MEMORANDUM OF DECISION AND ORDER ON DEFENDANT'S MOTION IN LIMINE TO EXCLUDE ALL RESULTS OF “ULTRA-DEEP NEXT GENERATION SEQUENCING” FOR DISTINGUISHING DNA FROM IDENTICAL TWINS WITH SAMPLES FROM DIFFERENT CELL TYPES
INTRODUCTION
The defendant, Dwayne McNair ("defendant"), is facing eight counts of aggravated rape 1/ and two counts of armed robbery. In advance of his trial, he has filed a "Motion in Limine to Exclude All Results of 'Ultra-Deep Next Generation Sequencing' for Distinguishing D A from Identical Twins with Samples from Different Cell Types," which motion the Commonwealth opposes. After a multi-day evidentiary hearing held pursuant to Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), and Commonwealth v. Lanigan, 419 Mass. 15 (1994) ("Daubert/Lanigan Hearing"), and for the reasons set forth below, the court ALLOWS the motion.
FINDINGS OF FACT
I. Procedural Background
The alleged facts underlying this case are as follows. On September 21, 2004, at approximately 9:45 p.m., the defendant and Anwar Thomas ("Thomas"), both wielding a handgun at various times, kidnapped a young woman ("M.F.") as she walked home in the forest Hills area of Boston. They forced her into their car, pistol-whipped her, and then drove her a short distance to a house she did not recognize at first but later identified as the defendant's home. In a shed and garage appurtenant to the house, the two men took turns raping M.F. before stealing her minidisk player and cellular telephone, driving her back to the Forest Hills area, and releasing her there. Nine days later, on September 29, 2004, at approximately 11 :20 p.m. and into the early morning hours of September 30th, the defendant and Thomas committed a nearly identical attack against a second young woman ("J.G."), who was walking in the Roxbury section of Boston. Both the defendant and Thomas, who was armed with a gun, drove J.G. to a wooded area and raped her in various ways. Thomas then took the victim's cellphone, identification, and $10.00 in cash before the two men drove away. J.G. wrapped in her bra the used condom that the defendant had discarded and put the two items in her jacket pocket.
After their respective attacks, both M.F. and J.G. reported to hospitals, where sexual assault evidence collection kits were performed on them. Deoxyribonucleic acid ("DNA"), the molecule that carries every individual's unique genetic code, was recovered from semen in both victims' kits. DNA from the semen in the condom J.G. retrieved and on her bra also was detected. The semen found in the condom is the subject of the testing at issue here.
In 2007, an investigation conducted by the Boston Police Department's Sexual Assault Unit focused on the defendant as a suspect in the crimes. A grand jury investigation resulted in the defendant and his identical, or monozygotic ("MZ"), twin brother, Dwight McNair ("Dwight"), submitting DNA samples. In 2008, a DNA analysis performed by the Boston Police Department Crime Laboratory matched the defendant's DNA profile to the DNA profile generated from sperm cells from the condom J.G. recovered and from her bra. Because standard Short Tandem Repeat ("STR") analysis is unable to distinguish between MZ twins, the DNA profile generated from those items also matched the defendant's identical twin, Dwight.
In 2010, the Combined Offender DNA Index System (CODIS) produced a match between Thomas' DNA and evidentiary items from both assaults. In 2011, he was charged in an eighteen-count indictment with the abduction and rape of both M.F. and J.G. See Commonwealth v. Thomas, Suffolk County Superior Court Docket No. l 184CR10230. On September 12, 2012, Thomas pled guilty to all eighteen counts against him 2/ and was sentenced to sixteen years to sixteen years and a day in state prison, followed by a ten-year probationary period. As part of his plea agreement, Thomas testified before the grand jury that he had committed the crimes with the defendant, whom he had known since high school and could distinguish from his brother, Dwight, with no difficulty.
As a result of Thomas' testimony (which he is expected to repeat at the defendant's trial), the defendant initially was indicted in connection with this case in 2012. See Commonwealth v. McNair, Suffolk County Superior Court Docket Number 1284CR11006. In April 2014, shortly before the defendant's trial, the Commonwealth learned of a new forensic test developed by Eurofins Medigenomics Forensics[, Ltd.] ("Eurofins"), a German forensic DNA testing laboratory, that supposedly could distinguish the defendant's genome, the genetic material of an organism, from that of his MZ brother. When the Commonwealth's motion to continue the trial in light of the newly-discovered DNA testing methodology was denied, the Commonwealth entered a nolle prosequi 3/ in order to seek a grand jury order for the defendant and his identical twin brother, Dwight, to submit saliva samples suitable for testing by Eurofins.
