Summary
finding FBI's method for calculating frequency of defendant's DNA profiles not generally accepted because of lively and current debate regarding existence of population substructure
Summary of this case from State v. VandebogartOpinion
January 9, 1992.
July 20, 1992.
Present: LIACOS, C.J., ABRAMS, LYNCH, O'CONNOR, GREANEY, JJ.
Evidence, Competency, Scientific test, Paternity. Paternity, Rape.
Evidence regarding two laboratories' forensic tests of the deoxyribonucleic acid (DNA) of criminal defendants would not be admissible at their trials for rape and other sexual crimes, where, based on evidence in the record of pretrial proceedings, as well as the reasoning and conclusions of other courts and the writings of experts, this court was unable to conclude that the laboratories' processes of estimating the frequency with which the defendants' DNA profiles occur in the population had found general acceptance in the field of population genetics. [159-164] Evidence of forensic tests of the deoxyribonucleic acid (DNA) of a defendant charged with rape and other sexual crimes, which tended to exclude the defendant as the father of an alleged victim's child, would be inadmissible at the defendant's trial, where the exclusion was based solely on the results of a single genetic probe and where the record of pretrial proceedings contained no scientific justification for the testing laboratory's departure from its customary practice of basing exclusions on two such probes. [164-166]
INDICTMENTS found and returned in the Superior Court Department, five on January 11, 1989, and thirty-four on July 27, 1989.
Consolidated pretrial proceedings were heard by Elizabeth B. Donovan, J. Applications for interlocutory appeals were allowed in the Supreme Judicial Court for the county of Suffolk by Wilkins, J., who ordered that the appeals be entered in the full court.
Stephanie Martin Glennon, Assistant District Attorney ( Marianne C. Hinkle, Assistant District Attorney, with her) for the Commonwealth.
William C. McPhee ( Dolores E. O'Neill with him) for Thomas Lanigan.
Brownlow M. Speer, Committee for Public Counsel Services ( Gordon A. Oppenheim, Committee for Public Counsel Services, with him) for Leo Breadmore, Jr.
After a consolidated pretrial hearing, a judge of the Superior Court ruled that evidence regarding forensic tests of the defendants' deoxyribonucleic acid (DNA) could not be admitted at the defendants' criminal trials. The Commonwealth sought leave to appeal from a single justice of this court. The single justice concluded that the order of the Superior Court judge was the equivalent of the allowance of a motion to suppress and, hence, allowed the Commonwealth leave to appeal to the full court. G.L.c. 211, § 3 (1990 ed.). See Mass. R. Crim. P. 15 (b) (2), 378 Mass. 842 (1978). We affirm the order of the judge of the Superior Court.
I. Facts. The parties do not dispute the basic facts underlying both cases.
A. Thomas Lanigan. In January, 1989, defendant Thomas J. Lanigan, a Caucasian, was indicted for rape of a child, and indecent assault and battery upon three minors. Pursuant to a court order, the Commonwealth obtained a blood sample from Lanigan and submitted it to the Federal Bureau of Investigation (FBI) for DNA testing and comparison with DNA extracted from semen on the clothing of the rape complainant. Through a process known as Restriction Fragment Length Polymorphism analysis (RFLP analysis), the FBI used four genetic probes to develop "profiles" of DNA from both the blood and semen samples. On the basis of these DNA profiles, the FBI concluded that Lanigan's DNA "matched" the DNA from the semen in certain areas specifically identified by the four genetic probes. See Commonwealth v. Curnin, 409 Mass. 218, 229-230 (1991) (describing the use of genetic probes to identify specific areas on the DNA molecule).
A description of Restriction Fragment Length Polymorphism analysis and the purposes for which it is undertaken can be found in Commonwealth v. Curnin, 409 Mass. 218, 227-231 (1991).
The individual areas on the DNA molecule that are targeted by genetic probes are generally referred to as "alleles". See id. at 228. A DNA profile is merely a series of alleles.
In order to determine the frequency with which Lanigan's DNA profile occurs in the general Caucasian population, the FBI compared Lanigan's alleles against allele frequency estimates it had developed from population databases of 200 and 700 Caucasians. The FBI estimated that, when considered against the data base of 200 Caucasians, the probability was four million to one that the DNA of an individual selected at random from the general Caucasian population would "match" the DNA found on the complainant's clothing. When considered against the data base of 700 Caucasians, the probability of a random "match" dropped to 2.4 million to one.
