1

9
2

CHAPTER 13

Admissibility of DNA Evidence in Court

ANDRE A ROTH

F orensic DNA typing has existed since the late 1980s, and has been admitted in
court cases as evidence of identity since the late 1980s (Kaye, 2010, pp. 60–​63).
Some DNA evidence admissibility questions are now relatively uncontroversial (such
as a “match” between robust single-​source polymerase chain reaction-​short tandem
repeat (PCR-​STR) profiles), and others are still contentious (such as results of “low-​
copy-​number” testing and interpretations of complex mixtures by expert systems).
This chapter offers a brief overview of the legal rules governing the admissibility of
forensic DNA typing results, primarily in US court cases.
Any discussion of the admissibility of DNA in court, it should be said, is really a
discussion of several different questions of admissibility. Before introducing evidence
of a DNA profile “match” at trial, for example, the proponent must show not only
that the DNA typing method is reliable, but also that the method of interpreting the
results and calculating the statistical significance of the results is reliable. Thus, in a
case involving a complex DNA mixture in which the prosecution alleges that a sus-
pect is a likely contributor and seeks to introduce a likelihood ratio (LR) reported by a
probabilistic genotyping software program like TrueAllele, the prosecution might be
called upon in a pretrial reliability hearing to establish the reliability of (1) the PCR-​
STR method used to compare alleles among various potential contributors to a mix-
ture; (2) the reliability of the expert system in estimating the number of contributors
and whether a peak is a true allele or an artifact; and (3) the reliability of the statis-
tical method the system uses to generate the LR, along with the reliability of the LR
itself as an expression to the jury of the statistical significance of the results.
It is also worth noting that even where forensic DNA typing results are admis-
sible as an evidentiary and constitutional matter, their meaning and probative value
might be vigorously contested at trial by the opponent. The parties might disagree
over whether a peak is a true allele or artifact and offer conflicting expert testimony
on the matter; whether a match statistic is grossly overstated and offer conflicting
expert testimony on the matter; the relevant population of potential contributors to

Andrea Roth, Admissibility of DNA Evidence in Court In: Silent Witness. Edited by: Henry Erlich, Eric Stover and Thomas J.
White, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780190909444.003.0014
2
9

a mixture; and the prevalence of phenomena like DNA transfer, which might offer an
alternative innocent explanation for the presence of a person’s DNA at a crime scene.
An opponent might also question the qualifications or conclusions of a DNA expert,
even if that expert succeeds in testifying. In short, the admissibility of DNA—​that
is, whether a judge or jury determining the facts of a case is even allowed to hear the
results of forensic DNA typing—​is only the first of many questions related to how
the legal system treats DNA evidence in court.

THE BASIC LEGAL RULES GOVERNING ADMISSIBILITY

OF DNA EVIDENCE

To be admissible in a civil or criminal trial in the United States, forensic DNA typing
results must comply with the jurisdiction’s rules of evidence (each state, as well as
the federal system, has its own rules of evidence), as well as provisions of the US
Constitution that give certain trial rights to those accused of a crime.
This section focuses on the rules of admissibility of DNA evidence applicable
at trial. Although not all court cases involving DNA go to trial, the rules related to
admissibility of evidence at trial loom large over settlement or plea negotiations,
which are conducted in the shadow of a trial. And while DNA typing results might
also be offered in legal proceedings beyond trial, such as sentencing proceedings,
the rules governing admissibility of evidence at sentencing are generally both simple
and permissive. For example, the US Sentencing Guidelines state that sentencing
courts “may consider relevant information without regard to its admissibility . . . at
trial, provided that the information has sufficient indicia of reliability” (U.S.S.G. §
6A1.3(a)).

Reliability Requirements for Expert Testimony

The first set of legal rules governing the admissibility of DNA relate to the require-
ment, under statutory or common law rules of evidence, that expert testimony based
on scientific methods be reliable. Nearly all DNA typing results offered to prove iden-
tity are presented through one or more expert witnesses: laboratory technicians,
DNA analysts, statisticians, population geneticists, and the like. As a result, the ad-
missibility of DNA will turn in part on the rules of evidence governing expert witness
testimony. In particular, nearly every state, as well as the federal system, requires
that expert testimony be based on reliable methodology.
Some courts—​following the so-​called Frye standard—​delegate the question of
reliability of DNA to the scientific community, allowing the admission of expert tes-
timony based on DNA typing and interpretation methods so long as those methods
are “generally accepted” within the relevant scientific community. Before the 1920s,
scientific evidence was treated like most other evidence, subject only to the usual
requirements of relevance, witness competence, and the like (Spring Co. v. Edgar,
U.S., 1878). But in 1923, the D.C. Circuit Court of Appeals in Frye v. United States

[ 292 ] Challenges and Debates

3
9
2

held that a criminal defendant accused of murder, James Frye, could not offer the
expert testimony of Dr. William Moulton Marston—​who would later create the char-
acter Wonder Woman—​that Mr. Frye had taken and passed a polygraph examination
(Frye; Lepore, 2015). According to the Frye court, a novel scientific methodology
like the polygraph should not be admitted unless it is “sufficiently established” as
a method “to have gained general acceptance” among the “authorities” in the field
(Frye, p. 1014). Because the polygraph “ha[d]‌not yet gained such standing and sci-
entific recognition,” it was properly excluded by the trial court. The Frye “general
acceptance” standard was highly influential and ultimately became the dominant
standard in US courts for admissibility of expert testimony based on such methods.
Not until 1993, with the US Supreme Court’s decision in Daubert v. Merrell
Dow Pharmaceuticals, would the Frye test’s dominance be challenged. In Daubert,
the Court held that the federal rule of evidence governing admissibility of expert
testimony—​Rule 702—​did not require that an expert’s method be “generally ac-
cepted.” The rule’s language required only that an expert’s “scientific” or other tech-
nical or specialized knowledge be helpful to the jury, which in turn required only
that a method—​if purportedly scientific—​be scientifically valid (Daubert). And
like all preliminary questions related to admissibility of evidence, the scientific va-
lidity of an expert scientific method must be determined by the trial judge, not by
the scientific community. Thus, the Court reasoned, expert testimony is admissible
so long as the expert is qualified and her method, if scientific, is deemed by the trial
judge to be sufficiently reliable. In setting forth the factors to be considered by trial
judges in determining scientific validity, the Daubert Court relied heavily on Karl
Popper’s view of the scientific method (Daubert, p. 593). Influenced by Popper’s
preoccupation with the concept of falsifiability, the Daubert Court set forth the
following nonexhaustive list of factors to be considered by a judge in determining
reliability of an expert method: (1) whether the method “can be (and has been)
tested”; (2) whether the method “has been subjected to peer review and publica-
tion”; (3) the method’s “known or potential rate of error”; (4) “the existence and
maintenance of standards controlling the technique’s operation”; and (5) whether
the method is generally accepted in the “relevant scientific community.”
In two subsequent decisions, the Supreme Court held that the Daubert reliability
test applies not only to the method an expert uses but also to the expert’s applica-
tion of that method (General Electric Co. v. Joiner, 1997) and that Daubert applies
not only to “scientific” methods but to all expert testimony, including nonscien-
tific “technical” fields like tire-​tread analysis (Kumho Tire Co. v. Carmichael, 1999).
Together, Daubert, Joiner, and Kumho Tire are typically called the “Daubert trilogy”
(Bernstein & Jackson, 2004). The language of Federal Rule of Evidence 702 has been
amended to reflect the trilogy’s holdings and now requires both that the “testimony
is the product of reliable principles and methods” and that “the expert has reli-
ably applied the principles and methods to the facts of the case” (F.R.E. 702(c),(d)).
Courts applying Daubert to scientific methods continue to apply the nonexhaustive
Daubert factors (testability, peer review, existence and extent of error rate, existence
of standards to govern the method, and general acceptance in the scientific commu-
nity) set forth in Daubert itself (see, e.g., Moss, 2015).

