple genotypes still had identical LRs at the TrueAllele limit of nine decimal places . This would be consistent with the MCMC process having identical seeds for both interpretations. In some instances (e.g., all 1.2a and 1.2b FRjoint interpretations), the duplicate requests were sequential. However, some of the duplications appeared in runs that were performed at different times.

o 32 pairs were identical for all three contributors. Examples of th is were observed for all three studies (FR, Rl , and SC.)

o 5 of the pairs identical for all three contributors involved requests that were uploaded to the TrueAllele server on different days

• FR requests 1.2a and 1.2b were uploaded concurrently on 8/20/2013. Duplicates were adjacent to each other in the sample order.

• FR requests 1.2c were uploaded on 12/30/2013. • FR requests 1.2d were uploaded on 1/212014. • Note: Multiple other sets of requests were uploaded between sets 1.2c and 1.2d.

The 0.75ng amp of FR_ 1-1-1 had identical LR for 1.2a[1.2b and 1.2c. The 0.375ng amp ofFR_ 1-1-1 had identical LR for 1.2a[1.2b and 1.2c and 1.2d. The 0.375ng amp ofFR_ 4.5-4.5-1 had identical LR for 1.2a[1.2b and I.2c.

• A closer examination of the STRmix FR(3) study results with the highestlliogLRH results. o This part of the FR study had three interpretations performed for each amp separately as well as

for the joint amplification. o In each case, one of the interpretations acts as an outlier, with the other two log LRH being within

- 2X of each other. o Note : The FR_6-3-1 _0.375 comparison that had the divergent log LRH was the interpretation that

appears overall closest to the expected mixture proportions. The other two interpretations were closer to 7-1.5-1.5 than 6-3- 1.

Mix Ampng Donor Amp, lnterp 6-3-1 1.5 2M 2a

2b 2c

Comparisons 2a - 2b 2a - 2c 2b - 2c

Mix Ampng Donor Amp, Interp 6-3-1 1.5 3F 2a

2b 2c

Comparisons 2a - 2b 2a - 2c 2b - 2c

Mix Ampng Donor Amp, lnterp 6-3-1 0.375 1M la

Ib Ic

Comparisons la - Ib la - Ic I b - Ic

Steven P Myers~I·~/IL{

10gLRH 16. 13 15.98 13.34

tJ.logLRH 0. 15 2.79 2.64

10gLRH 12.17 12.02 9.3 1

tJ.logLRH 0.16 2.87 2.7 1

10gLRH 16. 15 16.5 1 14.49

tJ.logLRH 0.36 1.66 2.02

Donor I 58% 58% 58%

Factor X 1.40

615.50 440.22

Donor I 58% 58% 58%

Factor X 1.44

740.45 513.57

Donor 1 68% 68% 51%

Factor X 2.29

45.72 104.73

Donor 2 31% 31% 31%

Donor 2 31% 31% 31%

Donor 2 16% 16% 33%

Donor 3 11 % 11 % 11%

Donor 3 11% 11 % 11 %

Donor 3 16% 16% 16%

Page 26 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000001

STRmix V 1.0.7.49 and V2.0 TrueAliele Casework

Mix Ampng 8-1-1 1.5


Donor Amp, Interp 10gLRH Donor I 2M 1.2a 4.95 76%

1.2b 6.47 76% 1.2c 6.17 76%

Comparisons Lllog LRH Factor X 1.2a - 1.2b 1.52 32.91 1.2a - 1.2c 1.22 16.44 1.2b - 1.2c 0.30 2.00

STRmix: Reaction to ~rofile ~eaks that were not ~ro~erl~ edited

Donor 2 13% 13% 13%

• During the analysis of the 3PM input files, improperly edited results were detected. o See folder "STRmix reaction to artifacts"

• Ex. I: "OL" allele call o See folder "3PM_SC_8-1-1 _ 1.5 - amp I with OL" o Input file: 3PM_SCI _STRmix.txt

• Sample name: 3PM_8-1-1-(1.5-ng)-(I) • 8- I - I mixture, 1.5 ng amp I

• Issue: I didn't edit out an OL in vWA. The OL was THOI pull-up. o STRmix runs

• 3PM_SC_8-1-1 _ 1.5_ 1.2a and 3PM_SC_8-1-IJ5_ 1.2b

December 30, 2014

Donor 3 11% 11 % 10%

• Joint interpretations of 3PM _ 8-1-1-( 1.5-ng)-( I) and 3PM _ 8-1-1-(1 .5-ng)-(2) • Run folders (respectively):

o 3PM-SC-STRmix-2013-II-13-21-32-18 o 3PM-SC-STRmix-2013-11-14-06-08-41

• STRmix performed a full interpretation. • Looking at the results, though, it's seen that amp I 's alleles were not

incorporated starting at vWA • CSV file created by STRmix: 3PM_8-1-1-(1.5-ng)-(I).csv

o 0 for allele Height and Size at locus II (vWA). o OL call not listed. o No loci after that

• CSV file created by STRmix: 3PM_8-1-1-(1.5-ng)-(2).csv o Alleles properly listed for all loci