For a description of the process of using population samples to estimate the frequency of particular alleles in a general population, see id. at 230-231.
The Commonwealth presented an expert witness who testified that, because of the magnitude of the numbers involved, the difference between a probability of four million to one and a probability of 2.4 million to one was not troubling from a statistical perspective.
B. Leo Breadmore, Sr., and Leo Breadmore, Jr. In July, 1989, defendants Leo Breadmore, Sr., and Leo Breadmore, Jr., each were indicted for rape of a child, assault with intent to rape a child, indecent assault and battery upon a child, incest, and attempt to intimidate and threaten a witness. All of the alleged victims of the sexual offenses are granddaughters of Leo Breadmore, Sr., and nieces of Leo Breadmore, Jr. One of the alleged victims became pregnant and had a child. This alleged victim testified before the grand jury that the only individuals with whom she had intercourse prior to the birth of her child were Leo Breadmore, Sr., and Leo Breadmore, Jr. She also testified that she had intercourse with both of the Breadmores near the time of the conception of her child.
The Commonwealth obtained blood samples from Leo Breadmore, Sr., Leo Breadmore, Jr., the alleged victim, and her child, all of whom are Caucasian. The blood samples were submitted to Cellmark Diagnostics laboratory (Cellmark) for RFLP analysis to determine whether the DNA of the child "matched" the DNA of one of the defendants. Cellmark used five genetic probes in its RFLP analysis, developing an autorad for each genetic probe, as well as an autorad of a "cocktail" of several probes mixed together. Autorads of the first four individual probes revealed that, according to Cellmark's "matching rules," see id. at 230, the DNA profiles of both Leo Breadmore, Sr., and Leo Breadmore, Jr., "matched" the DNA profile of the child. These four autorads, however, provided no significant differentiation between the DNA profiles of the Breadmore defendants. Accordingly, Cellmark could not yet exclude either of them as the father of the child.
An autoradiograph, or "autorad", is a piece of X-ray film which demonstrates the relative positions and lengths of alleles which have been identified by genetic probes. See Commonwealth v. Curnin, supra at 230.
One of the Commonwealth's experts, Dr. Robin Cotton, testified at the pretrial hearing that the "cocktail" is used because it can often demonstrate an "exclusion," or non-match, more rapidly than is possible with individual probes. The cocktail used for the Breadmores' DNA samples did not reveal any such exclusion.
The autorad of the fifth probe, referred to by Cellmark as G-3, was also inconclusive, but offered the possibility for an exclusion of Leo Breadmore, Jr. To account for the possibility that the first G-3 probe had run off the gel during electrophoresis, Cellmark prepared a second G-3 autorad by using a shorter period for gel electrophoresis. On the basis of the second G-3 autorad, Cellmark excluded Leo Breadmore, Jr., as the father of the child. On the basis of all five autorads, Cellmark concluded that it was "highly likely" that Leo Breadmore, Sr., was the father of the child. At the consolidated hearing, the Commonwealth provided an expert witness who testified that, on the basis of allele frequencies estimates developed by Cellmark, it was 2,500 times more likely that the DNA profile presented by the five autorads would occur in Leo Breadmore, Sr., if he were the father of the child than if he were not the father.
The autorad of the child's first G-3 probe revealed only one band on the autorad. Usually, a probe will reveal two bands: representing one allele inherited from the mother and one from the father. There are two apparent reasons for the absence of a second band on the first G-3 autorad: (1) the child could be homozygotic for the area identified by the G-3 probe, that is, she could have inherited the same allele from each parent; or (2) the second allele could have been so small that it travelled completely across and off the gel during electrophoresis. See id. at 228-229 (describing gel electrophoresis). The shorter a restriction fragment, the more rapidly it travels during gel electrophoresis.
Cellmark used population databases of between 208 and 325 individuals to develop estimates of the frequency in the general Caucasian, Black, and Hispanic populations of each of the alleles targeted by its genetic probes.
C. The hearing. The defendants filed motions in limine to prevent the Commonwealth from introducing any evidence regarding the DNA tests. The defendants argued that DNA test evidence would be unduly prejudicial and speculative. The defendants also argued that the Commonwealth had failed to establish: (1) that the "scientific community generally accepts the theory that DNA tests produce reliable results; (2) [that] techniques currently exist in DNA testing that are generally accepted by the scientific community and that are capable of producing reliable results; and (3) whether the laboratory that performed the tests used these techniques in analyzing the samples in the instant case."