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 293 ]

4
9
2

Today, most states and the federal system have shifted to the Daubert standard,
either through court decisions or through passage of a statute or rule similar
to Federal Rule of Evidence 702. Still, a significant minority of states, including
New York and California, continue to adhere to Frye in determining admissibility of
novel scientific evidence such as DNA (Jurilytics, 2017). Thus, in relying on prece-
dent related to admissibility of a particular DNA method, litigants should be aware
of which standard—​Frye or Daubert—​governed the precedential decision and which
standard governs in the litigant’s jurisdiction.

The Evidentiary Rule against Hearsay and the Constitutional

Right of Confrontation

The second set of rules that might preclude admission of DNA typing results is the
rule against “hearsay” and its corresponding constitutional rule, the confronta-
tion clause of the Sixth Amendment to the US Constitution. For better or worse,
the Anglo-​American system prefers that the claims of human witnesses, if offered
for their truth, be made live, in court, subject to the oath, physical confrontation,
and cross-​examination. Thus, the Federal Rules of Evidence exclude “hearsay”—​an
out-​of-​court statement offered “for the truth of the matter asserted in the state-
ment”—​as presumptively inadmissible (F.R.E. 801(c), 802). All 50 states have an
analogous rule (Broun, 2013, § 244). Hearsay is thus inadmissible unless the pro-
ponent lays a foundation for admissibility under an applicable exception to the rule
against hearsay, such as for “business records,” “statements against penal interest,”
or “dying declarations” (Broun, 2013, § 245 et seq.). Even if a hearsay statement is
admissible as an evidentiary matter under an exception, its admission in a criminal
case against the accused might still violate the confrontation clause if it is the “tes-
timonial” hearsay of a nontestifying declarant. Hearsay is generally “testimonial”
if it is a sufficiently solemn statement that is either facially accusatory or created
with the help of government officers, such as a stationhouse police confession of a
defendant’s alleged accomplice (Crawford v. Washington) or a formal affidavit of a
forensic chemist about the presence of a drug in a tested substance (Melendez-​Diaz
v. Massachusetts).
In the context of DNA, these two rules arise most often when a testifying DNA ex-
pert determines a match or calculates a match statistic based in part on the hearsay
report of another DNA expert who does not testify at trial.

The Fourth Amendment

The Fourth Amendment to the Constitution also gives rise to potential admissi-
bility challenges to DNA evidence in court when offered in a criminal case against
the accused. The Fourth Amendment protects against “unreasonable searches and
seizures” and prohibits the issuance of a search or arrest warrant unless supported
by “probable cause” (U.S. Const. amend. IV). Police may therefore apply for a search

[ 294 ] Challenges and Debates

5
9
2

warrant to obtain a nonconsensual DNA sample from a criminal suspect only if they
have probable cause to believe the DNA will show that the suspect has committed
a crime. Police can also obtain DNA from a suspect by consent without violating
the Fourth Amendment (Schneckloth v. Bustamonte; Will, 2003). To the extent forced
DNA sampling in the absence of any individualized suspicion to believe the suspect
is engaged in a crime is unconstitutional, any DNA test results stemming from such
a Fourth Amendment violation may be inadmissible under the “exclusionary rule” as
the “fruit” of a constitutional violation (Mapp v. Ohio; Maryland v. King). As discussed
further later in the chapter, the primary context in which criminal defendants have
argued that evidence of a DNA match violates the Fourth Amendment is in database
“cold hit” cases.

THE STATUS OF RELIABILITY-​B ASED ADMISSIBILITY

CHALLENGES TO DNA EVIDENCE

This section explores the status of reliability challenges to various forms of DNA ev-
idence, setting forth both areas of consensus, in which the reliability of DNA typing
results will not likely be disputed, and areas of controversy, in which the reliability of
DNA typing results or statistical methods is more contentious.

Single-​S ource PCR-​S TR Testing Results and Random Match

Probabilities (RMPs)

Some forms of DNA evidence are now universally accepted as evidence of identity in
US courts as a matter of reliability. The original forms of forensic DNA testing and
interpretation used in the 1980s and early 1990s were subject to much criticism
during the “DNA Wars,” the history of which has been ably told by others (Kaye,
2010; Lynch et al., 2008; see c­hapter 1). But these earlier techniques have been
replaced in forensic DNA analysis by PCR-​based STR discrete-​allele typing. Courts
now universally accept as generally reliable both the PCR process for amplification
of DNA and the STR-​based system of identifying and comparing alleles (Kaye, 2010,
pp. 190–​191).
The most common PCR-​STR–​based kits used in forensic analysis in crim-
inal cases in the United States are those manufactured by Applied Biosystems
(such as ProFiler/​CoFiler, testing 13 core STR loci, including amelogenin, or sex;
IdentiFiler, testing 16 loci, including amelogenin; and, most recently, GlobalFiler,
testing 24 loci)) and Promega’s PowerPlex kits (FBI, n.d.). These are the primary
kits accepted by the Federal Bureau of Investigation (FBI) for uploading to the
Combined DNA Index System (CODIS) for comparison purposes (FBI, n.d.).
All new reference samples taken by convicted or arrested persons for upload to
CODIS must contain 20 “core CODIS loci” and thus must be tested using the most
recent, most highly discriminating kits (FBI, n.d.). All evidence samples from
crimes or missing persons inquiries must be at least tested for the core CODIS

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 295 ]

6
9
2

loci, though results at all 20 loci are not necessary for comparison purposes (FBI,
n.d., §§ 19, 20).
In addition, the use of the RMP to express the statistical significance of a match
between two PCR-​STR single-​source (non-​mixture) profiles has also been universally
accepted by US courts.1 In general, evidence of a DNA match is inadmissible without
a corresponding match statistic expressing the statistical significance of the match
to the factfinder.2 In the words of one court, “[w]‌ithout the probability assessment,
the jury does not know what to make of the fact that the [DNA] patterns match: the
jury does not know whether the patterns are as common as pictures with two eyes,
or as unique as the Mona Lisa.” (United States v. Yee, 1991). The RMP is the product
of the probabilities of a person having each of the alleles represented in a single-​
source PCR-​STR profile (Butler, 2009, pp. 229–​230). While the RMP has been the
subject of some academic debate because its accuracy rests on the assumption of
statistical independence among the STR loci and minimal population substructure,
courts have universally accepted it as a reliable expression of the statistical signifi-
cance of a match between two single-​source samples under both Frye and Daubert
(see, e.g., Mueller, 2008). Only where the RMP has been mistaken by a prosecutor
as the chance of the defendant’s innocence—​the “prosecutor’s fallacy” or “fallacy
of the transposed conditional”—​have courts commented on its potential for undue
prejudice.3 Moreover, while expert witness assertions of source attribution based