In the summer of 2014, Eurofins conducted advanced DNA testing to identify the source of the semen left behind during the September 29-30, 2004, rape of J.G. using the semen in the condom and the twins' buccal swabs. According to Eurofins, the test results effectively excluded Dwight as the source of the semen and demonstrated that the defendant is two billion times more likely than his twin brother to be the source. On September 21, 2014, the defendant was reindicted on eight counts of aggravated rape and two counts of armed robbery arising from the two alleged assaults. The defendant's instant motion ensued.
The Daubert-Lanigan Hearing was conducted before this Court over eight days between February 13 and 27, 2017. Testifying for the Commonwealth were Burkhard Rolf, Ph.D., Head of Eurofins' DNA Forensics and Paternity Testing Laboratory and a Qualified Lecturer in Forensics Genetics ("Dr. Rolf'); Michael Krawzcak, Ph.D., Professor, Institute of Medical Informatics and Statistics, Kiel University, Kiel, Germany ("Dr. Krawczak"); Bruce Budowle, Ph.D., Professor, Department of Molecular and Medical Genetics, University of North Texas Health Science Center ("Dr. Budowle"); and Matthew L. Warman, M.D., Director, Orthopaedic Research Laboratories, Boston Children's Hospital, and Professor of Genetics and Orthopaedic Surgery, Harvard Medical School ("Dr. Warman"). Testifying for the defense were David E. Housman, Ph.D., Professor, Department of Biology, Massachusetts Institute of Technology ("Dr. Housman"); David H. Kaye, J.D., Professor, Penn State School of Law ("Prof Kaye"); Sandy L. Zabell, Ph.D., Professor of Mathematics and Statistics, Northwestern University ("Dr. Zabell"); and Paul Goldstein, Ph.D., Professor of Genetics, Department of Biological Sciences, University of Texas ("Dr. Goldstein"). The court also considered forty exhibits and the parties' stipulation as to testimony Drs. Warman and Housman would have given on the subject of paroxysmal nocturnal hemoglobinuria (PNH) and its relevance, or lack thereof, to issues in this case.
II. Eurofins' Testing
A. Basic, Underlving: Genetics Principles
DNA is material that holds our genetic instructions and determines our inherited traits, e.g., stature, eye color, and hair color. Because DNA can replicate, it makes a copy of itself during cell division (called mitosis). Consequently, all our body's cells, except for red blood cells and other cells without a nucleus, contain a complete sample of our DNA. ← Quibble No. 1: Gametes (spermatozoa and ova) contain only half the full complement of nuclear DNA. Because each sperm cell contains a random half of the mother's diploid DNA and another random half of the father's, however, forensic STR profiling of DNA from a mixture of many spermatozoa reflects the STRs in the full genome.
DNA molecules consist of two twisting, paired strands containing four chemical units, called nucleotide bases: adenine (A), thymine (T), guanine (G), and cytosine (C). The bases pair in a specific way: an A always with a T, and a C always with a G. 4/ The sequence of these bases is the "code" of DNA. The human genome that comprises a person's DNA contains approximately three billion base pairs, which are located in the twenty-three pairs of thread-like structures, called chromosomes, located within the nucleus of animal (and plant) cells; one set comes from your mother, and one set comes from your father (making humans "diploid" organisms). Chromosomes are bundles of DNA; they are the basic building blocks of life, carrying our genetic blueprint. ← Quibble No. 2: A blueprint is not a building block.
A locus is simply a location, or marker, on the chromosome; such locations, or loci, reside at specific places on the chromosomes. Inasmuch as each chromosome has a similar chromosome pair (with the exception of a male's X and Y chromosomes), each locus is duplicated. An allele is a variation at a particular site on a chromosome.
Chromosomes range in size from about 50,000 to 300,000 base pairs. DNA sequencing is a process of determining the precise order of the nucleotide base pairs. Although most of our DNA is identical to the DNA of every other human, variations, called polymorphisms, in DNA sequence can be used to distinguish individuals. DNA profiling (or fingerprinting, typing, or testing) is a method of identifying these variable elements within a person's DNA sequence. A machine "reads" a sequence of bases, much in the same way the human eye scans letters to read a sentence. ← Quibble No. 3: Reading" sentences is a metaphor for sequencing the base pairs, but the steps are not analogous to those that the eye and brain use to read letters.
