Mitochondrial DNA

1. Mitochondrial DNA • Observations on Frye/Daubert Issues Associated with Forensic mtDNA Typing. • What is mtDNA Typing? • Database and statistical issues • General unacceptance

2. FBI Laboratory- DNA Unit 2 Armed Forces DNA Identification Laboratory- AFDIL Regional Labs: New Jersey Connecticut Arizona Minnesota Mitotyping Technologies College Park, PA Bode Techology, Springfield, VA Labcorp, Research Triangle Park, NC SERI, Richmond, CA Reliagene, New Orleans Orchid Cellmark- Texas North American Forensic mtDNA Laboratories Government Labs Private Businesses

3. FBI Interpretation Protocol Although there are differences in mtDNA compared with nuclear DNA, forensic identity testing is fairly straightforward. Typically sequence concordance is assessed between reference and evidence mtDNA sequences. When heteroplasmy arises, careful analysis and direct comparisons between multiple reference samples and a questioned sample should, in most cases, alleviate interpretational differences. If the mtDNA sequences from two samples being compared demonstrate the heteroplasmy (i.e., both sequences are observed in each sample), the interpretation is cannot exclude (or concordance). If they share a common sequence (i.e., one sample is heteroplasmic and the other homoplasmic, and one of the heteroplasmic types is concordant with the homoplasmic type), then the interpretation is a failure to exclude. If both samples are deemed homoplasmic and differ slightly (i.e., typically at only one site), further investigation is warranted; if no resolution can be attained, the interpretation is inconclusive. However, when rendering an interpretation (for concordance, exclusion, or inconclusive and for assessing weight of the evidence) in any forensic comparison, one should be careful not to exceed current limitations in knowledge, mtDNA or otherwise.

4. When a mtDNA profile from an evidence sample and one from a known reference sample cannot be excluded as originating from the same source, it is desirable to convey some information about the rarity of the mtDNA profile. If the suspect’s mtDNA type and the evidence mtDNA type are considered concordant (i.e., one cannot exclude the samples as originating from the same source), the current practice is to count the number of times a particular sequence is observed in a database(s). A confi-dence interval can be placed on the observation. Thus, based on the size of the database(s) and observed sampling, a range is placed on the frequency of a mtDNA haplotype (however, only an upper bound estimate usually is provided).

5. The statistical weight of a sequence match is presented in the following way: the sequence is searched against a population database (currently consisting of mtDNA sequences from populations of Caucasian, African, Asian, and Hispanic descent). The statement to be reported will state the number of times a particular sequence is present in the database and will include the number of entries in the database from each of the populations listed above. In some instances, a statistical calculation with a 95% confidence interval may be performed. These values should be included in the examiner’s notes.

6. FBI Interpretation Guidelines

7. SEQUENCE FREQUENCY “Unless the discriminatory potential of a test can be objectively evaluated, an inclusion could mean anything. It is therefore incumbent on the forensic scientist to determine a means to evaluate and communicate the significance of an mtDNA inclusion or ‘match’.” Holland & Parsons, 1999. Forens. Sci. Intl.

14. Overall Search Results within Forensic Profiles Average Number of Differences = 6.681

15. African Origin Database(s) within Forensic Profiles Average Number of Differences = 9.881

16. Caucasian Origin Database(s) within Forensic Profiles Average Number of Differences = 4.111

17. A 95% Upper Confidence Interval Using the Normal Approximation of the Binomial P  1.96 [PQ/N]1/2 ; where P = X/N, Q = 1-P, N = Database Size, and X = number of times a matching sequence is found in the database. Frequency of 0 differences = 9.2% Frequency of 1 difference = 27.2%

18. CONFIDENCE LIMIT FROM ZERO PROPORTION P limit = 1 – a – 1/N, where N = Database Size N Maximum Frequency 50 0.058 1 in 17 100 0.030 1 in 34 200 0.015 1 in 67 500 0.006 1 in 167 1000 0.003 1 in 334 5000 0.0006 1 in 1670

19. Statistics Associated with Peterson Case Match Criteria Frequency 95%UCL Not Excluded Zero Differences 0.00725 0.01133 1 in 88 Zero+One 0.09662 0.11085 1 in 9

21. Testimony of Terry Melton in NY v Porco

23. Overall Search Results within Forensic Profiles Average Number of Differences = 1.459

24. Caucasian Database within Forensic Profiles Average Number of Differences = 1.416

25. African origin Database(s) within Forensic Profiles Average Number of Differences = 1.454

27. Cross of Terry Melton in Porco Case