1. See, e.g., State v. Roman Nose, 667 N.W.2d 386, 398 (Minn. 2003) (“[A]‌random match
probability statistic [product rule] is scientifically acceptable when applied to a known
single source sample.”); and People v. Smith, 132 Cal. Rptr. 2d 230, 233 (Ct. App. 2003)
(“Defendant concedes, ‘It is generally accepted the [polymerase chain reaction and short
tandem repeats] can be completely accurate in typing genetic material from single source
samples.’“). Cf. United States v. Silva, 889 F.3d 704, 718 (10th Cir. 2018) (expert described
single-​source sample analysis “as easy as you can get”).
2. See, e.g., State v. Tester, 968 A.2d 895, 909 (Vt. 2009) (“[A]‌dmission of DNA match ev-
idence, without additional evidence of the frequency with which such matches might occur
by chance, is error.”); Deloney v. State, 938 N.E.2d 724, 730 (Ind. Ct. App. 2010) (deeming
DNA evidence inadmissible without “accompanying testimony explaining the statistical
significance of those non-​exclusion results”); United States v. Davis, 602 F. Supp. 2d 658,
673 (D. Md. 2009) (“DNA evidence cannot be admitted in a vacuum; the Government must
also present some additional information with which a jury can accurately assess the sig-
nificance of the consistency between a defendant’s DNA profile and that of the evidence.”);
and Commonwealth v. Mattei, 920 N.E.2d 845, 858 (Mass. 2010) (“The challenged expert
[DNA] testimony concerning the nonexclusion results should not have been admitted
without accompanying statistical explanation of the meaning of nonexclusion.”). But see
State v. Hummert, 933 P.2d 1187, 1191 (Ariz. 1997) (noting that in Arizona, no numerical
statistic is required as foundation for expert testimony on a DNA match).
3. The “prosecutor’s fallacy,” or fallacy of the transposed conditional, occurs when a
lawyer (or judge or juror) mistakes the RMP (e.g., one in a million) for the probability that
the defendant is not the source. Put differently, the person hearing the statistic mistakes
one conditional probability (the chance the defendant would match the profile, given that
he is not the source of the DNA, or the RMP) for its transposed conditional (the chance the
defendant is not the source, given that he matches). See, e.g., McDaniel v. Brown, 558 U.S.
120 (2009) (noting that the prosecutor and the government’s DNA expert both engaged
in the fallacy of the transposed conditional in their statements before the jury); and (Roth,
2010), explaining the fallacy in laypersons’ terms. ().

[ 296 ] Challenges and Debates

7
9
2

on RMPs have been the subject of some defense challenges, courts generally allow
DNA experts to testify to their opinion, based on an exceedingly small RMP, that the
DNA profiles share a common source. For example, FBI analysts commonly testify to
source attribution above any RMP threshold of 1 in 300 billion (1,000 times the US
population) (Butler, 2009).

Y-​S TR and Mitochondrial DNA (mtDNA) Testing Results

Litigants have also made reliability challenges to Y-​STR and mtDNA testing results,
with little success. While forensic PCR-​STR typing looks at short repeated sequences
of DNA on the “autosomal” (nonsex) chromosomes (Butler, 2009), forensic Y-​STR
typing looks at certain short repeated sequences on the Y chromosome in male
DNA samples. Men inherit the Y-​STR profile of their father, and the profiles are not
believed to change much over generations. Thus, the statistical significance of a Y-​
STR match is not calculated by an RMP; the allelic frequency tables that generate
RMPs for traditional PCR-​STR profiles assume statistical independence of the STR
markers. In Y-​STR typing, in contrast, analysts use a “counting method” to generate
a match statistic. That is, they look for the Y-​STR profile or haplotype in a database
of Y-​STR profiles from a relevant population, take the resulting number of “hits”
(say, zero or one), and build a confidence interval around that number to express the
chances of seeing additional matches in a larger population sample. Thus far, courts
appear to have universally accepted Y-​STR typing results as reliable when offered
by the government in a criminal case, both under Frye4 and Daubert.5 Notably,
courts have excluded Y-​STR typing results in certain circumstances when offered by
a criminal defendant in postconviction proceedings as evidence of innocence, even
while acknowledging that Y-​STR typing is generally accepted.6 One recent court also

4. See, e.g., People v. Zapata, 8 N.E.3d 1188, 1195 (Ill. Ct. App. 2014) (holding that no new
Frye hearing was required because Y-​STR was already generally accepted); State v. Calleia,
997 A.2d 1051, 1065 (N.J. Ct. App. 2010), rev’d on other grounds, 206 N.J. 274, 20 A.3d 402
(2011) (ruling that Y-​STR is generally accepted); Commonwealth v. Jacoby, 170 A.3d 1065,
1095 (Pa. 2017), petition for cert pending (ruling as a matter of first impression that Y-​STR
is generally accepted); State v. Bander, 208 P.3d 1242, 1255 (Wash. Ct. App. 2009) (no new
Frye hearing required because Y-​STR is already generally accepted); and People v. Stevey,
148 Cal. Rptr. 3d 1, 12 (Ct. App. 2012) (no new Kelly-​Frye hearing required because Y-​STR
is already generally accepted).
5. See, e.g., People v. Tunis, 318 P.3d 524, 528 (Colo. Ct. App. 2013) (holding as an issue
of first impression that Y-​STR is reliable under Daubert); People v. Wood, 862 N.W.2d 7,
24 (Mich. Ct. App. 2014), judgment vacated in part on other grounds, 498 Mich. 914, 871
N.W.2d 154 (2015) (trial court did not abuse its discretion by admitting expert testimony
on Y-​STR after holding Daubert hearing); and State v. Maestas, 299 P.3d 892, 934 (Utah
2012) (same).
6. See, e.g., Commonwealth v. DiCicco, 25 N.E.3d 859, 869 (Mass. 2015) (holding that
the trial judge did not abuse her discretion by excluding defendant’s proffered Y-​STR ex-
pert witness, because the Y-​STR exclusion was based on a single potential allele, a method
that is discouraged, though not prohibited, under the SWGDAM Y-​STR guidelines); People
v. Stoecker, 10 N.E.3d 843, 849 (Ill. 2014) (holding that, given the DQ alpha testing done

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 297 ]

8
9
2

excluded evidence of Y-​STR results in a (currently pending) Texas murder trial, but
only apparently because the results did not incriminate the defendant other than the
tendency to show that the DNA was male (Winkle & Goard, 2018).
Challenges to mtDNA typing results have been similarly unsuccessful. Unlike
Y-​STR typing, mtDNA typing isolates a long sequence of DNA in a particularly
hypervariable region of DNA found in the mitochondria of one’s cells (outside
the nucleus). Like Y-​STRs, mtDNA is not recombinant; we inherit our mtDNA se-
quence from our mothers, and mtDNA sequences (like Y-​STRs) are not believed to
change much over generations. As with Y-​STR haplotypes, the statistical significance
of a match between mtDNA sequences is expressed through the counting method,
building a confidence interval around the number of matching haplotypes found in a
relevant mtDNA population database. To be sure, mtDNA match statistics have been
criticized for being misleading and inaccurate, given the amount of “clustering” of
haplotypes based on migration patterns (Kittles et al., 2006). Nevertheless, most if
not all challenges to mtDNA typing results or match statistics in US courts have been
unsuccessful under Frye7 and Daubert,8 with very few challenges even being brought
in the last decade.