Since its development forty years ago, the Sanger (chain-termination) method for DNA sequencing has been accepted widely in the genetics community. The newer, more rapid next-generation sequencing ("NGS"), also known as massively (or massive) parallel sequencing ("MPS") or high throughput (or "ultra deep") sequencing, is regulated by Standard 17025 of the International Standards Organization ("ISO"), the leading international organization responsible for issuing standards for forensic laboratories ("ISO"); is accepted generally in the genetics community; and is used extensively today in the field of biological research. ← Minor Comment No. 1: ISO 17025 regulates accredited laboratories generally. Many research laboratories are not accredited. This high-level standard contains no instructions on how to sequence genomes.
A female egg cell fertilized by a male sperm cell is called a zygote, which turns into a blastocyst as the cells start dividing and which, after additional cell divisions, becomes an embryo. MZ twins are produced when one sperm cell fertilizes one egg cell (hence, "monozygotic"), after which the zygote splits and develops into two babies; the babies share exactly the same genetic information because this twinning process occurs at the very earliest stage of development. As cells grow in an embryo, they differentiate and become distinct cell types. Cell differentiation occurs after twinning.
In the course of cell division, a rare (1 in 10 million), transmissible mutation can occur when a DNA gene, the basic physical unit of heredity, is changed in such a way so as to alter the genetic instructions carried by that gene. ← Comment No. 1: Mutations are not limited to DNA in genes, and the rate is not the same for every locus. See Elie Dolgin, Human Mutation Rate Revealed: Next-generation Sequencing Provides the Most Accurate Estimate to Date, Nature, Aug. 27, 2009, doi:10.1038/news.2009.864 (“Every time human DNA is passed from one generation to the next it accumulates 100–200 new mutations, according to a DNA-sequencing analysis of the Y chromosome. This number — the first direct measurement of the human mutation rate — is equivalent to one mutation in every 30 million base pairs, and matches previous estimates from species comparisons and rare disease screens. ... But the mutation rate might be somewhat different on other chromosomes.”); Jedidiah Carlson et al., Extremely Rare Variants Reveal Patterns of Germline Mutation Rate Heterogeneity in Humans, 9 Nature Commun. 3753 (2018), doi: 10.1038/s41467-018-05936-5 (“The gold standard for studying the germline mutation rate in humans is direct observation of de novo mutations from family-based whole-genome sequencing (WGS) data 9–12. These studies have produced accurate estimates of the genome-wide average mutation rate (~1 − 1.5 × 10−8 mutations per base pair per generation) and uncovered some of the mutagenic effects of genomic features. However, the inherently low-germline mutation rate means family-based WGS studies detect only 40–80 de novo mutations per trio sequenced 9,10,12, making it difficult to accumulate a dataset large enough to precisely estimate mutation rates and spectrum at a fine scale and identify factors that explain genome-wide variability in mutation rates.”)
If a mutation occurs after the twinning process, it will show up in one twin or the other but not both. ← Minor Comment No. 2: Well, the other twin could randomly acquire the same mutation, but that would be an unlikely event.
A somatic (or acquired) mutation occurs in the general body cells and cannot be passed to the next generation; a germ (or germ line) mutation occurs only in the sexual reproductive cells (i.e., sperm or egg) and will be hereditary.
← Quibble No. 4: If the mutation occurs in the germ cells that make gametes (sperm and ova), then the gametes will transmit the mutation to the next generation.
If only one copy of the gene contains the mutation, it is called heterozygous; if both copies of the gene contain the mutation, it is called a homozygous.
B. The Eurofins Test Itself
It is undisputed that Eurofins is a recognized, forensic DNA laboratory, accredited by the German national accreditation body, the Deutsche Akkreditierungsstelle (DAkkS), in accordance with standards issued by the ISO. Eurofins is a commercial company with a financial stake in the acceptance of its test in courts of law and the global genetics community. ← Minor Comment No. 3: Considering the more than 400 million tests Eurofins performs each year, the financial stake in this boutique testing is quite small.