• 3PM SC 8-1-1 1.5 la - - --• Single interpretation of 3 PM _ 8-1-1-( 1.5-ng)-(l ) • Run folder: 3PM-SC-STRmix-2013-ll-14-06-14-16 • No iterations were perfonned. • Results were essentially blank/aborted. • CSV file created by STRmix: 3PM_8-1-1-(1.5-ng)-(l ).csv

o 0 for allele Height and Size at locus II (vWA). o OL call not listed. o No loci after that

o New input file: 3PM_SC I_STRmix (8-1 -1 corr.).txt • OL deleted

• Ex. 2: Undetected pull-up/down and "<" allele designation o Seefolder "3 PM BK 4.5-4.5-1 1.5 - Off Lad and PU-PD" - - -o Input file: 3PM_BKsC\ _STRmix.txt

• Sample name: Amp2_ 4.5-4.5-1 _ 1.5 • 4.5-4.5-1 mixture, 1.5 ng amp 2

Note: Because this set of mixtures was not originally amplified in duplicate, a second set of amplifications were performed. These were labeled with "amp 2 and "amp 3" . For the purposes of th is mixture study, the BK "amp 2" was

· \ "\ ~ll Steven P. Myers rz,l~O 7 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000002

STRmix V 1.0.7.49 and V2.0 TrueAllele Casework

Interpretation of Complex DNA MixturesPART II: 3-Person Mixtures

December 30, 2014

treated as amp I when comparing to other studies, while the BK "amp 3" was being treated as amp 2.)

• Issues:

-..---..........-.-.. ,. I-

-..----"-"'----.. 1_ I_ pi.

1-- 1-' .. -_ .......... _--.. ,. I

."'"

.. ......-... · ............... .-.. 1. I_

·

o A D 13S317 stutter peak labeled "<8" was incorporated into that donor's interpreted genotype. STRmix requires allele designations to be numbers with no additional symbols. This should have been converted to a "7".

o A D8S 1179 "9" peak appears to be a PUIPD peak that overlaps the leading edge of the associated D3S1358 " 16" allele. NOTE: The "9" was observed in both amps and centered in the bin. However, additional low-level peaks in other colors display that this sample was having spectral issues of this sort.

Wi

' !""

/\ D C\/\ ~ I l1li1 .

:1 '~ id )~ c" z \ I \ 2\S ~ " 2\ 1 \ "" ill 19 B E3 Ej' -

- ...... ---.. ~ mill!! ' =" .. · ,. .

i \ ~I [j f\... ) l ~ 1 ~ ~ II '" " '" ['.,

1B EI 1'3 _ .. - ........ _ ...... ·1. I_

n.1

~I · .:] \

,.

HJ \ hJ\ • 1 ~ 4 1 ~ c ~ '" '" .00

ill ~ 1'3 8 B B 8 :jJ

Steven P. Myer~oll\.{ Page 28 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000003


o STRmix runs


December 30, 2014

• "<": 3PM_BKscl_ 4.5-4.5-1 1.5 la and 3PM BKscl_ 4.5-4.5-1 1.5 Ib • Single interpretations Amp2_ 4.5-4.5-1_1.5 • Run folders (respectively):

o 3PM-BKsc I-STRmix-Alt-Var-20 14-02-13-07-29-56 o 3PM-BKsc I-STRmix-Alt-Var-20 14-02-13-09-23-I I

• Interpretations ceased at D 13S317 • CSV fil e created by STRmix: Amp2 _ 4.5-4.5-1 _ 1.5.csv

o Nothing was listed for DI3 or any loci after that. • PU/PD: 3PM BKscl 4.5-4.5-1 1.5 la - -

• Single interpretation Amp2_4.5-4.5-1_1.5 • Run folder: 3PM-BKsc I-STRmix-Alt-Var-20 14-02-14-12-37-03

Note: This run was performed after the "<" designation was corrected for D13, but before the Pu/PD peak at D8 was noticed. The PUIPD issue was, however, also observed in the two run folders listed above for the "<" issue.

• STRmix performed a full interpretation. • All possible minor contributor genotypes at D8S 1179 were incorrect. They all

included a "9" allele, which this person does not have. • LR> I was obtained for the two major contributors, but the minor contributor

was LR ~ O. o Files: 3PM BKscl STRmix 4.5-4.5-1 1.5 la IF NC.xls - - - ----

3PM BKscl STRmix 4.5-4.5-1 1.5 la 2M NC.xls - - - - ---

3PM_BKsc 1_ STRmix _4.5-4.5-1_1.5 _la_3M _ NC (O).xls • CSV file created by STRmix: Amp2 _ 4.5-4.5-1 _ 1.5.csv

o All loci have results. o The 54 rfu "9" PU/PD peak was included for D8S 1179. There is a 65

rfu "10" peak, but that's too low to allow the "9" to be stutter. Hence why the "9" was identified as allelic by STRmix.

o New input file: 3PM_BKscl(ammendedLSTRmix.txt • Converted the DI3S317 "<8" stutter peak label to "7". • Deleted the Pu/PD peak in D8S1179.

