Opinion
2939-2016
03-05-2018
The People of the State of New York, Plaintiff, v. David Seepersad, Defendant.
For the defendant David Seepersad Marnie Leigh Zien, Esq. The Legal Aid Society Criminal Defense Division 49 Thomas Street New York, New York 10013-3821 For the People Office of the New York County District Attorney Nabilah Hossain, Esq. 80 Centre Street New York, New York 10013-4306
For the defendant David Seepersad Marnie Leigh Zien, Esq. The Legal Aid Society Criminal Defense Division 49 Thomas Street New York, New York 10013-3821 For the People Office of the New York County District Attorney Nabilah Hossain, Esq. 80 Centre Street New York, New York 10013-4306 Mark Dwyer, J.
Defendant Seepersad has asked this court to endorse a subpoena for certain records of the Office of the Chief Medical Examiner ("OCME") of New York City. This court has signed the subpoena, and writes this opinion to explain its decision.
A
Defendant Seepersad is charged with possessing a loaded gun recovered from his bedroom. The OCME lab found that DNA on the gun was a mixture of three individuals' DNA. An analysis of the mixture using OCME's Forensic Statistical Tool was highly incriminatory, but was ruled inadmissable by this court in an oral decision rendered on October 16, 2017. That decision will be explained in a written opinion which the court hopes to release soon. The court's ruling was grounded in the views expressed by this judge in People v. Collins, 49 Misc 3d 595 (Sup Ct Kings Co 2015).
OCME has stopped using the Forensic Statistical Tool on DNA mixtures. As of January 1, 2017, OCME now instead uses a program called STRmix. In the wake of this court's ruling of October 16, 2017, OCME used the new program to re-examine the DNA at issue in this case. Once again the analysis produced a highly inculpatory conclusion. Defendant's subpoena is directed to the "raw data" underlying the STRmix results. B
A word about the processing of a DNA sample will provide helpful background to this court's decision to sign defendant's subpoena. After DNA from a sample has been quantified and "amplified," a laboratory analyst seeks to identify the DNA profiles of the contributor or contributors to the sample. The genetic material is fed through a gel in an "electrophoresis" machine. The machine determines which of the many possible alleles are present at the relevant points in the genome, based on the speed at which the sample's alleles pass through the gel. What emerges from electrophoresis is "raw data"—and that is what concerns the parties here. The raw data produced, especially in mixture cases, is for practical purposes usually impossible for humans to read. The data is therefore analyzed through computer software. The software transforms the raw data into peaks atop various points on a flat line. If the software works properly, the peaks reveal which alleles are present. If the alleles indicate that more than one person contributed DNA to the sample, software estimates the odds that a suspect (or other person of interest) was one of them. For this part of the analysis OCME has used the Forensic Statistical Tool, and now STRmix in its turn.
The laboratory procedures up to this point are described in more detail in People v. Collins, 49 Misc 3d 595, 598-99, 604 (Sup Ct Kings Co 2015).
Defendant is not happy with the highly inculpatory results of the STRmix analysis in this case. The Forensic Statistical Tool had found it 172 million times more likely that one mixture on the gun was from defendant and two unidentified, unrelated individuals than from three unidentified, unrelated individuals. STRmix has found it even more likely that defendant was a contributor to a mixture—up to 10 trillion times more likely than not. Critically, however, STRmix based its assumptions on its assessment that there were only two contributors to the mix. The Forensic Statistical Tool analysis was based on a conclusion that there were three contributors. And, quite naturally, defendant is concerned by the inability of the two software packages used in the same laboratory even to agree on the number of contributors.
Defendant wants OCME to provide, in electronic form, the raw data from the electrophoresis run that informed the STRmix analysis. Defendant seeks the data so that his own experts can use different software to process it. Defendant of course hopes that new numbers will at least be less inculpatory than the STRmix numbers. In that regard, and as the Forensic Statistical Tool and STRmix results in this case show, different software can yield highly different estimates of the relevant odds. Indeed, use of one type of software as opposed to another can be the difference between an exculpatory, and an inculpatory, result. C
This court thinks that the potential significance of the raw data makes the issue presented here simple. Neither the People, nor OCME, has a privileged property interest in the raw data. No proprietary OCME software is involved. Nor is the raw data even the result of skilled human effort. The raw data is produced by feeding DNA material into a machine. This court has opined that the procedure involved is so routine that the constitution does not require testimony at trial from the analyst who conducted it. People v. Corey, 52 Misc 3d 987 (Sup Ct NY Co 2016).