In the case at bar, Eurofins proposed to identify the source of the semen left behind at the scene of the September 29-30, 2004, rape by extracting and testing DNA from the saliva samples taken from the defendant and his MZ twin, Dwight, and comparing those profiles with the unchallenged STR profile of the semen sample. Eurofins based its planned methodology on a 2013 article, "Finding the needle in the haystack: Differentiating 'identical' twins in paternity testing and forensics by ultra-deep next generation sequencing," authored by Eurofins scientists and published in the peer-reviewed journal, Forensic Science international: Genetics ("Needle in the Haystack Article"). In the Needle in the Haystack Article, Eurofins scientists reported that an identical twin's paternity of a child was determined by comparing the DNA profiles of the twins' sperm samples with the DNA profile of the blood sample of one twin's child. Eurofins reported its testing in this case in Forensics Science international: Genetics; and Dr. Budowle characterized the test's "risk of false identification [as] . . . infinitesimally low" in an editorial in the journal, Investigative Genetics. To date, no article or report has been published in any scientific journal to either confirm or criticize Eurofins' methodology.
In conducting its experiment, Eurofins performed the extraction and then conducted DNA quantification of all extracted DNA. After preparing an "NGS shotgun library generation" (a method for sequencing long strands of DNA by breaking them up into segments and then taking multiple overlapping "reads") for each of the saliva samples, Eurofins then performed NGS/MPS sequencing of the saliva samples pursuant to ISO Standard 17025. Eurofins next used bioinformatics (an interdisciplinary science to analyze complex biological data, such as genetics code) in order to, in Dr. Rolfs words, "map" the results "to the human reference genome sequence" and, thereafter, employed scanning technology "to identify somatic, twin[-]differentiating mutations." (The deeper one goes in examining DNA, the more likely one will find somatic mutations.) Finally, using Sanger sequencing, Eurofins verified nine of the ten single nucleotide polymorphisms (or SNPs) (i.e., variations at a specific position in the genome, akin to copying errors) identified by the NGS/MPS sequencing. 5/ The presence in the semen sample of somatic mutations at Positions 1 and 4 in the defendant's saliva and the absence in the semen sample of any somatic mutation present in Dwight's saliva supported Eurofins' conclusion that the defendant is the source of the sperm sample.
Eurofins then commissioned Dr. Krawczak to conduct a statistical assessment of the relative likelihoods that the defendant and his twin brother, Dwight, were the source of the forensic semen sample in this case. Employing a "likelihood ratio," Dr. Krawczak concluded that it was more than two billion times more likely that the defendant was the source of the semen than Dwight. See Commonwealth v. Mattei, 455 Mass. 840, 842 (2010). ← Minor Comment No. 4: In Mattei, an expert testified that the defendant could not be excluded as a possible source of some of the DNA in a mixed sample. The Supreme Judicial Court held that "accompanying statistical explanation of the meaning of nonexclusion" was essential.
When asked to calculate the same likelihood ratio making the most conservative possible assumptions, Dr. Krawczak concluded that, under those assumptions, the defendant was more than 12,000 times more likely than Dwight to be the source of the trace evidence in question. See id. ← Comment No. 2: One can almost always contrive more conservative assumptions. The next sentence demonstrates that is the case here.
When asked to calculate the same likelihood ratio without assuming independence of the mutations at Positions 1 and 4, Dr. Krawczak determined that the defendant was 320 times more likely than Dwight to be the source of the semen. See id.
The Eurofins' test, which is laborious and cost $120,000 to conduct, and Dr. Krawczak's concomitant statistical analysis are novel. Distinguishing between twins is not a part of normal forensic testing. The methodology has never been replicated by any laboratory in the world; and there have been no peer-reviewed studies of the Eurofins test (other than Eurofins ' own Needle in a Haystack Article), either to support or refute it. Furthermore, the forensic community has not developed guidelines for analyzing the Eurofins test either scientifically or statistically.
C. The Defendant's Challenges to the Eurofins Test
Dr. Housman acknowledged that, using a combination of the two types of sequencing employed by Eurofins (NGS/MPS and Sanger), it is possible to identify mutations that distinguish MZ twins from one another; but he took issue (as did Dr. Goldstein but without Dr. Housman's elaboration) with Eurofins' use of saliva as a proxy tissue for semen. Dr. Housman testified that his concerns in this regard might be "mooted or dispelled" if Eurofins had a semen sample from both twins.
