NIST Inter Laboratory Study 2013

Recently Barry Scheck spoke at some length at a meeting on the New York State Commission on Forensic Science ^[1]. Amongst other matters he stated that a certain software was the only software that gets the NIST Mix 13 ^[2] exercise correct. We are a little disappointed at Mr Scheck’s statements at this important meeting. His statement about NIST Mix 13 is untrue.

We did not participate in MIX13 at the time that the study was undertaken. We have subsequently analysed the five cases within the study. The STRmix™ results are summarised in the table below.

Case		Likelihood ratio for person of interest ^[3]				Outcome
Case		a	b	c	d	Outcome
1		2.60x10¹⁹	NA	NA	NA	Correct
2		2.12x10¹⁶	465,000	822,000	2.35x10^-7	Correct
3		21,000	0	NA	NA	Correct
4		1.71x10¹⁸	NA	NA	NA	Correct
5	3 person solution	295,000	50,000	0	NA	Correct
5	4 person solution	1,400,000	1,000,000	0	NA	Correct

The outcome is considered ‘correct’ based on a high likelihood ratio for true contributors and a low one for non-contributors.

Case 5 was challenging. The mixture is actually a 1:1:1:1 mix of individuals a and b and two other individuals selected to maximise allelic overlap (the genotypes follow). 0.5ng of each was amplified (total 2ng).

	Suspect 05A	Suspect 05B	"not tested"	"not tested"
AMEL	X,Y	X,Y	X,Y	X,Y
CSF1PO	12,12	12,13	11,11	11,12
D13S317	11,13	8,8	11,12	11,11
D16S539	11,12	11,11	11,13	11,12
D18S51	13,15	15,15	13,15	17,17
D19S433	13,14	14,15	12,14	14,14
D21S11	30,31	30,31.2	31.2,32.2	31,31
D2S1338	17,23	17,23	18,20	18,23
D3S1358	16,17	16,16	16,17	17,17
D5S818	11,13	11,13	11,13	11,12
D7S820	8,10	9,10	10,10	8,11
D8S1179	10,15	14,14	10,10	12,14
FGA	22,24	21,21	21,21	21,22
TH01	7,7	7,7	6,9.3	6,9.3
TPOX	8,9	11,11	8,8	8,11
vWA	15,16	17,18	16,17	17,17

The epg follows.

epg

The allelic overlap and 1:1:1:1 ratio creates a perfectly viable three person solution. The comparison of the peak fits for the 3 and the actual 4 person solutions is shown below. The three person solution is a closer fit by least squares.

Locus	Allele	Observed heights (rfu)	Expected Three Donor Solution	Expected Four Donor Solution
D8S1179	9	57	69	69
	10	1008	1086	1082
	12	314	295	303
	13	67	65	65
	14	1105	1039	1043
	15	387	365	358

D21S11	29	62	46	46
	30	933	791	783
	31	937	1023	970
	31.2	690	688	749
	32.2	336	368	364

D7S820	8	626	532	532
	9	333	356	351
	10	1118	1138	1137
	11	212	230	236

CSF1PO	11	1000	1013	965
	12	1230	1155	1212
	13	314	316	310

D3S1358	15	115	125	125
	16	2085	2078	2072
	17	1801	1797	1804

TH01	6	1394	1328	1221
	7	2338	2328	2313
	9.3	1012	951	1073

D13S317	8	1223	1216	1210
	10	111	141	143
	11	2346	2249	2274
	12	614	655	649
	13	662	616	603

D16S539	10	141	180	173
	11	2904	2883	2784
	12	956	957	1068
	13	590	571	566

D2S1338	16	57	53	53
	17	1007	892	882
	18	816	849	773
	20	355	342	421
	22	78	75	76
	23	1102	1180	1184

D19S433	12	423	417	504
	13	653	630	610
	14	2358	2394	2237
	15	417	386	470

vWA	15	595	550	545
	16	1002	1015	1085
	17	1947	1971	1820
	18	421	398	484

TPOX	8	1889	1741	1666
	9	441	349	423
	11	1149	1203	1208

D18S51	12	56	57	56
	13	924	894	881
	14	125	113	112
	15	1800	1765	1759
	16	61	46	47
	17	630	721	741

D5S818	10	87	111	111
	11	1742	1770	1770
	12	486	460	470
	13	1400	1374	1363

FGA	20	119	120	120
	21	1995	1919	1927
	22	697	712	712
	24	359	394	386

Looking at the genotypes, if one considers, say, CSF1P0, the four contributors are 12,12 and 12,13 and 11,11 and 11,12. There is an equivalent solution as a 2:1:1 mix of 11,12 and 11,12 and 12,13. This is true for all loci.

STRmix requires the assignment of the number of contributors. With no prior information, we would assign case 5 as a three person mixture. STRmix™ gives mixture ratios 2:1:1 for this solution.

If we assume the presence of reference a, as some software programmes do, we force the four person solution in STRmix and obtain the mixture proportions 1:1:1:1. This is a valid assumption under H_p and we do not argue with the software programmes that do this. However it would be wrong to make this assumption under H_d and the most parsimonious H_d solution is three contributors. The only obvious way to decide that this is a four person mixture is to assume the presence of one of the persons of interest (POI): reference a, which we were not prepared to do.

However, whether assuming three or four contributors all POI have been correctly assigned valid likelihood ratios and included or excluded as contributors to the mixture correctly.

[1] June 19, 2015 meeting available here

http://www.criminaljustice.ny.gov/pio/openmeetings.htm (external link)(external link)

[2] http://www.cstl.nist.gov/strbase/interlab/MIX13.htm (external link)(external link)

[3] LRs were calculated using the NIST Caucasian allele frequencies

Back to the news

NIST Inter Laboratory Study 2013

Developer tools

Debugging

Cookies

Developer tools

Debugging