In conducting DNA analysis, OCME plainly acts as an agent of the prosecution, as much as the police department does. See e.g., People v. Gills, 52 Misc 3d 903 (Sup Ct Queens Co 2016). The raw data is therefore a "report" or "document" that must be disclosed by the People pursuant to CPL Section 240.20. That section requires the People to disclose every "written report or document . . . concerning a...scientific test... relating to the criminal action... which was made by, or at the request or direction of a public servant engaged in law enforcement activity. . . ." See, People v. Gills, supra; CPL Section 240.20 (1)(c).This court is not certain whether the raw data is often or always printed on paper. But that cannot be decisive. Any report on a computer that gives the results of a scientific process is a written report or document—whether or not anyone hits the "print" button. Thus, for example, the People could not refuse to produce during discovery the final conclusions of STRmix analysis of a DNA mixture simply because the lab did not print them. A document is "written" if it can be printed from a computer.
But what is at issue here is not really just what might be "printed." As noted, humans cannot make sense of raw data without using software to translate that electronic information so that it can be run through a program like STRmix. In this court's view, the electronic data from the electrophoresis run is the real "report" or "document" that discloses the results of a scientific test. The CPL words "report" and "document," employed for decades, must be interpreted in light of 2018 reality.
Moreover, the prosecution and its agents are constitutionally obliged to disclose arguably exculpatory information to the defense. This court does not believe that data critical to the assessment of a laboratory's inculpatory conclusions could be outside the rule, particularly given that different software used to analyze the data will yield different results. Until other software is employed, the People cannot conclude that the raw data will not undercut or refute their case. Neither the People nor this court should, blindly, assume that data in the People's possession is worthless to the defense. Notably, this court has received no legal objection from the People or OCME to defendant's motion for the court to endorse its subpoena—although the People and OCME concede nothing.
The People have opined that the subpoena request is a "delaying" tactic. The court has no reason to think so. Defendant is incarcerated pending trial, and the People have produced DNA results from time-consuming consecutive analyses. --------
This court will conclude with a brief discussion of the relevant precedents. First, a number of appellate opinions agree that defendants have a statutory right to material or information like the raw data here. See, e.g., People v. White, 40 NY2d 797 (1976) (discovery rules require that a defendant be able to test alleged narcotic drugs); People v. Davis, 196 AD2d 597, 598 (2nd Dept 1993); People v. Freshley, 87 AD2d 104, 112 (1st Dept 1982). Many trial judges have agreed, on the precise question posed here. See, e.g., People v. Jones, 55 Misc 3d 743 (Sup Ct Bx Co 2017); People v. Gills, 52 Misc 3d 903 (Sup Ct Queens Co 2016). See also, ABA Criminal Justice Standards on DNA Evidence, Standard 4.1 (a) (viii) (the prosecution should disclose "all raw electronic data produced during testing").
A number of opinions differ. Some judges do not think that material necessary for defense testing is discoverable. But as noted, the Court of Appeals does not agree. Some other judges have thought that production of the raw data will be burdensome to OCME. See People v. Mohammed, 52 Misc 3d 242 (Sup Ct Bronx Co 2016). But production of the same data, in the course of DNA analysis, is not too burdensome when it is done at the prosecution's behest; it is unclear to this court why a copy of the data cannot be created, at the same time or later. In any event, the notion that reproducing the raw data is burdensome has been discredited. See People v. Jones, 55 Misc 3d 743 (Sup Ct Bronx Co 2017). * * *
If this court's analysis is correct, the raw data should be disclosed even without a subpoena. But the subpoena requested by the defense should certainly be signed. ENTER: ____________________________ MARK DWYER Justice of the Supreme Court