Dr. Housman, a respected geneticist, claimed that, if the relevant mutation events occurred before twinning, it would be possible for one twin to have the relevant mutations in his semen but not his saliva and vice versa (a phenomenon known as "mosaicism"), resulting in a misidentification of the source of the semen. However, Drs. Krawczak, Budowle, and Warman all testified persuasively that this outcome its extremely unlikely. In any event, Dr. Krawczak testified that Dr. Houseman's improbable scenario would not affect his statistical calculations, a point supported by Dr. Budowle, a highly regarded expert in the field of genetics who has authored over 580 articles on genetics and DNA sequencing (but who has testified in court mostly for the prosecution or under the auspices of the Federal Bureau of lnvestigation ("FBI")). 6/
Dr. Housman also criticized the testing assumption that the blastocyst was homozygous in the ground state at each of the nine critical loci in this case. Drs. Krawczak and Warman countered effectively that, if the blastocyst had been heterozygous in the ground state, the heterozygous readings at any location would have shown, but did not, allelic readings of fifty per cent for each allele; they also calculated that, based on the relevant reads, the chance of the blastocyst being heterozygous was very low.
Dr. Goldstein, who provides litigation services mostly to the defense bar on a broad range of topics, initially contended that 92 per cent "coverage" (i.e., the number of reads of a given nucleotide in a sequence) of Dwight and 95 per cent coverage of the defendant introduced uncertainty into the testing but then conceded that this level of coverage was standard for NGS/MPS sequencing. Dr. Goldstein's additional criticism that technical restrictions in the interpretation of the reading at Position 4 7/ cast doubt on the test results was overcome by Dr. Rolf’s assertion that the combination of NGS/MPS sequencing and Sanger sequencing resolved any such doubts.
Drs. Housman and Zabell, who almost always testifies for the defense, attempted unsuccessfully to pick apart Dr. Krawczak's statistical approach by disputing four assumptions made by Dr. Krawczak in arriving at his likelihood ratio calculations: (1) that the mutations at Positions 1 and 4 were independent events, (2) that the mutation rate is constant, (3) that the staging of the mutations could be known, and (4) that the correct mutation rate was uncertain.
Dr. Krawczak explained away each objection. First, there is no basis for an assumption of dependence, given that Positions 1 and 4 are on different chromosomes; besides, Dr. Krawczak provided a likelihood ratio assuming that the mutations were entirely dependent and making the most conservative assumptions possible. Second, Dr. Krawczak relied on an average mutation rate, not a constant one. ← Comment No. 3: Dr. Krawczak used the same number for the probability of a mutation at each locus. That the number is the (estimated) mean probability across the genome does not make it any less of a constant in the Bernouilli model for a mutation that Dr. Krawczak used.
Third, Dr. Zabell agreed that diagrams prepared by Dr. Krawczak resolved his concern about the staging of the mutation rates. ← Comment No. 4: The phrase "staging of mutation rates" betrays some confusion about the computation of the likelihood ratios. Dr. Zabell discussed two issues -- the probability of mutations and the stages at which mutations might have occurred. With regard to the first issue, Dr. Zabell suggested that a range of values should have been used and that the report Dr, Krawczak submitted to Eurofins cited but a single study of mutations in one person. With respect to the second issue -- the timing or stages in the development of the embryos at which mutations occurred in the McNair twins -- Dr. Zabell testified that the report was "very unclear" about what the prosecutor called "the staging of the mutations." He agreed that diagrams shown at the hearing finally "clarifie[d] the framework of [Dr. Krawczak's] analysis" and "the process that he was carrying out." Dr. Zabell did not agree that any diagrams resolved (or even had any bearing on) the uncertainty in the estimated mutation rate that was treated as a known constant in arriving at the likelihood ratios.
Finally, the testimony overwhelmingly supported Dr. Krawczak's assumption of the correct mutation rate.
Prof. Kaye, a law professor, who is not trained as a geneticist, biologist, or embryologist and who never has conducted nor interpreted sequencing results himself, offered little additional grist for the mill. He echoed the concerns of other defense witnesses that, without empirical testing, general acceptance in the relevant scientific community, peer review, and published standards, the Eurofins methodology lacks foundational validity.