• Ex. 3: Undetected pull-up o See folder "3PM FR 6-3-1 1.5 - PU" o Input file: 3PM JR _ STRmix.txt

• Sample name: 1.5_A2_6-3-1 • 6-3-1 mixture, 1.5 ng amp 2

• Issue: A CSFIPO " 11.2" peak appears to be a PU peak overlapping the associated D2S 1338 "20" allele.

o STRmix run • Run 3PMJR_6-3-1 1.5 1.2a

• loint interpretation of 1.5_AI_6-3-1 and 1.5~2_6-3-1 • Run folder: 3PM-FR-STRmix-Alt-Var-2014-01-24-04-34-2I • STRmix performed a full interpretation. • All possible minor contributor genotypes at CSFIPO were incorrect. They all

included a " 11.2" allele, which this person does not have. • LR > I was obtained for the two major contributors, but the minor contributor

was LR ~ o. Looking locus-by-Iocus, only CSFIPO was LR ~ O. All other loci were LR > I.

o Files: 3PM FR STRmix 6-3-1 1.5 1.2a 1M NC.xls -- - ----

3PMJR_STRmix _6-3-1 _1 .5 _1.2a _2M _ NC.xls 3PMJR_STRmix_6-3- U .5_ 1.2a_3F _NC (O).xls

• CSV file created by STRmix: 1.5_AI _6-3-l.csv

Steven P Myers ~~O\\~ \..z, I '

o All loci have proper results.




December 30, 2014

• CSV file created by STRmix: 1.5_A2_6-3-l.csv o All loci have results. o The 64 rfu "11.2" PU peak was included for CSFIPO. There are no

other microvariant "#.2" alleles or peaks. Hence why the" 11.2" was identified as allelic by STRmix.

o New input file: 3PM_FRammended_STRmix.txt • Deleted the PU peak in CSF I PO.

• Ex. 4: Undetected pull-up o See folder "3PM FR 4.5-4.5-1 1.5 - PU" - - -o Input file: 3PM_FRammended_STRmix.txt

o

• Sample name: 1.5_ A 1_4.5-4.5-1 • 4.5-4.5-1 mixture, 1.5 ng amp I

• Issue: A 68 rfu D21SI1 "35" peak appears to be a PU peak overlapping the associated 4119 rfu DI3S317 "I I" allele.

STRmix run • Run 3PM FR 4.5-4.5-1 1.5 1.2a -

• Joint interpretation of 1.5 _A 1_4.5-4.5-1 and 1.5 _A2_ 4.5-4.5-1

• Run folder: 3PM-FR-STRmix-Alt-Var-20 14-02-03-1 0-06-02 • STRmix performed a full interpretation. • The minor contributor was assigned the incorrect genotype 32.2,35 for all

genotype combinations at D21 S II. That person does not have that genotype. • LR > I was obtained for the two major contributors, but the minor contributor

was LR ~ O. Looking locus-by-locus, only D21SI1 was LR ~ O. All other loci were LR> 1.

o Files: 3PM]R_STRmix_4.5-4.5-1_1.5_1.2a_IM_NC.xls 3PM FR STRmix 4.5-4.5-1 1.5 1.2a 2M NC.xls

-- - ----3PM]R _ STRmix _ 4.5-45-1 _ 1.5 _ 1.2a_3F _NC (O).xls

• CSV file created by STRmix: 1.5_AI _4.5-4.5-l.csv o All loci have results. o The 68 rfu "35" PU peak was included for D21 S II. There are no other

peaks within 4 bases. Hence why the "35" was identified as allelic by STRmix.

• CSV file created by STRmix: 1.5_A2_ 4.5-4.5-l.csv o All loci have proper results.

o New input file: 3PM_FRammend2_STRmix.txt • Deleted the PU peak in D21 S II.

Discussion and Conclusions

Summary: Tested in a comparable manner, STRmix had better sensitivity and precision than TrueAllele Casework with these 3-person mixtures.

Sensitivity was measured as the proportion of comparisons (spanning allmixture-to-contributor comparisons) that gave + log LRH for true contributors.

• Not surprisingly, sensitivity and LRs tend to go down with lower amounts of template DNA, when comparing to a minor contributor, and when interpreting more even mixtures (e.g., I-I-I). In this study, TrueAllele had the advantage of being able to use results as low as the 10 rfu analytical threshold, whereas the STRmix threshold was 50 rfu. This might explain why the 0.375 ng 1-1-1 graphs displayed some higher LRs for TrueAllele than STRmix. Future testing of STRmix using reduced analytical thresholds could lead to increased performance, but drop-in parameters would need to be established.

• STRmix can give LR « 1.0 or even 0 for true contributors. While a locus LR for TrueAllele can never fall below 0.0 I, this sometimes masks that the system didn 't assign any probability to the contributor ' s genotype. While the TrueAllele detailed report might show that the genotype was not included, the other tabular reports created by the software would not.

- ~ (11.-\ Steven P. Myers~-t \ 4j) Page 30 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000005


[nterpretation of Complex DNA Mixtures PART II: 3-Person Mixtures

December 30, 2014

• Overall, STRmix had a high degree of sensitivity with + 10gLRH in 96.59% of the comparisons. 0.55% out of the 3.4 1 % of -log LRHcomparisons were the result of complete fal se exclusions (LR = 0). These false exclusions, however, were solely the result of the software exceeding the Java cap on iterations. When rerun in a manner that kept the iterations below the cap, all of the LR = 0 comparisons became LR » 0 comparisons. (This information can be found in the "Sample Entry Order" section of the STRmix validation.) Of the remaining 2.86% of comparisons with 0 < LR < 1.0, all occurred with 0.3 75 ng amplifications, and seemed to be the result of overlapping issues: poor estimates by STRmix of the mixture proportions; few or no loci where more than 2(N - I) alleles were detected; and/or multiple donor alleles that fe ll below the analytical threshold (see especially SC_I-I-I _0.375 amp I compared to Donor I .) Care should be taken when interpreting such mixtures, especially if most/all of the indicators that they consist of 3 people fall below the analytical threshold. In such cases, jointly interpreting replicate amplifications could prove helpful to correct for this, especially in regard to mixture proportions estimates.