Admissibility Challenges to “Low Copy Number” DNA

Typing Results

The primary context in which admissibility challenges to single-​source nuclear DNA

comparison results are still successful is low copy number (LCN) DNA testing. Many
laboratories have a different set of protocols for testing DNA samples involving an
amount of input DNA lower than 1 nanogram, a level below which the identification

before trial, there was no “reasonable likelihood of more probative results” using Y-​STR
typing after conviction); and People v. Barker, 1-​12-​3238, 2015 WL 2069736, at *8 (Ill. App.
Ct. Apr. 30, 2015) (same).
7. See, e.g., People v. Stevey, 148 Cal. Rptr. 3d 1, 11 (Ct. App. 2012) (“mtDNA evi-
dence . . . has also gained general acceptance within the scientific community”); People
v. Klinger, 713 N.Y.S.2d 823, 831 (N.Y. Co. Ct. 2000) (holding that mtDNA is generally
accepted and reliable based on holdings in a number of other jurisdictions); Magaletti
v. State, 847 So. 2d 523 (Fla. Ct. App. 2003) (same); State v. Pappas, 776 A.2d 1091, 1108
(Conn. 2001) (same); and People v. Holtzer, 255 Mich. App. 478 (2003) (same).
8. See, e.g., State v. Brochu, 949 A.2d 1035, 1049 (Vt. 2008) (“Although mtDNA evidence is
relatively new, all jurisdictions to have considered the issue have uniformly found mtDNA
to be reliable.”); United States v. Beverly, 369 F.3d 516, 531 (6th Cir. 2004) (holding that
trial court did not abuse its discretion in admitting mtDNA evidence because the “scientific
basis for the use of such DNA is well established”); United States v. Coleman, 202 F. Supp. 2d
962 (E.D. Mo. 2002) (holding mtDNA admissible under Daubert); Wagner v. State, 864 A.2d
1037, 1044 (Md. Ct. App. 2005) (same); and State v. Council, 515 S.E.2d 508, 518 (S.C.
1999) (same). Cf. State v. Griffin, 384 P.3d 186, 203 (Utah 2016) (noting, while discussing
probative/​prejudicial balancing, that “every state that has been confronted with the ques-
tion of whether mtDNA is admissible under its applicable rules of evidence has answered
the question in the affirmative”).

[ 298 ] Challenges and Debates

92

and interpretation of alleles becomes more difficult and controversial (Butler, 2015,
pp. 159–​160; ISHI Conference, 2017). Several trial and appellate courts have ruled
LCN testing reliable, under both Frye9 and Daubert.10 Nonetheless, at least a handful
of courts have excluded LCN testing results on reliability grounds.11

Admissibility Challenges to Mixture Interpretations

by Human Analysts

Criminal defendants in the United States challenging the admissibility of DNA

typing results have perhaps had the most traction in cases involving DNA mixtures
and the presentation of a match statistic called the combined probability of inclu-
sion, or CPI, to the factfinder.
Because DNA mixtures involve more than one contributor, it may be difficult
for analysts to determine how many contributors there are to a mixture, and which
alleles at each locus belong to which contributor. As a result, a simple calculation
of an RMP using allelic frequencies and the product rule, as analysts do for single-​
source sample comparisons, is not possible in mixture statistics. And while new
probabilistic genotyping software programs hold great promise for mixture deconvo-
lution and the calculation of highly discriminating match statistics based on the con-
sideration of thousands of permutations, analysts are more limited in their mixture
interpretation abilities. Thus, analysts calculate a match statistic in mixture cases by
(1) identifying all alleles at all loci, (2) determining whether the reference sample of
interest (such as a criminal suspect) has alleles that are consistent with the alleles

9. See, e.g., Phillips v. State, 226 Md. App 1 (Md. Ct. Spec. App. 2015) (holding LCN DNA
analysis admissible under Frye and that any attack on its reliability went to its weight
rather than admissibility); People v. Garcia, 963 N.Y.S.2d 517 (N.Y. Sup. Ct., Bronx County
2013) (holding both that LCN DNA testing is not a novel science that would require a Frye
hearing before being admissible and that LCN DNA testing conducted by the OCME in
New York is generally accepted); and People v. Megnath, 898 N.Y.S.2d 408 (N.Y. Sup. Ct.,
Queens County 2010) (holding that LCN DNA testing as conducted by the OCME is admis-
sible under Frye). Cf. People v. Lazarus, 190 Cal. Rptr. 3d 195, 239 (Ct. App. 2015) (holding
that defendant was not entitled to a Frye hearing because there was no evidence that LCN
was not generally accepted).
10. See, e.g., United States v. Morgan, 53 F. Supp. 3d 732 (S.D.N.Y. 2014), aff ’d, 675
F. App’x 53 (2d Cir. 2017) (holding that the OCME’s LCN testing methodology is reliable
under Daubert); and United States v. Barton, 8:14-​CR-​496-​T-​17AEP, 2016 WL 4921036,
at *6 (M.D. Fla. Sept. 14, 2016) (holding that magistrate did not abuse its discretion in
denying Daubert challenge to LCN results). Cf. United States v. Sleugh, 14-​CR-​00168-​YGR-​2,
2015 WL 3866270, at *3 (N.D. Cal. June 22, 2015), appeal pending (holding that no Daubert
hearing was necessary to introduce LCN results).
11. See United States v. McCluskey, 954 F. Supp. 2d 1224 (D.N.M. 2013) (excluding LCN
testing under Daubert because the New Mexico Department of Public Safety laboratory
used different procedures and methods than the New York OCME, and there was no
evidence that the NM procedures and methods would yield reliable results); and People
v. Collins, 15 N.Y.S.3d 564 (N.Y. Sup. Ct., Kings County 2015) (holding LCN testing using
the OCME’s “FST” software was not admissible under Frye).

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 299 ]

03

in the mixture, and (3) if so, calculating the chance that a randomly selected person
would also be consistent with the combination of alleles present in the mixture—​the
CPI (sometimes referred to equivalently as the “random man not excluded” statistic)
(Bieber et al., 2016; Butler, 2015).
The CPI has been the subject of considerable criticism, from both sides of the
American criminal justice system. On the one hand, the CPI is a much less discrimi-
nating statistic than the RMP and tends to ignore a significant amount of relevant in-
formation about likely number of contributors and likely contributor profiles (Curran
& Buckleton, 2008). At the same time, some argue that the CPI carries too great a risk
of falsely inculpating innocent suspects, because it removes any loci from its statistical
calculation that exhibit signs of “allelic dropout”: based on a suspect’s allele being ab-
sent from the mixture, rather than considering the possibility that the absence of the
suspect’s allele reflects that the suspect is simply not a contributor (Murphy, 2015,
pp. 92–​94; Butler, 2015; Curran & Buckleton, 2010).
In 2010, in response to the critiques of the CPI from the scientific community, the
Scientific Working Group on DNA Analysis Methods (SWGDAM) changed its DNA
mixture interpretation guidelines to require laboratories to remove certain loci from
their CPI calculations. SWGDAM’s concern was with peaks that were above the profiling
system’s “analytical threshold” (AT) (the height below which a peak carries too great a
risk of being an artifact rather than genetic material) but below the system’s “stochastic
threshold” (ST) (the height above which stochastic effects,12 such as allelic dropout, are
unlikely to occur). The new guidelines required laboratories to remove any locus from
the CPI calculation that contained any peak above the AT but below the ST.13In 2016, a
distinguished group of scientists reiterated this call to remove any locus from the CPI
calculation where any peak was within a certain range (Bieber et al., 2016; PCAST, 2016,
p. 78 (citing Bieber et al. 2016)). And later in 2016, the President’s Council of Advisors
on Science and Technology (PCAST) concluded that the CPI method is “clearly not
foundationally valid” under Daubert (PCAST, 2016, p. 78).
While the CPI has been introduced as a valid match statistic under both Frye14 and