CONCLUSIONS OF LAW
I. Underlying Legal Principles
In its groundbreaking Daubert decision, the United States Supreme Court overturned seventy years of jurisprudence and ruled that the so-called Frye "generally accepted as reliable in the relevant scientific community" test, Frye v. United States, 54 App.D.C. 46, 47, 293 F. 1013, 1014 (1923), was displaced by Rules 401 ("Test for Relevant Evidence"), 402 ("General Admissibility of Relevant Evidence"), and 702 ("Testimony by Expert Witnesses") of the Federal Rules of Evidence. Daubert, 509 U.S. at 587. Specifically, Rule 702 then provided:
"If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise."Fed. R. Evid. 702 (1991). 8/ The Daubert Court determined that trial judges, when faced with a proffer of expert scientific testimony, must serve as gatekeepers to ensure that any and all scientific evidence or testimony admined is not only relevant but reliable. 9/ Daubert, 509 U.S. at 589. Accordingly, trial judges must evaluate whether the reasoning or methodology underlying the proposed testimony is scientifically valid and whether that reasoning or methodology properly can be applied to the facts in issue. Id. at 592-93. Pertinent considerations in such an assessment are whether the scientific theory or technique (1) can be (and has been) tested, (2) has been subjected to peer review and publication, (3) has been shown to have a known or potential rate of error, (4) has established standards controlling the technique's operation, and (5) has been accepted generally in the relevant scientific-technical community. Id. at 593-94. See Commonwealth v. Patterson, 445 Mass. 626, 635-36 (2005). All these factors are relevant to the determination of reliability but are not "indispensable predecessor[s] of admissibility." Lanigan, 419 Mass. at 25. See Hicks 's Case, 62 Mass. App. Ct. 755, 760 (2005).
The Supreme Judicial Court adopted the basic reasoning of the Daubert opinion in Lanigan, viz., that a proponent of scientific opinion evidence may demonstrate reliability or validity without establishing general acceptance. Lanigan, 419 Mass. at 26. See Commonwealth v. Sands, 424 Mass. 184, 185-86 (1997). "A determination of the reliability of the testing process entails a fact-based inquiry, including questions of credibility ... The analysis calls on a judge to determine whether testing was properly performed .. . and whether an expert's conclusions based on clinical experience and observations were sufficiently reliable." Commonwealth v. Gaynor, 443 Mass. 245, 264 (2005) (internal citations omitted). The proponent of the expert testimony in question has the burden to establish that the expert's opinion is reliable. Commonwealth v. DiCicco, 470 Mass. 720, 729 (20 15). See Commonwealth v. Curnin, 409 Mass. 218, 223 (199 1). "[T]he touchstone of admissibility is reliability, and not necessarily general acceptance within the scientific community." Commonwealth v. Vao So, 425 Mass. 787, 796 (1997), quoting Sands, 424 Mass. at 185-86. See DiCicco, 470 at 729. However, it may very well be that general acceptance in the relevant scientific community will continue to be the significant, and often the only, issue. Lanigan, 419 Mass. at 26. See Canavan 's Case, 432 Mass. 304, 310 (2000). Indeed, when addressing issues relative to DNA evidence, the Supreme Judicial Court has relied heavily on authoritative reviews by independent organizations such as the National Academy of Sciences' National Research Council, which authored the reports, "DNA Technology in Forensic Science" (1992) and "The Evaluation of Forensic DNA Evidence" (1996), and the book, Strengthening Forensic Science in the United States: A Path Forward (2009). See Commonwealth v. Rosier, 425 Mass. 807, 815-17 (1997); Vao Sok, 425 Mass. at 801; Lanigan, 419 Mass. at 21. See also Gaynor, 443 Mass. at 266 (citing guidelines published by the Technical Working Group for DNA Analysis Methods in 1995).
II. Eurofins' Testing
A. Credibility Determinations
Dr. Budowle, who donated his time here, is a widely recognized pioneer in the field of DNA research; this court found his testimony eminently credible, even though he has done extensive work with the FBI and typically testifies for the prosecution. All the other Commonwealth witnesses, who also donated their time, proved knowledgeable and believable, even though the court is mindful of the fact that Dr. Rolf has a financial stake in the outcome of the court's decision.
Dr. Housman's credentials and background in the field of genetics similarly were impressive. However, his opinions were undermined by the persuasive testimony of the Commonwealth's witnesses. Moreover, Dr. Housman's misgivings about the Eurofins test seemed to emanate chiefly from the lack of additional samples for testing, not from any doubts as to the underlying scientific principles applied.
Prof. Kaye and Drs. Zabell and Goldstein, who testified on behalf of the defendant, displayed something of a defense bent. In addition, Prof. Kaye's marked reliance on the controversial and debatable report of the President's Council of Advisors on Science and Technology (PCAST) gave this court pause. ← Comment No. 5: The affidavit I provided called the court’s attention to the demand in the PCAST report for multiple studies by independent researchers quantifying the accuracy of certain forensic-science identification procedures. I opined that the identification in this case plainly lacked validity under PCAST's requirements. The state’s cross-examination established that prosecutorial and other organizations disputed some of the report's conclusions or questioned the competence of its authors to review the forensic science literature. And, it proposed that because PCAST did not try to publish the report in a peer-reviewed scientific journal, the document could not be regarded as scientifically credible.