• Overall, TrueAliele Casework had + log LRH in 94.88% of the comparisons, which is similar to the sensitivity of STRmix. However, when the locus minimum LR was bypassed (Le., locus LR < 0.0 I allowed), the sensitivity decreased to 90.79% with 5.56% of the comparisons giving LR = 0 for ~ I locus.

Precision was measured as the proportion of pairwise comparisons that were in the ranges of 0 - 0.3 and 0 - 1.0 log LRH units. This corresponds to LRs within 2X and lOX of each other, respectively.

• Overall, STRmix had 83.85% within 0.3 log units, and 96.87% of the pairwise comparisons within one log unit. When the pairs with at least one LR = 0 result are removed from consideration, the maximum difference was 2.87 log units, which corresponds to a factor of - 740. As observed in the graphs, these largest deviations occurred in pairs with a minimum log LRH > 7 (LR > 10 million). LRs of 10 million and 10 billion are likely to lead to the same conclusions about the strength of the evidence. Below this level, deviations ranged up to - I OOX, which could possibly lead to moderately different conclusions. The precision results for the FR(3) study would support any of the following strategies:

I. Always perform three interpretations and rep0l1 the LR that falls in the middle of the range; or 2. Perform two interpretations. If the LR fall within a factor of # (value to be set in the protocol),

report the lower LR. If, however, they diverge by more than a factor of #, perform a third interpretation and report a. the LR that fall s in the middle of the range; or b. the lower LR of the two that are more similar.

• Overall, TrueAliele Casework had 41.67% within 0. 3 log units, and 70.09% of the pairwise comparisons within one log unit. These values are elevated somewhat by the pairs of interpretations that obtained identical LRs that should not have occurred given the randomness of the MCMC process. When the pairs with at least one LR = 0 result are removed from consideration, the maximum difference was 15.8 log units, which corresponds to a factor of - 6.5 quadrillion. As observed in the graphs, many deviations were observed that could lead to different conclusions about the strength of the evidence (e.g. , LR « I in one interpretation becoming LR » I in another). It is acknowledged that some of these results may be due to interpretations where the MCMC process didn ' t explore the space well, or where the chains had not reasonably converged. A more in-depth examination of the Mx chains and genotype weights might have el iminated some runs from comparison. However, this highlights that more subjective evaluations and more computer interpretation time (3-4 days/interpretation for TrueAliele vs. < I day/per interpretation for STRmix) would be required for TrueAliele than for STRmix.

Artifacts and non-numerical peak labels in the sample profiles are, as noted by ESR, a significant issue for STRmix. The profiles must be thoroughly edited prior to import, or the results could be ...

• fully or partially aborted runs; or • incorrect genotype assignments that might lead to false exclusions and less likely false inclusions. This

appears to be the biggest concern for minor contributors, as their allelic peaks are more likely to be in the rfu range of pull-up peaks.

TrueAliele would not suffer from some of these issues in the same manner. • TrueAliele has its own process for identifying and labeling peaks; and • It has a locus LR threshold of 0.0 I that would prevent complete false exclusions (LR = 0).

However, TrueAliele interpretations can still be skewed by artifacts, and the system has a mechanism for the manual removal of peaks or loci.

~D\\'-\ Steven P. Myers /~A\\1-t,.

\--t Page 3 1 of47

epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000006

STRmix V1.0.7.49 and V2.0 TrueAllele Casework

Appendix 1


Electropherograms for interpretations that were -log LRH in STRmix.

SC 1-1-1 0.375 VS. Donor I fIofdt_r .... __

PII:t~~PI>I.-.. 300 -11311_

-- I. '(1

l!!I'iI?t 'gill! i !!!1-

:~I ,. '" '" ~~ " ~.~ ,. '" _ m ~ m >It m ~ ~ M m M

mftllU·I.I.~_..<JICD"'lJ-",IIlJ"

lilIilffi

~I ,. .., '. '. '. ~i M-~l

''''.ilI

,@,""

i iUt ....

I@"

'~I ,. ,. ,. , .. , .. " 0 , " "" . 11 , .. no NIl I0Il ,.

!'IoUl_r .... _HOb

Pli:is..v:jC;:;\"",,-.. 3l!!1 - r3 _--=w 1MCf..l .IIIilJ

-- ft "

@IM IIMI I wpn iltHWW

L :~· 1111 1M II, ,. ,., ... ...

mftlO~·' .. __ ..arcau."-'='I. I_ @I@ "_, il!!!#!!i i ..... U

nil '" '10 ... .., ... .., NIl

, .~ ,l ·~ r ~,. : . ,: .......... Q ...... .... etOtf·II..-u,;. ~

1IlJ· ... i 11M! !D0p ..

'~I ,. o. ,. ", ,. ,., ,. no ,. .'"