The motion judge held a consolidated hearing to determine whether "the community of scientists involved generally accepts the theory or process" involved in the DNA testing of the defendants. The Commonwealth presented four expert witnesses at the hearing: (1) Dr. Harold Deadman, a supervisory special agent and chemist for the FBI's DNA Analysis Unit; (2) Dr. David Housman, a molecular biologist and professor at the Massachusetts Institute of Technology who conducts research regarding DNA and its role in human disease; (3) Dr. Robin Cotton, a molecular biologist and deputy director of Cellmark's laboratory; and (4) Dr. Neil Risch, a population geneticist and professor at Yale University. The defendants presented a transcript of the testimony of Dr. Laurence Mueller, a population geneticist, given in Commonwealth v. Curnin.
After hearing the testimony of the four expert witnesses, the judge ruled that "the scientific theory underlying DNA testing is generally accepted in the scientific community and satisfies the Frye standard." See Curnin, supra at 220 (parties did not dispute that "the underlying theory and at least the general processes" of DNA testing are accepted in the scientific community). See also Frye v. United States, 293 F. 1013 (D.C. Cir. 1923). The judge ruled, however, that the results of the particular DNA tests in the instant cases were inadmissible. The apparent basis of this ruling was the judge's finding that there was "disagreement within the scientific community regarding the validity of the [Cellmark and FBI] data base[s]." Accordingly, the judge ruled that the statistical probability estimates in both cases were inadmissible. In addition, the judge found that Cellmark's conclusions as to paternity were unreliable, and thus inadmissible, because Cellmark had excluded Leo Breadmore, Jr., solely on the basis of the second G-3 probe, thereby violating an internal Cellmark procedure of excluding only on the basis of two probes.
II. Admissibility of DNA testing evidence. In Commonwealth v. Curnin, we stated that "[w]e have usually applied the Frye test in deciding whether evidence produced by scientific theory or process is admissible, that is, whether the community of scientists involved generally accepts the theory or process." Id. at 222. With regard to DNA testing, we stated that "we would not permit the admission of test results showing a DNA match (a positive result) without telling the jury anything about the likelihood of that match occurring." Id. at 222 n. 7. By way of contrast, "evidence of the absence of a match (a negative result) could properly be admitted without any need for a showing of the likelihood of a match occurring." Id. In the present case, we shall consider the admissibility of both the DNA "matches" of all of the defendants, followed by consideration of the DNA exclusion of Leo Breadmore, Jr. And, as in any case in which we consider the general acceptability of novel scientific evidence, we make our "own determination without regard to the conclusions of the trial or motion judge." Id. at 223. In reaching this determination, we "may properly consider not only the evidence in the record but also the reasoning and conclusions of other courts and the writings of experts." Id.
A. DNA "matches." Because evidence regarding a DNA "match" cannot be admitted without appropriate statistical support, we must consider the general scientific acceptance at this time of the method by which Cellmark and the FBI calculate DNA profile frequencies. The statistical validity of standard DNA profile frequency estimates depends on the accuracy of the assumption that large population groups such as the Caucasian and Black populations "randomly mate" with regard to the alleles that are targeted by genetic probes. W. Thompson S. Ford, DNA Typing: Acceptance and Weight of the New Genetic Identification Tests, 75 Va. L. Rev. 45, 82, 85-87 (1989). In the context of RFLP analysis, random mating means that an individual does not choose his or her mate based on the presence or absence of particular alleles which might occur in a DNA profile. Id. In a random mating population, alleles are distributed proportionately across the population and the likelihood of receiving a particular allele is not affected by the presence or absence of other alleles. Id.
Populations that meet these conditions are said to be in "Hardy-Weinberg equilibrium." See Curnin, supra at 225 n. 11.
The competing theory of population substructure, however, states that there are significant genetic subgroups within larger heterogeneous populations such as the Caucasian or Black population. See United States v. Yee, 134 F.R.D. 161, 181-183 (N.D. Ohio 1991). These subgroups most likely occur along racial and ethnic lines and present significant pockets of genetic variation within the larger population group. Id. Population substructure, therefore, presents the possibility that using allele frequencies of larger population groups might produce an inaccurate frequency estimate for members of substructure groups. Both Cellmark and the FBI calculate allele frequencies within three large population groups only; both appear to proceed on the assumption that population substructure, if present at all, does not affect their frequency calculations.