B. The Reliability of the Eurofins Test Itself
Based on the credible evidence presented at the multi-day Daubert/Lanigan Hearing, the court determines that the Eurofins testing and Dr. Krawczak's statistical analysis were based on generally accepted, valid scientific or statistical principles, including, inter alia, Sanger sequencing, NGS/MPS sequencing, and the likelihood ratio method, and were unshaken by defense testimony. Moreover, there is no scientific or forensic support for the defendant's suggestion that merely because the Eurofins' methodology is cutting edge it is unreliable. See Rosier, 425 Mass. at 812. Nevertheless, a sound scientific or statistical methodology is not necessarily an admissible one. ← Comment No. 6: Why would testimony from qualified experts based on a scientifically sound and properly applied method be inadmissible? Rule 702, as construed in Daubert, surely is satisfied. Consequently, such expert testimony must be admitted unless some other rule stands in its way.
Of the five Daubert factors, the Commonwealth arguably has established only one here, viz., the proposed methodology has been shown to have a known or potential error rate, i.e., Dr. Krawczak's statistical likelihood ratio. See Daubert, 509 U.S . at 593-94. ← Comment No. 7: Whether a likelihood ratio is an error rate is an interesting question. The Daubert Court seemed to be pointing to empirical studies of false positive and false negative test results. The likelihood ratios here were not that. Nonetheless, a broader reading of Daubert is that if the uncertainty in relying on an inference from the scientific data can be determined, then the trial court should use that information in assessing validity. The likelihood ratio here is (or should be) the ratio of the probabilities of the NGS data under competing hypotheses as to the responsible twin. These conditional probabilities are akin to "error rates."
Conversely, the Commonwealth has failed to meet its burden with regard to the four other factors: the Eurofins test has not been replicated by any laboratory anywhere in the world; it has not been subjected to robust, objective peer review and publication; no standards, protocols, or even guidelines controlling the technique's operation have been established; and the methodology has not been accepted in the relevant scientific-technical community. See id.; Patterson, 445 Mass. at 635-36. Although the Daubert considerations are not "indispensable predecessor(s] of admissibility" Lanigan, 419 Mass. at 25, the absence of four of the five factors vitiates admissibility.
With regard to the first lacking factor, the true test of a methodology's soundness is whether its results "are highly reproducible when appropriate quality-control methods are followed." United States v. Shea, 957 F.Supp. 33 1, 338 (D.N.H. 1997). The smallest possible cohort (one pair of twins) was used here. As several witnesses testified, validation employing up to 100 pairs of MZ twins would be the gold standard of test verification. Even the Commonwealth admitted that validation through additional testing would be preferable. ← Minor Comment No. 5: The first Daubert factor is "whether a theory or technique ... can be (and has been) tested." 509 U.S. 579, 593 (1993). Testing for reproducibility is not the same as testing for validity.
With regard to the second lacking factor, the results of the Eurofins test have not been the subject of peer-reviewed literature in the genetics community. Until a test has had the benefit of independently-conducted validation studies and the results have been scrutinized by the relevant scientific community, the admissibility of such evidence should be questioned. See People v. Young, 425 Mich. 470, 476 (1986). ← Minor Comment No. 6: It would seem that the court applied the PCAST Report's approach to empirical validation after all.
Eurofins' own articles in Forensic Science International: Genetics hardly can be considered independent review; and Dr. Budowle authored merely an editorial on, not an in-depth study of, Eurofins' new technique. The absence of criticism of Eurofins' articles or Dr. Budowle's editorial is not the equivalent of vigorous, independent vetting. "When other scientists analyze and repeat the tests, they counteract the dangers of biased reporting." Id. at 499.
With regard to the third lacking factor, it is undisputed that, although Eurofins' sequencing was conducted under ISO guidelines, no standards or protocols have been established yet relative to Eurofins' unprecedented technique or Dr. Krawczak's newly commissioned statistical analysis. "Common sense dictates that higher academic and professional standards increase the chances that an expert will properly follow the objective criteria and properly employ his subjective consideration to the facts at hand." Patterson, 445 Mass. at 654. All DNA identification methods that heretofore have survived Daubert/Lanigan scrutiny have external standards or guidelines that address the steps in the process. See, e.g., Patterson, 445 Mass. at 642; Gaynor, 443 Mass. at 266; Rosier, 425 Mass. at 812 n.9; Shea, 957 F.Supp. at 340 n.22. As Dr. Budowle acknowledged, forensic science requires the development of protocols under certain circumstances.