,~ ~

~\ '\'\ Steven P. Myers ~'" ",1JO \-z,

1m .....

December 30, 2014

1 !CRIM ! no ... .,1 >It ... .. 110 ..

, k:iPiP!! HiI ..... ... ..,

,~

i glll8 ... _ .. 0 alii ... __

!!IIlmo



.... tdt_T.,.. ..... HotI

PIot-'~PId-' 31!!J -fI3l "'-

-- O' III

.. JftUo.~.-.ctI,._._u ........ ~ ",11M] , riA


~ e ~ ~ , . W .

i ''M i wLiA!

December 30, 2014

t A -

.. .• .• • •• ~ m •• • • M .. • m .. •• m .. ~ m -~A¢~ ~ ~ .!~&L~ :~,l,--""'"'"""I'''~''",,-.• '' -" -" """"il" ~-~~~~~-..........

,- IN ."

" ... ~Q .. __ ., ......... _u_ ....... ..-

,bliM i EWA i 11M i .,DU'

'~lL __ " ___ "'~#."~'_' __ " ___ " ___ " __ '_'~.~~'~" ___ • ___ •• __ m __ ~~~~.~~_m __ M ___ •• _______ m#~~_~~~._. __ .. ___ .. ___ - __ -___ ---"

~ (UIS .... "' -!II!

... f<It_r .............

~- ....... ~- 3 I!!! II -1"'""3!! ...... ~!IiJ:r..! ! JIII1l1J 1ll. !~!.1 .. ~!! .. 1-- 1'" I'll "JI"V·~"'~l""'-U __ " . •

iIiIKl pWD '" ,.. .to '. .to ,. 11"

+ ttlJI"V'!.,. ___ .GlmtUu._r--rw-

IIIIrJ "'"Ii ...

... .. _--.QCl«._..__u._... ~

(!'g] ibiD" ...

(1( 310.81 '1' ·14/

A~ '\,-\ Steven P. Myers __ ,/'\. I~ \ t ,

i m m M m _ m _ m a m _ _ ~ ~ _

. "" . , ,

, , ,



SC 4.5-4.5-1 0.375 vs. Donor 3

~~UJ.a.MeOfI"~"" . 1_ M"'" · - . " .. t. t. I.

~I ~ ! ,1· ~ ;; ;;;

~.~IIhIHIUft __ .u •• ~

IIIIII!! · - ... .. .. ~I • ~ . ,~, t ~ :

.. --..Q ........ .....,...,.~~IA r-liI 1I:!lI1II2I!

'~r - '"~ .. • .. ..

1 ·1 IX:tIUI Y .•

"'" ~ _ TaoiI-' "*'

1-" j" I'" ....,...,........., .. _UOV--.U.JIII • • ·IM I · - ." ." .• .. ..

:~I 0 - ~~ ~

................ I .. ~.-..u ... . • II!UIII I • - .- . - .. .. .. -." .

I II k . BI"jB B

.. .PllQ .... _ ............. u."..,..~ .. • • mill I • - m .. M .. ..

'~I ~ .!J

-

-

ftM' m

0

I!I!!!! M

a_I M

Interpretation of Comp lex DNA Mixtures PART II: 3-Person Mixtures

'M'"' 'MM .- - m m - w - -,~~ M~~ " I ..

1111111 1m .. .- - .. -l~

- - - - m -~. ~

IMIIII 1l1l'i. M - - - - - w - - m -

.~ .~ ~~

.pmlt 'WA .- m .. - - - - w - - --~, 1 .~ b" '"

IIIIImI! II!HR .- --- - m - --- --! J EJ " , " .n - ~ .. '" , ..

.~

ItmIIl:! I .. .. - M - - - - - - - m -! ~ ~ M

~ '\\~ . -1P Steven P. Myers \"t- \

December 30. 20 14

i "IN I - - - • - - . - •

~ I · . I !CRB I - -- • - -- - ~

~ :.:.; I

I s:aB -- - -1 o~ ....