For the purposes of DNA testing, the FBI calculates allele frequencies for Caucasian, Black, and Oriental populations; Cellmark calculates allele frequencies for Caucasian, Black, and Hispanic populations.
Many population geneticists have joined in the dispute over the extent of population substructure and the consequences that it might hold for the determination of allele frequencies. At the consolidated hearing, Dr. Risch testified that his research and the research of others demonstrates that, even if population substructure exists in a larger group such as the Caucasian population, it does not produce any significant deviation in the frequency of the alleles that usually are targeted by genetic probes. Doctor Risch also testified that both Cellmark and the FBI utilize certain estimation methods, known as "bin" procedures, that produce inherently conservative estimates of allele frequency. In opposition, the defendants presented a transcript of the testimony of Dr. Laurence Mueller given in Curnin that certain studies demonstrate that population substructure does exist among population groups and could have significant consequences for determining the frequencies of alleles. See Commonwealth v. Curnin, supra at 225 n. 11.
None of the parties disputes that the field of population genetics is the relevant field for considering the general acceptance of the process for estimating DNA profile frequencies. See Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).
This dispute has been heard in other court systems as well. For example, in the case of United States v. Yee, supra, a Federal magistrate heard more than six weeks of expert testimony regarding DNA testing, with much of the discussion centering on the debate regarding the extent and the effect of population substructure. Although the magistrate ultimately concluded that DNA test evidence should be admitted, he described his choice as "the most difficult single decision in this case." Id. at 204. "Scientists of indisputable national and international repute and stature, aided and confronted by lawyers of unusual skill and understanding of the issues, took diametrically opposed views on the issue of general acceptability, and those views reflected the division of opinion on the merits of the underlying scientific disagreements." Id. at 206. The magistrate's report contains a particularly careful and comprehensive description of the conflicting positions that certain population geneticists have taken with regard to population substructure. Id. at 180-187, 204-206.
Most recently, the National Research Council has spoken out on the issue of population substructure. In April, 1992, the Council's Committee on DNA Technology in Forensic Science released a report entitled "DNA Technology in Forensic Science." The report, which addresses all aspects of forensic DNA testing and provides specific recommendations for improvement, states that the question "whether actual populations have significant substructure for the [alleles] used for forensic typing . . . has provoked considerable debate among population geneticists." Id. at 3-6. In addition, although the report states that recent studies indicate that substructure does not have a significant effect in the calculation of allele frequencies, it proceeds on the assumption that population substructure may exist. Id. at 3-7.
The Commonwealth is correct that it need not demonstrate absolute unanimity in the scientific world in order to demonstrate general acceptance of Cellmark's and the FBI's processes of estimating DNA profiles. In Commonwealth v. Lykus, 367 Mass. 191, 198 (1975), we stated that "neither infallibility nor unanimous acceptance of the principle need be proved to justify its admission in evidence." However, the lively, and still very current, dispute described above regarding the role of population substructure constitutes something much more than a lack of unanimity. We cannot say that the processes by which Cellmark and the FBI estimated the frequency of the defendants' DNA profiles has found "general acceptance" in the field of population genetics. Accordingly, evidence of the estimated frequencies of the defendants' DNA profiles is not admissible. Because the frequency estimates are inadmissible, evidence of a match between profiles is also inadmissible. See Commonwealth v. Curnin, supra at 222 n. 7.
It bears noting that the National Research Council's report provides several specific recommendations for DNA testing which would, apparently, account for the possibility of any excessively large frequency estimates caused by population substructure. First, the report recommends immediate studies to sample ethnic groups directly, and to compare observed allele frequencies in order to quantify the extent of population substructure. Id. at 3-8 — 3-10. Second, and more importantly for the purposes of the judicial system, the report recommends the use of a "ceiling frequency" for all DNA frequency calculations. Id. at 3-10 — 3-14. The use of a ceiling frequency, which would represent the greatest observed frequency of particular alleles within a given number of randomly selected population groups, automatically would provide for the greatest, and therefore most conservative, estimate of the frequency of a DNA profile. One of the principles behind the recommendation of a ceiling frequency is the notion that any error in calculating profile frequencies that is caused by population substructure should accrue to the benefit of the individual against whom the DNA testing is being used. Id. at 3-13. This principle is entirely in keeping with the hope that we expressed in Commonwealth v. Curnin, supra at 227, that the scientific community would "generally agree on a means of arriving at a conservative estimate of the probability of another person having the same alleles and thus resolve all uncertainties and variables in favor of the defense." The national call for considered, conservative approaches to DNA testing, such as the use of ceiling frequencies, and the absence of such an approach in the present cases, underscore the wisdom of the motion judge in excluding the test evidence.