12. Stochastic effects are those due to sampling issues caused by the low number of
events.
13. See Scientific Working Group on DNA Analysis Methods, “SWGDAM Interpretation
Guidelines for Autosomal STR Typing by Forensic DNA Testing Laboratories” (approved
January 4, 2010, § 4.6.3) (“When using CPE/​CPI (with no assumptions of number of
contributors) to calculate the probability that a randomly selected person would be
excluded/​included as a contributor to the mixture, loci with alleles below the stochastic
threshold may not be used for statistical purposes to support an inclusion. In these
instances, the potential for allelic dropout raises the possibility of contributors having
genotypes not encompassed by the interpreted alleles.”).
14. See, e.g., State v. Bigger, 227 Ariz. 196, 205, 254 P.3d 1142, 1151 (Ariz. Ct. App. 2011)
(CPI is “generally accepted,” even when applied to LCN samples); and Phillips v. State, 126
A.3d 739, 751 n.11 (Md. Ct. Sp. App. 2015), aff ’d on other grounds, 152 A.3d 712 (Md.
2017) (holding in footnote that the use of a CPI statistical computation for a steering
wheel DNA sample was admissible because the laboratory “analyzed the steering wheel
sample in a generally accepted manner.”).

[ 300 ] Challenges and Debates

1
0
3

Daubert15 in numerous criminal trials in the United States, at least one recent court has
rejected the CPI as unreliable,16 and other courts have recently reversed convictions
based on the presentation of mixture statistics to a jury that, viewed in retrospect,
are vastly more inculpatory—​sometimes by several orders of magnitude—​than
they would have been under the post-​2010 guidelines (see, e.g., Moran, 2017; Texas
Forensic Science Commission, 2015). Moreover, some courts have excluded a CPI
on undue prejudice grounds in cases where the statistic is only minimally discrimi-
nating (such as excluding only 50% of the population).17 Litigants involved in DNA
mixture cases should therefore be aware of the guideline change and how it might
affect match statistics presented at both past and current trials.

Admissibility Challenges to Complex Mixture Interpretations

by Expert Systems

Some laboratories have begun to address the problems of the CPI by employing ex-
pert systems to interpret DNA mixtures. Unlike human analysts, expert systems, or
probabilistic genotyping software (PGS) in this case, can consider much more infor-
mation, including data sets estimating allelic dropout averages at various loci, than
human analysts can. Instead of calculating the CPI, these expert systems calculate an
LR or similar statistic that purports to compare the probability of seeing the mixture
given the competing hypotheses that the suspect (or other person of interest) is or is
not a contributor to the mixture. The resulting LRs tend to be much more discrimi-
nating than the CPI; a typical LR reported by the program TrueAllele would state, “A
match between Mr. [Defendant] and the fingernails is 189 billion times more prob-
able than a coincidental match to an unrelated Caucasian.” (Perlin, 2010). Expert
systems have also been wielded by lawyers for criminal defendants as evidence of
innocence in high-​profile exonerations (see, e.g., McCall, 2018). While some PGS are
open source, most are proprietary. One proprietary program, the “FST” software de-
veloped by New York’s Office of the County Medical Examiner (OCME), was excluded
by one trial judge under Frye, prompting the OCME to both make the source code
public and shift to using a different program (Jacobs, 2016). Now, the two main

15. See, e.g., State v. Haughey, 3 A.3d 980, 992 (Conn. Ct. App. 2010) (defendant could
not demonstrate that the trial court abused its discretion in admitting CPI evidence).
16. United States v. Williams, 3:13-​CR-​00764-​WHO-​1, 2017 WL 3498694, at *13 (N.D.
Cal. Aug. 15, 2017) (excluding DNA analysis from the SERI laboratory, which used a
“suspect-​centric” form of CPI analysis “not based on sound methodology.”). Cf. People v.
Smith, C062513, 2011 WL 4528254, at *22 (Cal. Ct. App. Sept. 29, 2011) (unpublished)
(holding that “the CPI was an improper statistical analysis to be used in this case because
[the expert] knew he was dealing with a mixture that had significant allelic dropout,” but
finding no prejudice).
17. See, e.g., People v. Pike, 53 N.E.3d 147, 170 (Ill. App. Ct. 2016), reh’g denied (May 2,
2016), appeal denied, 89 N.E.3d 761 (Ill. 2017) (holding that it was error, but not plain
error, for the trial court to admit a “50% inclusion probability statistic” derived from CPI
calculations “because the statistic was irrelevant”).

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 301 ]

2
0
3

programs used in forensic DNA testing in the United States are TrueAllele, owned by
the Pittsburgh, Pennsylvania, company Cybergenetics; and STRMix, owned by the
New Zealand research institute ESR (PCAST, 2016, p. 80).
The 2016 report of the President’s Council of Advisors on Science and Technology
(PCAST) cited expert systems as an improvement over existing human anal-
ysis of complex mixtures, concluding that such methods have been established as
foundationally valid for mixtures with three or fewer contributors, where the minor
contributor constitutes at least 20% of the intact DNA in the mixture and the DNA
amount exceeds the minimum required by the method for analysis (PCAST, 2016,
p. 82). The report suggested, however, that use of the software beyond its empirically
established range could be problematic. To be sure, the PCAST report has itself been
subject to criticism both for failing to more fully solicit the participation of forensic
examiners and law enforcement and for placing a premium on properly designed
“black box” validation studies as a prerequisite for foundational validity (National
District Attorney’s Association, 2016; Budowle, 2017).
So far, the LRs from both STRMix and TrueAllele have been admitted in nu-
merous courts across the country, in both Frye and Daubert jurisdictions.18 In fact,
TrueAllele has not been excluded on reliability grounds by any court, although
one California trial court—​later reversed by a higher court—​had deemed the
failure to disclose TrueAllele’s source code a barrier to its admissibility (People
v. Chubbs, 2015). STRMix has been deemed inadmissible in two cases. In the first,
the New York Hillary case, the trial judge excluded the evidence under Frye not
based on a ruling that STRMix is an inherently unreliable method, but on the lack
of internal validation studies by the local laboratory that conducted the testing
(Hillary Order, 2016, pp. 9–​10). Notably, the inculpatory LR generated by STRMix
in Hillary was contradicted by TrueAllele results on the same sample, which in-
dicated that Mr. Hillary was likely not a contributor (Roth, 2017, p. 2019). In
the second case, in June 2018, a Texas trial judge excluded under Daubert the
STRMix results from male DNA found on the thigh of a female murder victim,
after human analysis “came up inconclusive” (Winkle & Goard, 2018). In another
recent case, currently pending appeal, a California state judge has conditioned the
admissibility of STRMix under Frye on the government providing the source code
to the defense, which it thus far has refused to do (People v. Dominguez, 2018).
While the research institute ESR offers limited access to STRMix’s source code to
defense experts before trial under a nondisclosure agreement, the company has
declined to allow broader access, citing a trade secret privilege. While some legal
commentators have suggested that no such trade secret privilege should exist in
criminal cases (Wexler, 2018; Chessman, 2017), no appellate court has yet been

18. See, e.g., People v. Bullard-​Daniel, 54 Misc. 3d 177, 191 (N.Y. Co. Ct. 2016) (holding
STRMix results admissible under Frye). The Cybergenetics website lists numerous cases
in which TrueAllele has been admitted under either Frye or Daubert. See TrueAllele
Admissibility, cybgen.com, https://​www.cybgen.com/​information/​admissibility/​page.
shtml. Likewise, the STRMix website lists numerous cases around the country and globe
in which STRMix has been admitted. See https://​strmix.esr.cri.nz/​#news.