.8

.oJ I Mil. I - -

~~~ - - - - ..

A I

I Bllm I - -- - - - • - ~

~~h A ... . n '. ~

I .. III I - - - . - - • - -~ ~ I

.



"'"1!dt_T"'-'~

!'IciSollhg:t:='"",se., ilE!!l -1'31"-

wLfRd i rii


i tI!i4 'IMP'

December 30, 2014

tOO ,to tOO tOO ,. ' "

~~ ,-,~l ,~'-·n: ......... I .. .. .. .. ..

_1

IJ"AA ' iW4 i IlNj t. '" ,It ,. ,. t" - '" '. '. M ~ ~ - ~

~M~ 1 ~ W@l i .... 4 i 11M

'. ' " . It t .. ' .. tOO , . .. ,. ,. M .. m - o.

~ - ~ ~ rnU.'IV"""

Aofdl_r .......... "'"

P'd~jC;W01i l'lci-' 31!D -r3...... II!Cr _11 m.;) p-a

!III;l PdD

i ItQ ......... ".

C t st

i M '" ..........

a.. :".e e c

i m

·MW-

i II8IP' • - m - - - ., . - - -

at_~ i@iDl - - .. - - - 0' M . - -

M

.. .. m .. .. .. ... .. .. .. ..

t1 •

.. m .. .. ... ... .. .. .. ..

I -

-

III

-1 t ' J

,. t .... no -~ -,~ --. -------. -1

Steven P. Myer " ~\~o)\ '-\ ~-zJ


STRmix VI .0.7.49 and V2.0 TrueAlIele Casework

SC 6-3-1 0.375 vs. Donor 3

,. :: .

,. ,

" I I

""[of;_t ... ,...~

",.-.1(=> ..... -. 31!!1 -14-:31 ..

...... .. III

ilMI ,. '.

@i'_ i 11M' •• ... ,. ,.. ,. tOO ...

~_o\\V\ Steven P. Myers ,tV'~1-r' ,

Interpretation of Complex DNA Mixtures ~ PART II: 3-Person Mixtures

December 30, 2014

* •

III j I!

i I pM I !!:P'N

PilI! ,p- i !tP'N I ,. ,. N - N - • ,. - - m - - - - - - - - -,~ .~ ~ ~ ,M ::

• Mum 'M. i gllg

-~: -T~ - --~~w - --~~- . -i l!!lijll ill"g , Dllp

1I., ...... ~tlJlata.

_ _ m _ _ _ m ~ _ M _ _

!U ~.~- !.~ ,~, I I!!@" iMp I Mil-,. . N - ., N • , . - - ., - - -

i ~ ~ M ~ ~ .8


STRmix V I.0.7.49 and V2.0 TrueAllele Casework

... (dt _ T .......... .....

"""-..IC=* ..... -. ill! -1i""'""'31..-


December 30, 2014

"iM "Drip i rii i •• , i

:~t--..I - ' .~~ _" ,. _" - ~;' .. I~. ,. -----'Ilr-i '-/,\--~ ~ _~ m ~~\~~~'A1 M ____ - -_---..J- -

....-. ... ,.~.tIIOftUI.u,....~ O!f@l \ &iii i tIM

... CdI;_T ......... t\otI

"""-.IC*MM'\",,,-. :::II!!! -~ ...... -OIl I!Cf IIilLJ 1!1

,. 'II

..........,...I._IJIOI~U_-.u . • (l1li;] gHp ... ,.. t.

" i~' . • +

........,...I.~ .. _-.-.U __ • • (III;] PMP "" .. 100 " .

.,r ... ..

_.)ft(.an.,._ .• _u-u.-..._ ~ !mid .... 11 i N

'111 H I ' !II ... '.

~ ~~ M




December 30, 20 14

SC 8-1-10.375 vs. Donor 2

~_ ...... ~_ 3111 11 -.----3l1.~III!Cf ... I_ J.J 1Il.1I~1I1lIJ.al1i 1l " 1-- 1-' lOCI .............. ,.-.ovot.-lI~ . •

ctUI'IIV-l.I. __ ...... U-'"""~

Pillff , . II.

• !!III!! - ,. M N m

1LI

"'fdt_r .... __

~_ ...... ~_ 3111 11 -.----311_ 'M ''' III!Cf..! I_ 1J 1Il. '.!!J!!!!!IIIJ.al1ill " 1-- .. It(! -u>IC.. .. ,~ .. _Utt.".....II __ • I_

@'QH I., , A'@II

N

:~f ,. ,. ,. ,. ,.

'" , . ,. ,. ,. M N m N

~~. ~ tMJIIO-.. a.. ___ .u.".....~

lDiI_

"'~~_ ........... II--... ~

e !

11M' ,.. . .

A •

gu... I '!W' • _ N

. ~ij ~oO~! 'lDQiJli

... .. , ... too 110 M

. ~! .. ~ !@gll

•• .. M

i SAiN I M .. tI, M _ M

-

I ptM • "¥fib! I M N -

~: N - "

;1-

- - - -

i-Hi , .,. - N - m _ N - "' M - - --

!.~~ :::r; .~~ 1M..... i Dill-

~ ~ .. ~ M M .. ~ • • _ .. ..

L~ .~. ! 0"·

,. .. " . .. .........

'~!'-------"#;-.~ --M;M~~ '~!/;----------/l~ I,,\,--~ ---I.W;-~ -(XliV! V·UIIi

StevenP. Myers ~\4F\\~ \.-z,



Win... i rii If ,10 " . ,. ' 10 ,. ,.. 'If

:~I

ctVfIIV-l'I.~"_"_.II-",",~ pita i !ti!4

Interpretation of Complex DNA Mixtures PART II: 3·Person Mixtures

i dfOI 'M" ~ M ~ ~ ~ • m -

~~ ~ i AX ! "liP' - . I. ~ - - - -

• -'ti - -

