United States v. Morgan
53 F. Supp. 3d 732
| S.D.N.Y. | 2014Background
- Defendant Johnny Morgan indicted for unlawful possession of a firearm (18 U.S.C. § 922(g)) after a .40 caliber Glock was recovered near where he was stopped; three DNA swabs were taken from the gun and an oral swab from Morgan.
- The New York City OCME performed Low Copy Number (LCN) PCR/STR DNA testing on the gun swabs; the sample at issue contained ~14 pg of DNA (LCN by OCME definition).
- OCME uses 31 PCR cycles for LCN (versus 28 for high-copy testing), increasing stochastic effects (allelic drop-in, drop-out, stutter, heterozygote imbalance).
- Morgan moved to exclude OCME’s LCN results under Daubert/Fed. R. Evid. 702, challenging OCME’s validation, sample quantity/quality, peak-height reliance, number-of-contributors analysis, and representations to the NY Commission; the Court held a multi-day Daubert hearing.
- OCME presented internal validation studies (sensitivity, mixtures, degraded/touch samples), ASCLD accreditation, SWGDAM-guided protocols, peer-reviewed publications, and NY State Commission/DNA Subcommittee review/approval of LCN methods.
Issues
| Issue | Plaintiff's Argument (Gov’t) | Defendant's Argument (Morgan) | Held |
|---|---|---|---|
| Admissibility of OCME LCN DNA results under Daubert/Rule 702 | OCME’s methods are validated, peer-reviewed, accredited, and reliably applied to the facts; expert testimony helps trier of fact | OCME’s LCN methodology is insufficiently validated for this sample and unreliable; results should be excluded | Admissible: Court found OCME’s LCN methods sufficiently reliable under Daubert/702; objections go to weight, not admissibility |
| Adequacy of OCME validation studies | Validation (sensitivity, mixture, degraded samples) provides scientific basis for interpretation protocols | Validation used primarily pristine buccal swabs and not crime-stain degraded mixtures of the exact type/quantity here; grants too much inferential leap | Adequate: Court held studies meet Daubert; lack of exact same-test studies affects weight, not admissibility |
| Sample quantity/quality (14 pg) and stochastic effects | Quantity alone is not dispositive; stochastic effects determine interpretability; OCME validation included lower amounts and relevant studies | 14 pg is too low; OCME previously stated it would not test below 20 pg; results unreliable | Admissible: Court agreed no categorical lower limit; stochastic effects evaluated case-by-case; DNA Subcommittee found no absolute lower limit |
| Use of peak-height data and mixture interpretation | Peak heights are accepted in SWGDAM guidance and validated by OCME’s studies to inform heterozygote/major contributor assessments | Peak-height measures unreliable in LCN and should not be relied upon for inferences | Admissible: Court held use of peak-height ratios consistent with best practices and validated; criticisms go to weight |
Key Cases Cited
- Daubert v. Merrell Dow Pharm., 509 U.S. 579 (gatekeeping standard for scientific expert admissibility)
- Kumho Tire Co. v. Carmichael, 526 U.S. 137 (trial judge’s broad discretion in assessing expert reliability)
- Amorgianos v. National R.R. Passenger Corp., 303 F.3d 256 (Daubert factors nonexclusive; focus on methodology)
- General Elec. Co. v. Joiner, 522 U.S. 136 (court may exclude expert opinion if analytic gap exists)
- United States v. McCluskey, 954 F. Supp. 2d 1224 (contrast: LCN inadmissible at one lab lacking validation/accreditation)
- United States v. Zajac, 749 F. Supp. 2d 1299 (Daubert does not require every conceivable validation test)
- Graham v. Playtex Prods., 993 F. Supp. 127 (proponent bears burden to establish admissibility under Daubert)