[ 302 ] Challenges and Debates

3
0

persuaded by such arguments to uphold a trial court’s order requiring source code
disclosure.
In sum, the results of expert systems TrueAllele and STRMix have thus far been
widely held admissible as evidence of both guilt and innocence. Nonetheless, if de-
fense requests for access to source code continue to be granted, and if disclosure
of source code is deemed a condition of admissibility, then the proprietors of these
programs may have to subject their source code to further scrutiny or face possible
exclusion of results in certain cases. Moreover, to the extent these systems con-
tinue to be used on complex mixtures beyond the empirically established range
of the software—​with multiple contributors, extreme peak height differential, or
low template DNA—​and to the extent different systems continue to generate con-
tradictory results, courts might be more receptive to reliability challenges in the
future.

Admissibility of Evidence Related to DNA “Transfer”

Under certain circumstances, DNA can “transfer” from one individual or surface to
another (direct transfer), or even from that second person/​surface to a third person/​
surface (secondary transfer) (Butler, 2009, p. 80; 2011, pp. 18–​19). The likelihood of
transfer occurring is a function of a number of factors, including the type of sur-
face touched and whether the individuals involved are “shedders” or “non-​shedders”
(Fonneløp et al., 2017). In a given case, the likelihood of transfer might be a critical
issue for the factfinder, in terms of what inference to draw from the presence of a
person’s DNA at a crime scene. For example, in one recent case, a state appellate
court ruled that the presence of a defendant’s DNA on a handgun found in his house
was insufficient evidence to convict him for possession of the gun, because of the
high likelihood of transfer (Finley v. State, 2014). And in a high-​profile California
case, a homeless man, Lukis Anderson, accused of killing a wealthy Silicon Valley
investor, was eventually exonerated after the presence of his DNA on the victim’s
fingernails was explained by DNA transfer; the same EMTs who responded to the
murder scene had assisted Anderson earlier in the day and could have transferred
traces of Anderson’s DNA to the victim (Worth, 2018). While numerous courts have
allowed expert testimony as to transfer, several courts have also denied motions for
postconviction relief filed by defendants claiming that a DNA transfer expert would
have made a difference at trial.19

19. See, e.g., Adams v. State, 161 Idaho 485 (Ct. App. 2016) (holding that the addition
of expert testimony on DNA transfer would not have made a difference to the trial out-
come); and Sancier v. Comm’r of Correction, 139 Conn. App. 644 (2012) (acknowledging
that DNA transfer is “theoretically possible” but ruling that transfer evidence would not
have affected outcome). Cf. State v. Freeman, No. 28150, 2008 WL 142299, at *1 (Mo. Ct.
App. Jan. 16, 2008), rev’d en banc, 269 S.W.3d 422 (Mo. 2008) (rejecting defendant’s claim
that the DNA evidence against him was insufficient because of the possibility of transfer).

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 303 ]

4
0
3

THE STATUS OF CONSTITUTIONAL ADMISSIBILITY

CHALLENGES TO DNA EVIDENCE

This section explores the status of constitutional challenges to DNA evidence under
the confrontation clause and Fourth Amendment.

Confrontation Clause Challenges to Reliance on Hearsay

DNA Reports of Nontestifying Analysts and to Proprietary
Expert Systems

Because of the rule against hearsay and the confrontation clause, the proponent of
a forensic DNA report cannot offer the report itself into evidence without calling the
report’s author to the witness stand. However, the proponent of the testimony can
circumvent the hearsay rule by having the testifying analyst simply explain to the
factfinder that her expert opinion is based upon the other analyst’s report. Under
Federal Rule of Evidence 703 and its state analogs, so long as the hearsay report it-
self is not offered as evidence, a testifying expert is free to “base an opinion” upon
hearsay or other inadmissible evidence (Williams v. Illinois; F.R.E. 703).
Although the evidentiary rules allow a testifying DNA analyst to rely on another
analyst’s hearsay report in rendering an opinion, there is still an open question as to
whether such testimony might violate the confrontation clause in a criminal case if
offered against the accused. In Williams v. Illinois (2012), the Supreme Court heard
a rape case in which the state’s DNA analyst testified that the defendant’s PCR-​STR
profile, which was tested and developed at the testifying analyst’s state laboratory,
“matched” the profile developed from the victim’s vaginal swabs, which were tested
and analyzed at a different laboratory, Cellmark. The defendant argued that the
analyst’s testimony violated the confrontation clause because the analyst’s expert
opinion was based in large part on the analysis and conclusions of Cellmark’s analyst,
who did not testify. A majority of justices of the Supreme Court concluded that the
testimony did not violate the confrontation clause, but no one argument received
a full five votes. Four justices concluded that the nontestifying expert’s report did
not implicate the clause because it was technically offered only to explain the basis
of the testifying analyst’s opinion, rather than for its “truth.” An additional justice
concluded that the testimony did not implicate the clause because the hearsay report
of the nontestifying analyst was not sufficiently formal or solemn to count as testi-
monial hearsay, given that the Cellmark analysis was conducted before a suspect had
been identified.
Because neither of the theories of admissibility in Williams garnered five votes,
and because of the changing composition of the Supreme Court, the Williams deci-
sion leaves unresolved whether future DNA cases with slightly different facts might
present a confrontation clause problem. For example, a recent decision by New York’s
highest state court reversed a burglary conviction on confrontation clause grounds
where the testifying DNA analyst, who opined that the defendant’s DNA matched
the DNA from the crime scene but who had not conducted, witnessed, or supervised

[ 304 ] Challenges and Debates

5
0
3

the DNA testing in the case, simply read to the jury the hearsay report of another,
nontestifying DNA analyst colleague (People v. Austin, 2017).
The other potential confrontation clause challenge to DNA evidence relates to the
results of proprietary expert systems offered at trial. A reported LR from TrueAllele is
not considered “hearsay” under American rules of evidence because it is not an asser-
tion by a human witness; TrueAllele cannot be placed under oath, “cross-​examined,”
or physically confronted. But some commentators have argued that expert systems
offering accusatory claims against criminal defendants should perhaps be considered
“witnesses against” the defendant for purposes of the confrontation clause (see, e.g.,
Roth, 2017). While “confrontation” of a proprietary algorithm would not be synony-
mous with “cross-​examination,” it might involve disclosure of source code; disclosure
of prior statements of the algorithm related to the same subject matter; or a right
to some sort of technical transparency report that reveals relevant assumptions
of the program, such as the program’s estimate of allelic dropout rates, or stutter
percentages, at various loci. With the exception of a dissenting California Supreme
Court justice, however, no appellate court has yet been persuaded that machine-​
generated results might implicate the confrontation clause (People v. Lopez, 2012).
Indeed, at least one US Supreme Court justice has intimated that “raw data” from
a machine would likely not implicate the confrontation clause (Bullcoming v. New
Mexico, 2011, Sotomayor, J., dissenting).