~ ~ "'~ ........ 0 ... __ .....-..-.. ,.... ... ~

!!!I!idJ , ..,., .. , ........ ..

~~ 3111 11 III. II 101

.1

I. ~ I. I

: ·

A • • •

I-I. ""!':' .• ,

I.

': A • .A

I. ""!'! .• I , ,. · - ! I

~ tI El .!l

~o'\\i\ Steven P Myers /~-i\'?1

I

i 11M ' ....

.. . -

I. _

. . " .

.k!

ffi

M

AI ~El

December 30, 2014

iii

. , • • M - .

I .. - - M - .

I

~

I

I

I

.


STRmix V1.0.7.49 and V2 .0 TrueAllele Casework

RI 6-3-1 0.375 vs. Donor 3

~ ~ -ra ...... ,..

"""~~I'\:II-';' illl!lll -I"'illl_ 1lI.1I .'iliiJi" 1I (;j~ .. _. .. " .. _<_ •• .}IV< ___ ........... ~..ru ....... • T_

PIll1f ,. m '" .. .• .. .• -• -'. I A · :~[:l] "

__ ._ ................ _o:.1t_oa4J ....... ..ru ...... .t~

IlIm~' .. - .. .• .. , . , . --- I ~ - ! · " " •

1., (lC"U7V.'1I5l

... ldI;_T ......... "'"

..

..

Interpretation of Complex DNA Mixtures PART II : 3-Person Mixtures

,- 1OCI..I lliliilJ

'g,,1! I MM , . .. , . ~ .. ~ ~ ~ ~ ~ '" •

, I! . ~Fld " .• . . ~

11111111 I I!Z!IDI .. .. .. m .. ~ • , . ~ . ,. -

I I ~ ~ ~ '. ~l:,~,

~-=~- illl!l ll -1"'il11_~IIOCI..! l liIiilJ 1lI.1 1 .'iliiJi" 11 (;j~II " 1'-- I" i'Q

-@iM I OM' i wptt IUU ,. ,. .• .. .. .. .. .. , . .. ~ o • m .. , . m - '" -

I ! I ! I ! I ! · ~~ 'q!;! '~ , ~ 1114 , ~

@ ..... I ifl!8! j !PPllIi I"'D ..... . .. .. .. - , . .. .• .• ,. .. m O. m .. , . m • .. •

--.. \ A I ! ! I • I , · ... " " ,~l ±J 1111 .~ 33E

, ,

d'i ,\~ Steven P. Myers 'l~tl-;O

December 30, 20 14

ml_

'" I WB I

• m '" g ~ M * •

! ~ ,.

I !!:IIB I ~ . .. m ~ M - -

B " . .

I Mgw I - - .. m - M * -

j I ! , BeJ:fJ ~

I @1U1i I • - O. m - M - -

\ ! A " if] . .. - ~



f'Io[of;_r .... __

P'd-.a~P'd--.. 3m -Ii"""'"'"3 "-

.... ... .JOU. ___ nNJQ4I ...... .J......... • w"M W

Interpretation ofCompJex DNA MixturesPART II: 3-Person Mixtures

- 'Mp. .. - ' W

II!I!BD I I!!!i 111m I_I - ,. .. .• .. ." - ,. '" .. - m m - ~ - - - -

· 1 ! .l It ,A J

-

lA .. ~, ~ B " fI i:I' :B EffI

.... ~_T .... __

1'IaI-..jC;;:;;t .... -.. ilBll _ 1i""'""3 .....

... cJOl.JIIV·.JIIII_ ... .-.. ...... .JU ...... ·.·iir,,·=======:::J ~ O'P ... - .. . - " .. .. - ,. .. - m m - - - - -

ill ! II! · , . :;'I~ ~ .ltffi " . - PR" I ... - - .. .• .• .. . " .. .. .• .. M .. m - .. - - m - -

· I H! , EJ" .. .

EE~ . '"

,.\ I "\ Steven P. Myers ~1-1f' \

December 30, 20 J 4

Jill

I - - - - - - •

III

- ..... - ..

I • - .. .. - - •

Page 4 J of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000016



December 30, 2014

SC 6-3-1 0.375 vs. Donor 3

... [.t;_roohAloloo~

1"<C __ ~PId-' 3m! -14"3 . Iii. 1!!IiiJi' IIliJoWiI .. -- a. , ..

"',,11' iUIili i pM i i<:itlbl i

~I,---" '" '~. :~,~~ ~--,W~.-~ ,_w ~ ~-~.M,l,----,,;--:1~:,~ -_* -... ~nl:lP..u ... .Jao;QlOf_~.~

!!!111m

"-"EUmUn"'Jf'I_OIfJ*ClQ".,NMIJ'~ "lihi

1'Io~_r"""'_ ...

'1*'111 i M. i !¢PIN I

'''' .. Ii i 1!t1 .... i lei",..,

- . -Pllis..v!C=\JIO:CS-UO :!1H!J 11 -r3II_~IIi!Cl~ I_1J

--@iIl" i liM' i @MI! "ONW ' pigs , " ~ ~ , M ~ ~ m _ m ~ ~ _ m m M W ~ _ m _ m a m m _ M _ _

~n""..u..-oarcKQ'l~·I. I_ WIM

... ~nal'....,._ftf..l""""-"'~ IIIma

~I , . ... '" '. ,.

~ .~ w

i ii.'

I I!II!!!I ,. ,.

i PP""

Ulf - l~-liI!IBIl' ,. ,. ,. ~ N m - w

! ~.~

. -

II!IID I 1111111 - - m - - . .. m --- -M .~

I

J

I .8

A71'"~ol \"1 Steven P. Myers ? ,-z,,1 ~ Page 42 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000017

STRmix V 1.0.7.49 and V2.0 TrueAlIele Casework


December 30, 2014