Finally, with regard to the fourth lacking factor, it cannot be gainsaid that, although studies have demonstrated that NGS/MPS sequencing successfully can differentiate between MZ twins, the use of mutations in saliva to identify the source of a trace semen sample is not generally accepted in the genetic community. To be sure, a proponent of scientific opinion evidence may demonstrate reliability or validity without establishing general acceptance. Lanigan, 419 Mass. at 26. See Commonwealth v. Sands, 424 Mass. 184-185-86 (1997). However, general acceptance in the relevant scientific community remains a significant issue in determining reliability. Lanigan, 419 Mass. at 26. See Canavan 's Case, 432 Mass. 304, 310 (2000); Rosier, 425 Mass. at 812. Of course, the Commonwealth could argue that unvalidated but otherwise admissible evidence is a matter of weight to be decided by the jury. Commonwealth v. Bly, 448 Mass. 4 73, 489 (2007), quoting Daubert, 509 U.S. at 596 ("Vigorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence").← Comment No. 8: At this point, the opinion falls apart. Daubert does not countenance dispensing with scientific validation to admit otherwise shaky evidence. It does not permit the court to admit "unvalidated but otherwise admissible evidence." Daubert explicitly demands that scientific evidence be scientifically valid to be admissible. The "Daubert factors" are just guideposts for assessing validity. If the trial judge justifiably believed that Eurofins' results and inferences were scientifically valid despite the state's inability to establish that these factors pointed toward validity, then she should have rejected the defendant's Rule 702 arguments. In that situation, "the absence of four of the five factors” would not "vitiate[] admissibility."
Nonetheless, the substantial complexity of DNA evidence and concomitant advanced mathematical computations here may leave jurors confused and uncertain. See Whack v. State, 433 Md. 728, 747-48 (2013). Although the court has the utmost respect for the ability of jurors to comprehend complicated scientific principles, they would not have the luxury of many days of rumination, as this gatekeeper has needed, to untie this Gordian knot. ← Comment No. 9: Having tied itself into knots about Daubert, the court may have merely decided that the balance of probative value compared to the usual counterweights of confusion and time-consumption warranted exclusion. Justice Elspeth B. Cypher of the Supreme Judicial Court denied the state’s interlocutory appeal with the terse observation that "The Commonwealth has not demonstrated that the trial judge abused her discretion." Commonwealth v.McNair, No. SJ-2017-0186 (Mass. July 27, 2017). This one-sentence conclusion begs the question whether the acceptable exercise of discretion pertained to Judge Giles' application of Rule 702, to Rule 403, or to both.
Notes in the Memorandum Opinion
1. The aggravating factors are alleged to be the commission of the rapes as part of a joint venture, in conjunction with a kidnapping, and/or during the commission of an assault by means of a dangerous weapon.
2. At a later proceeding, Thomas was allowed to vacate his guilty plea to eight of the counts; and those eight counts were dismissed with the Commonwealth's assent.
3. Eight of the original eighteen counts against the defendant were dismissed at this time.
4. Because the bases pair in this fashion and the identity of one of the bases in the pair determines the other member, it is customary for scientists not to have to report both bases of the pair.
5. A candidate mutation at Position 7 could not be verified usi ng Sanger sequencing and, therefore, was not used to differentiate the twins.
6. Dr. Budowle's work has been cited in no fewer than three Supreme Judicial Court decisions addressing DNA and genetics matters: Commonwealth v. Rosier, 425 Mass. 807 , 817 n.19 ( 1997); Commonwealth v. Sok, 425 Mass. 787, 802 (1997); and Commonwealth v. Comes, 403 Mass. 258, 272 (1988).
7. Contrary to Dr. Goldstein's testimony, Position 2 was not used in Eurofins' conclusion.
8. The 2017 version of Rule 702 provides as follows:
"A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if:Fed. R. Evid. 702 (2017). See Mass. Guide to Evidence § 702 (2017).
(a) the expert's scientific, technical, or other specia lized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue;
(b) the testimony is based on sufficient facts or data;
(c) the testimony is the product ofreliable principles and methods; and
(d) the expert has reliably applied the principles and methods to the facts of the case."
9. The Supreme Court later applied the Daubert gatekeeping obligation not only to "scientific" testimony but to all expert testimony. Kumho Tire Co., ltd. v. Carmichael, 526 U.S. 137, 147-49 ( 1998).
No comments:
Post a Comment