Fourth Amendment Challenges to DNA Database

“Cold Hit” Results

Thus far, Fourth Amendment challenges to forcible DNA sampling of those convicted
of or arrested for certain crimes, and the uploading of the resulting DNA profiles to
CODIS for comparison purposes, have been unsuccessful. Each state, as well as the
federal system, maintains a DNA database, authorized by statute, containing the
PCR-​STR profiles of people convicted of, or arrested for, various crimes (Roth, 2013).
These official statutory databases are all interconnected through CODIS, allowing
local police anywhere in the country to compare a crime scene DNA profile to the
14+ million profiles in CODIS to look for a match, or “cold hit,” to an unsolved case.
In 2013 the Supreme Court upheld the constitutionality of Maryland’s arrestee data-
base, on grounds that states can reasonably require arrestees to give DNA for identi-
fication purposes, just as they are forced to give fingerprints (Maryland v. King, 2013).
Most recently, the California Supreme Court held that California’s arrestee database
was also constitutional under King, a critical decision, given that California’s data-
base differs significantly from Maryland’s database in that it does not allow for au-
tomatic expungement of an arrestee’s record and is more expansive in the crimes it
covers (People v. Buza, 2018).
To the extent that Fourth Amendment challenges to database “cold hits” might be
successful in the future, they will probably relate to more controversial tactics such
as familial searching, searching of genealogy websites, or collection of “abandoned”
DNA. Familial searching entails searching a DNA database not just for a perfect match,

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 305 ]

6
0
3

but for a partial match, indicating that a person with partially matching might be
related to the perpetrator (Murphy, 2010; Chapter 4). While familial searching has
been banned in a few states, including Maryland, it is explicitly permitted in 12
states, including California (Rainey, 2018). While proponents argue that the practice
helps catch elusive criminals (Rainey, 2018), critics argue that it has a significant ra-
cially disparate impact and unfairly allows law enforcement to scrutinize people who
are not in a criminal or even an arrestee database, in the absence of any suspicion of
wrongdoing (Murphy, 2010).
No Fourth Amendment challenge has yet been successful against the collection of
“abandoned” DNA, which people inadvertently leave on coffee cups, cigarette butts,
and the like. The prevailing logic is that because the DNA has been “abandoned,” it
is akin to garbage, which is devoid of Fourth Amendment protections because the
owner has abandoned it and thus has no “legitimate expectation of privacy” in the
contents under existing constitutional doctrine (Joh, 2006). A Fourth Amendment
challenge against government searches of commercial genealogy databases, onto
which members have voluntarily posted their DNA profiles for comparison with
other members, might be precluded on the same grounds. California police recently
identified a suspect in the Golden State Killer case based on a search of the open-​
source genealogy website GEDMatch, conducted without a warrant, without prob-
able cause, and in a way that violated the terms of service (by creating a fake name
associated with the crime scene DNA profile they uploaded for comparison purposes)
(Zhang, 2018; see ­chapter 15).
The Supreme Court’s recent decision in Carpenter v. United States might breathe
new life into such challenges, however. In Carpenter, the Court held that the gov-
ernment cannot subpoena historical cell phone location records without a warrant,
even though the defendant had shared his location with his cell phone company
(Carpenter v. United States, 2018). In doing so, the Court significantly limited the
reach of the “third-​party doctrine” that a suspect has no legitimate expectation of
privacy in information he shares with others.20 The full implications of Carpenter for
DNA database search challenges remains to be seen.

CONCLUSION: DNA ADMISSIBILITY ISSUES

ON THE HORIZON

The next decade will inevitably bring further DNA admissibility issues as the tech-
nology advances. For example, the use of Rapid DNA machines—​which can develop
a profile from a sample in as little as 90 minutes—​on crime scene evidence samples
will surely be the subject of Daubert and Frye reliability challenges. While reference
samples developed with Rapid DNA are eligible for upload to CODIS, crime scene

20. See Carpenter v. United States, 585 U.S. _​_​(2018), slip op. at 11 (“Given the unique
nature of cell phone location records, the fact that the information is held by a third party
does not by itself overcome the user’s claim to Fourth Amendment protection.”).

[ 306 ] Challenges and Debates

7
0
3

Figure 13.1 Composite sketch of a crime suspect created by Maryland police using DNA
phenotyping. Source: “Maryland composite suspect sketch created through phenotyping.
Courtesy of Montgomery County Police.

(evidence) samples are not (FBI, n.d.). Likewise, the use of DNA phenotyping to iden-
tify crime suspects may well face admissibility challenges. Phenotyping involves
estimating the physical characteristics of a suspect based on a crime scene DNA pro-
file (Southall, 2017). As figure 13.1 shows, the technique is already being used to
develop composite sketches of suspects.
Because phenotyping is used only to initially identify a suspect, not to ultimately
prove that the suspect matches the DNA from a crime scene, it is unlikely that the
practice will trigger reliability challenges at trial. However, the reliability of the tech-
nique may well be relevant to Fourth Amendment challenges to searches and seizures
based on a suspect’s alleged similarity to an estimated phenotype.
In sum, the foundational validity of PCR-​STR forensic DNA typing for single-​
source robust samples and the use of the RMP to express the statistical significance
of two matching single-​source profiles are well established as reliable. Nonetheless,
several aspects of forensic DNA typing may still continue to raise significant admis-
sibility issues.

REFERENCES

Bernstein, D., & Jackson, J. (2004). The Daubert trilogy in the States. Jurimetrics, 44,
351–​366.

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 307 ]