B. Cellmark's exclusion of Leo Breadmore, Jr. In Commonwealth v. Curnin, supra, we stated that in appropriate circumstances evidence of a non-match, or exclusion, could be admitted without frequency estimates. Accordingly, we must now consider whether evidence of Cellmark's exclusion of Leo Breadmore, Jr., would be admissible. The Commonwealth argues that the judge's finding that Cellmark's determination of paternity is unreliable is based on a fundamental misunderstanding of the DNA testing process. In its brief to this court, the Commonwealth states that "[t]he motion judge was flatly wrong to state that `problems . . . necessitated a rerun of [the] G-3 probe,' and constituted a `violation of Cellmark's internal procedure.'" The Commonwealth points to expert testimony that Cellmark has a standard procedure of running the G-3 probe twice, and that the second G-3 probe was conducted not because of any "problems" but in order to obtain additional information. Accordingly, the Commonwealth argues, the judge's findings that Cellmark violated its internal procedures and that the test results are unreliable are erroneous and should be overturned.
While it may be that the expert testimony supports the Commonwealth's contention that the second G-3 probe was run in order to gain additional information, and not due to any problems with the first G-3 probe, it is patently clear from the judge's memorandum of decision that the Commonwealth has inaccurately paraphrased the judge's finding. The motion judge did not find that Cellmark violated any of its internal procedures regarding the running of the G-3 probe; instead, the judge specifically found that Cellmark violated its internal procedure in excluding Leo Breadmore, Jr., solely on the basis of the results of the second G-3 probe. The hearing testimony amply supports the judge's finding. The deputy director of Cellmark's laboratory, Dr. Cotton, testified that it was not only Cellmark's procedure, but the procedure of "every paternity testing laboratory I know" to exclude on the basis of two probes. Dr. Cotton testified that two probes are recommended by the American Association of Blood Banks to avoid an artificial exclusion due to gene mutation.
The Commonwealth, however, argues that Cellmark was justified in excluding Leo Breadmore, Jr., on the basis of one probe because "this was not a normal paternity case, but an incestual mating, where the potential fathers not only were both related to the child's mother but also were father and son, making the ability to exclude quite different than in a `normal' case." We note that the Commonwealth's reliance on the "incestual mating" between one of the defendants and the alleged victim assumes the truth of the charge of incest that is sought to be proven through the DNA tests.
Even without regard for this reference, however, we conclude that we are not persuaded by the Commonwealth's arguments. The Commonwealth's expert witnesses testified that an exclusion in a paternity case involving relatives is "more difficult" and "different" than it would be in a "normal" paternity case because relatives share more of the same gene types than do unrelated individuals, thereby making "matches" by individual genetic probes more likely. The Commonwealth has not offered any scientific justification for allowing exclusions in these "difficult" and "different" cases to be judged on the basis of a less restrictive standard, one probe, than is allowed in a "normal" paternity case. The Commonwealth's expert, Dr. Risch, testified that in a paternity case involving relatives "[i]t may be difficult to distinguish between [relatives] unless you use a large number of [genetic probes]". This statement suggests that reliability in these cases is not served by making exclusion decisions on the basis of fewer, rather than more, genetic probes.
In Commonwealth v. Curnin, supra at 222 n. 7, we suggested that challenges to the admissibility of DNA testing evidence "should focus on the soundness and general acceptance of the particular testing process for forensic use, and, if raised, on the proper implementation of that process in the given case." In the absence of any scientific justification for Cellmark's deviation from its two-probe exclusion standard, we agree with the motion judge's finding that the exclusion of Leo Breadmore, Jr., is unreliable and therefore inadmissible.
The Commonwealth does not question, nor do we consider, whether the Commonwealth should be allowed to submit additional allele frequency estimates which have been determined on the basis of a revised process. In addition, the parties have not briefed, and therefore we do not consider, the materiality of paternity test results. The crimes charged do not require paternity to be established. These issues are for the trial judge.
III. Conclusion. The ruling of the motion judge is affirmed. The cases are remanded to the Superior Court for proceedings consistent with this opinion.
So ordered.