'!PM i &1U

r.jili I!!"AU i !?WI

,. .oo .. .. ," -

~- ...... ~- :£1l1li 11 -I'"3!1 I1LJ11~ III!CI..! I IIII/lLl 1Il.!I .!!l!!!!!.IIlJ~II " i-- .. I" MV,"_Qnr~ ... ____ ..-.. . I_

i ItNX

i "I

,-,

~.- -- ~--J. ' .. '

I ...........

eM

...........

• liM! i ... ... m ... ... ... m ... ... .. ... ...

IiIml p., I '" i -----=r ---.- --~ ,~ ----

-vn ........... .P . .....-......-... . •

IIml PWit

..... ll'timUn ......... n • .J_II_..,.,.~

IiIII.1 PdP i 1M .11 ,. ,.. ,.. ... ,,. '" m 100 , ..

~~o \'1 Steven P. Myers"""-/I--t I Page 43 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000018


Interpretation of Complex DNA Mixtures PART ll: 3-Person Mixtures

December 30, 2014

SC 4.5-4.5-1 0.375 vs. Donor 3

No',._T ......... ~ JIO:I;Sooehg:jCOOOWIrtPlol-"" 31!!1 -~ I _

_ ..-umaaego...,_ru.J ......... -"'N,.~ "Iiilft , .. 1111 I Iff... , !mlN

-..,n_n ......... _a • .,lGIIrD .. -'*'"'~

pnLl. 'pn" i I!!!@ i S:V!fU '

~-·· - · r -··

Nofdt_''''*_''' N SChg: iC=WCf\ """ -.. il l!!! -1""'"3 _ iii:!' IIII1lJJ

-- O. IQ

@II" mI' iil!Q¥!!i 'II". ! 0'•

~ -. -Iff. ~ r D "'r MA'· ~~ N * "' .~ • • ~i

... ..-umaD ... .,IIlI_eu-'_ .. .J*U,.~ Ii"'@ H'I! i 1!!QQII I .... HI i '",pi

_~n ... .,....n.j....ul" ___ '~ 1m".. !1Ii1 i w!JpIi i!l!'iAA i @I"",



:: .

Wi .. .• w . U

E1El [,j lX_m'l~

.... (dI;_TooIs ..... 1IIb

.

..... SoItt-Q: je;;WOrt Plat -. 3 E!I -,.---:;] 1 ___ 1ll. I!!llIWIIIll..wo:1 .. -- .. "


i 1M ' ... ' ~ - m • m m

. ~AI "~M ,

I ~ .1 A A

1E ,

111M i !!!PDI

A ~ E1~

,- EI IIIIC!LJ

IiHI!;] IIi-Y , ...

December 30, 2014

- - - .-

.oJ j .. • m ~ m m ..

,1 I £

• ,

I!Il

j

~ 1-.-. -i~: ... - - - m m - : m ~ : m I pall I '"

I~·~ · --·· · ·~ · M " n::" •• • It •

pall i ... .lI III ,., t OO tOO " . ItO

AfVY'- \\'-\ Steven P. Myers .......... f\ -V\1f' Page 45 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000020


FR 6-3-1 0.375 vs. Donor 3

"*lof;_r ........ ..-. "..-.o:~I'Ial-" :::II!!! -11""3 .

pm"


i P !11i i pM i ICRI& _ m m _ ~ _ _ ~ _ _ _ _

~- ...... ~_ 31!!1 11 - r3I1_ ,. ''' IJm..! I_JJ 1li. 1.!'l!!!!l11<~1I " 1-" !_. I'"

OOU .... .J.I..W-.... • • mll_ 111M' i _ Ii iun i g"g

December 30, 20 14

- - -

- ,. ~,---~:'~1 "-~ -~I ~.~~ m ~iJ -----,--l~~~r--,,;· -A~~,~r ;-'---', --

woo 1 11M' i ""'P" lfiWW I AilD - ,.

@il_ i awl i !!!QP!! ,un i DiM

Steven P. Myers ~¢) \ "" Page 46 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000021

STRmix Vl.O.7.49 and V2.0 TrueAllele Casework

Interpretation of Complex DNA Mixtures PART II : 3-Person Mixtures

December 30, 2014

.. [dIt_r.,.. ........ PIaI$oIt>J~!'I:II ___ ::Jlm ll -r-3I1..-- !'oIIll ll!C'f.J IJI&LJ

1Il.II !iliIIi' IIIl~!I " 1-" ," jllli

i iff. i tiM iliUM' -,l!· .....-

¢lit@] i -Wi

"

M>k I- I_ wli@ i "". i tlH!t i Wi! - '. '. - - .. ..

'~l ~~ ~ ~ "~_'"NoIoo''' PIaI~~PIaI-' 3E!!! _143 OJ ' l1li:[ _ Lf

.... .m~u.. • • IIIIi1 Rid!! 'M ,o. 100 .. , .1:11 lit ... _ . ..

"

".''''..AI.f."',," I_ 1_ 11m) PijII Iq -. · ,l~:

.....-P_" I ...

.. .. .. ~ ... .. M .. ..

.. III .........

...............

I m .. "'" .. rn .. .. .. .. ..

't . ..

m .. .. .. .. ... .. .. .. ..

1 •

.. .. m .. .. .. .. .. .. .. .. ..

Steven P MyerS~r~o \ \"\ Page 47 of 47 epic.org EPIC-16-02-02-CalDOJ-FOIA-20160219-Validation-Study-Part2 000022

myers~i·~/il{ - epic › state-policy › foia › dna-software › epic...2002/02/16 · 2.87...

Documents