8
0
3

Bieber, F., et al. (2016). Evaluation of forensic DNA mixture evidence: Protocol for
evaluation, interpretation, and statistical calculations using the combined
probability of inclusion. BMC Genetics, 17, 125–​140.
Broun, Kenneth S., et al., eds. (2013). McCormick on evidence. 7th ed. Eagan, MN: Thomson
Reuters.
Budowle, B. (2017, June 17). Open letter and affidavit re: PCAST Report. University of
North Texas Health Science Center.
Bullcoming v. New Mexico, 564 U.S. 647 (2011).
Butler, J. (2009). Fundamentals of DNA typing. Cambridge, MA: Academic Press.
Butler, J. (2011). Advanced topics in forensic DNA typing: Methodology. Cambridge,
MA: Academic Press.
Butler, J. (2015). Advanced topics in forensic DNA typing: Interpretation. Cambridge,
MA: Academic Press.
Carpenter v. United States (2018, June 22). Slip opinion. Retrieved from https://​www.
supremecourt.gov/​opinions/​17pdf/​16-​402_​h315.pdf.
Chessman, C. (2017). A “source” of error: Computer code, criminal defendants, and the
Constitution. California Law Review, 105, 179–​228.
Crawford v. Washington, 541 U.S. 36 (2004).
Curran, J., & Buckleton, J. (2008). A discussion of the merits of random man not excluded
and likelihood ratios. FSI Genetics, 2, 343–​348.
Curran, J., & Buckleton, J. (2010). Inclusion probabilities and dropout. Journal of Forensic
Sciences, 55, 1171–​1173.
Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).
Federal Bureau of Investigation (FBI). (n.d.). Rapid DNA—​CODIS fact sheet. Retrieved
from https://​www.fbi.gov/​services/​laboratory/​biometric-​analysis/​codis/​rapid-​dna.
Finley v. State, 139 So. 3d 940 (Fla. Dist. Ct. App. 2014).
Fonneløp, A., et al. (2017). The implications of shedder status and background DNA on
direct and secondary transfer in an attack scenario. FSI Genetics, 29, 48–​60.
Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).
General Electric Company v. Joiner, 522 U.S. 136 (1997).
International Symposium on Human Identification (ISHI) Conference. (2017, July 24).
Under the microscope—​interview with Bruce Budowle. Retried from https://​www.
ishinews.com/​under-​the-​microscope-​bruce-​budowle-​2/​.
Jacobs, S. (2015, January 5). Judge tosses out two types of DNA evidence used regularly
in criminal cases. N.Y. Daily News. Retrieved from http://​www.nydailynews.com/​
new-​york/​nyc-​crime/​judge-​tosses-​types-​dna-​testing-​article-​1.2065795.
Joh, E. (2006). Reclaiming “abandoned” DNA: The Fourth Amendment and genetic
privacy. Northwestern University Law Review, 100, 857–​884.
Jurilytics (2017, March 2). Daubert and Frye in the 50 states. Retrieved from https://​
jurilytics.com/​50-​state-​overview.
Kaye, D. (2010). The double helix and the law of evidence. Cambridge, MA: Harvard
University Press.
Kittles, R., et al. (2006). Database limitations on the evidentiary value of forensic
mitochondria DNA evidence. American Criminal Law Review, 43, 53–​88.
Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999).
Lepore, J. (2015). On evidence: Proving Frye as a matter of law, science, and history. Yale
Law Journal, 124, 1092–​1158.
Mapp v. Ohio, 367 U.S. 643 (1961).
Maryland v. King, 133 S. Ct. 1958, 569 U.S. _​_​(2013).
McCall, M. (2018, April 18). Two men exonerated by DNA evidence in gang rape sue
Hammond, Indiana State Police. Chicago Tribune. Retrieved from http://​www.
chicagotribune.com/​suburbs/​post-​tribune/​news/​ct-​met-​lawsuit-​against-​indiana-​
police-​st-​0419-​story.html.
Melendez-​Diaz v. Massachusetts, 557 U.S. 305 (2009).

[ 308 ] Challenges and Debates

9
0
3

Moran, G. (2017, October 24). Judge reverses murder conviction, saying crucial DNA
information not disclosed. San Diego Union Tribune. Retrieved from http://​www.
sandiegouniontribune.com/​news/​courts/​sd-​me-​dna-​evidence-​20171024-​story.
html.
Moss, K. (2015). The admissibility of TrueAllele: A computerized DNA interpretation
system. Washington & Lee Law Review, 72, 1033–​1076.
Mueller, L. D. (2008). Can simple population genetic models reconcile partial match
frequencies observed in large forensic databases? Journal of Genetics, 87, 101–​108.
Murphy, E. (2010). Relative doubt: Familial searches of DNA databases. Michigan Law
Review, 109, 291–​347.
Murphy, E. (2015). Inside the cell: The dark side of forensic DNA. New York,
NY: Nation Books.
National District Attorney’s Association. (2016, November 16). Letter to president
regarding PCAST report. Retrieved from https://​csidds.files.wordpress.com/​2016/​
11/​ndaa-​pcast-​response-​final.pdf.
PCAST. (2016). President’s council of advisors on science and technology. Forensic
Science in Criminal Courts: Ensuring Scientific Validity of Feature-​Comparison
Methods. https://​obamawhitehouse.archives.gov/​administration/​eop/​ostp/​pcast/​
docsreports.
People v. Austin, New York Court of Appeals, slip opinion no. 07300, Oct. 19, 2017.
People v. Buza, 4 Cal. 5th 658 (2018).
People v. Dominguez, Super. Ct. of Cal., No. SCD#230596, Order Granting Motion to
Compel Discovery (March 29, 2018).
People v. Hillary, Decision & Order on DNA Analysis Admissibility (Catena, J.). New York,
St. Lawrence County Court No. 2015-​15 (2015).
People v. Lopez, 286 P.3d 469, 494 (Cal. 2012) (Liu, J., dissenting).
People v. Superior Court ex rel. Chubbs, No. B258569, 2015 WL 139069 (Cal. Ct. App.
Jan. 9, 2015).
Perlin, M. (2010, December 29). Explaining the likelihood ratio in DNA mixture
interpretation. Promega Proceedings. Retrieved from https://​www.promega.ee/​~/​
media/​files/​resources/​.../​ishi%2021/​.../​perlin.pdf.
Rainey, J. (2018, April 28). Familial DNA puts elusive killers behind bars: But only 12
states use it. NBCNews. Retrieved from https://​www.nbcnews.com/​news/​us-​news/​
familial-​dna-​puts-​elusive-​killers-​behind-​bars-​only-​12-​states-​n869711.
Roth, A. (2010). Safety in numbers? Deciding when DNA is enough to convict. New York
University Law Review, 85, 1130.
Roth, A. (2013). Maryland v. King and the wonderful, horrible DNA revolution in law
enforcement, Ohio State Criminal Law Journal, 11, 295–​309.
Roth, A. (2017). Machine testimony. Yale Law Journal, 126, 1972–​2053.
Schneckloth v. Bustamonte, 412 U.S. 218 (1973).
Southall, A. (2017, October 19). Using DNA to sketch what victims look like: Some call it
science fiction. New York Times. Retrieved from https://​www.nytimes.com/​2017/​
10/​19/​nyregion/​dna-​phenotyping-​new-​york-​police.html.
Spring Co. v. Edgar, 99 U.S. 645 (1878).
State v. Tester, 968 A.2d 895 (Vt. 2009).
Texas Forensic Science Commission. (2015, August 21). Letter to Texas criminal justice
community. Retrieved from https://​www.tidc.texas.gov/​.../​Memo-​Presentation-​
from-​Texas-​Forensic-​Science-​Commission.pdf.
United States Sentencing Commission (2016/​2018). United States sentencing guidelines.
Retrieved from https://​www.ussc.gov/​.
United States v. Davis, 602 F. Supp. 2d 658 (D. Md. 2009).
United States v. Yee, 134 F.R.D. 161 (N.D. Ohio 1991).
Wexler, R. (2018). Life, liberty, and trade secrets: Intellectual property in the criminal
justice system. Stanford Law Review, 70, 1343–​1429.

A d mi s s ibi l i t y of DNA E v i de n c e i n  C o u r t [ 309 ]

0
1
3

Will, J. (2003). DNA as property: Implications on the constitutionality of DNA dragnets.

University of Pittsburgh Law Review, 65, 129–​143.
Williams v. Illinois, 567 U.S. 50 (2012).
Winkle, K., et al. (2018, June 20). Judge throws out DNA evidence in UT murder suspect’s
case. KXAN. Retrieved from http://​www.kxan.com/​news/​local/​austin/​judge-​
throws-​out-​dna-​evidence-​in-​ut-​murder-​suspect-​s-​case/​1250971292.
Worth, K. (2018, April 19). Framed for murder by his own DNA. Wired Magazine.
Retrieved from https://​www.wired.com/​story/​dna-​transfer-​framed-​murder/​.
Zhang, S. (2018, June 1). How a tiny website became the police’s go-​to genealogy
database. The Atlantic. Retrieved from https://​www.theatlantic.com/​science/​
archive/​2018/​06/​gedmatch-​police-​genealogy-​database/​561695/​.

[ 310 ] Challenges and Debates