IN THE COURT OF CRIMINAL APPEALS
OF TEXAS
NO. PD-1079-19
WILBER ULISES MOLINA, Appellant
v.
THE STATE OF TEXAS
ON APPELLANT’S PETITION FOR DISCRETIONARY REVIEW
FROM THE FIRST DISTRICT COURT OF APPEALS
HARRIS COUNTY
HERVEY, J., delivered the opinion of the unanimous Court.
OPINION
This case presents the question of whether the admission of expert testimony about
a DNA-comparison analysis violates the Confrontation Clause when the analysis is based
on computer-generated data from the expert’s laboratory and data from another
laboratory. We agree with the court of appeals that it does not, and we will affirm its
judgment.
FACTS
Molina–2
a. The Offense
On February 27, 2000, the victim and four of her friends traveled from Port Arthur
to Houston to go to the rodeo. They checked into their hotel around 10:00 p.m. and
decided to ride around the Richmond/Westheimer area, which the victim described as a
“drag” that people would “drive up and down.” The group drove around until about 2:30
a.m., at which point they stopped at a 24-hour diner called Mama’s Café so some of them
could go to the restroom before going back to the hotel. When the victim returned and
started to get into her car, a man in a hooded sweatshirt approached her and asked for a
cigarette. Before she could respond that she did not smoke, the man told her, “Let me
have your car,” and she felt something pushing against her side. It was a gun. The man
pushed her into the car and across the center console into the passenger seat. After a
second man entered the car and sat down in the backseat, the first man drove the car
away. While they were driving, the victim was forced into the backseat, and the man in
the backseat sexually assaulted her. The victim testified that the two people who
kidnapped her met up with two more people, and three of them sexually assaulted her at
gun point while she was blindfolded. Eventually they stopped and drove away, leaving
her in an empty field. The victim walked to a nearby business and asked someone to call
the police. When police arrived, they took her to the hospital where a nurse performed an
examination, collected samples, and took the victim’s clothes.
The evidence was outsourced to Reliagene for genetic testing, and the DNA profile
Molina–3
it developed was entered into CODIS, but police were not able to identify a suspect until
2017 when Appellant voluntarily gave a cheek swab to the Houston Police Department.
Appellant was subsequently indicted for aggravated sexual assault and convicted based
on a DNA analyst’s testimony that the profile developed from the victim’s clothing by
Reliagene was probably Appellant’s because the chances that a random person other than
Appellant was the contributor were in the trillions and quadrillions.
b. Forensic Evidence
The sexual assault nurse examiner (SANE) collected samples during the
examination: a reference blood sample, a pulled head and pulled pubic hair, a loose head
hair, vaginal swab(s) and smear(s),1 right and left-hand fingernail scrapings, two cuttings
from an undergarment, a nasal sample, oral swab(s) and smear(s),2 and two items of
“loose evidence collection.” When the SANE kit was sent for testing in 2003, the
biological section of the Houston Police Department Crime Lab was closed due to quality
control issues, so the SANE kit was outsourced to a private-sector laboratory called
Reliagene. In processing the evidence, analysts at Reliagene were able to obtain
epithelial-cell and sperm-cell fractions from the vaginal swab(s) and two undergarment
cuttings. From those, they were able to develop two DNA profiles. The profile developed
from the epithelial-cell fraction found on the vaginal swab(s) was consistent with the
1
The report is not clear about how many vaginal and oral swabs and smears were in the
SANE kit.
2
See supra, note 1.
Molina–4
victim’s known profile, but the other profile was from an unknown donor. A suspect had
not yet been identified.
Thirteen years later, Appellant agreed to give a buccal swab for testing. That swab
was sent to the Houston Forensic Science Center and processed by Lloyd Halsell, III.
Halsell compared the DNA profile he generated from the buccal swab to the DNA profile
developed by Reliagene. His analysis showed that the sperm-cell fraction on the vaginal
swab(s) was unsuitable for comparison due to insufficient data, but the following
information was obtained, which overwhelmingly indicates that the unknown profile
developed by the Reliagene analysts was probably Appellant’s:
Item Tested Results
Epithelial-cell fraction The probability that a randomly chosen unrelated
from undergarment individual would be included as a possible
cutting #1 contributor to this partial DNA profile is
approximately,
• 1 in 170 trillion for Caucasians,
• 1 in 20 quadrillion for African Americans,
• 1 in 26 trillion for Hispanics, and
• 1 in 1.2 quadrillion for Asians
Epithelial-cell fraction The probability that a randomly chosen unrelated
from undergarment individual would be included as a possible
cutting #2 contributor to the major component is
approximately
• 1 in 38 quadrillion for Caucasians,
• 1 in 3.8 quintillion for African Americans,
• 1 in 3.9 quadrillion for Hispanics, and
• 1 in 100 quadrillion for Asians
Molina–5
Sperm-cell fractions The probability that a randomly chosen unrelated
from both undergarment individual would be included as a possible
cuttings contributor to this DNA profile is approximately
• 1 in 38 quadrillion for Caucasians,
• 1 in 3.8 quintillion for African Americans,
• 1 in 3.9 quadrillion for Hispanics, and
• 1 in 100 quadrillion for Asians
c. Trial
The State called Halsell to testify, but the defense objected and argued that Halsell
was merely a surrogate for testimonial statements included in the Reliagene report. In
response, the trial judge allowed the parties to question Halsell outside the presence of the
jury, after which she ruled that Halsell could testify about his own analysis and
conclusions. During the evidentiary hearing, Halsell testified that evidence must be
processed before a DNA profile can be developed and that processing evidence involves
finding areas of interest on the evidence, conducting presumptive tests to find out if
genetic material is present, extracting the material, and amplifying the pertinent genetic
markers. He also testified that his laboratory uses various controls to ensure the reliability
of the data generated in his laboratory. For example, his laboratory uses a reagent blank to
ensure that the chemicals used to process the evidence are not tainted with DNA. He said
that it also uses a known sample during the amplification step to ensure that the genetic
markers were properly amplified. Halsell testified that both techniques were used to
verify the data from Reliagene. He also testified that, if there was an error in processing
the evidence, he would not expect a useable profile to be developed and that another
Molina–6
person’s profile would not be mistakenly generated. Using Reliagene’s
computer-generated data, Halsell said that he was able to independently verify the profile
developed by Reliagene, then compare that profile against the one he developed from
Appellant’s buccal swab. In ruling that Halsell could testify but that the Reliagene report
was inadmissible,3 the judge considered our decisions in Burch v. State, 401 S.W.3d 634
(Tex. Crim. App. 2013) and Paredes v. State, 462 S.W.3d 510 (Tex. Crim. App. 2015).
PROCEDURAL HISTORY
Appellant was indicted for and convicted of aggravated sexual assault. He was
sentenced to 55 years’ confinement. A split panel of the First District Court of Appeals
affirmed his conviction. Molina v. State, 587 S.W.3d 100 (Tex. App.—Houston [1st
Dist.] 2019). Appellant filed a petition for discretionary review, which we granted, asking
whether the court of appeals’s opinion conflicts with our decision in Burch, 401 S.W.3d
at 634.
CONFRONTATION CLAUSE
The Confrontation Clause of the Sixth Amendment guarantees the accused the
right to confront the witnesses against him. U.S. CONST. amend. VI; Crawford v.
Washington, 541 U.S. 36, 42 (2004); Paredes, 462 S.W.3d at 514. The Confrontation
Clause applies to in-court testimony as well as out-of-court testimonial statements.
Testimonial statements are those “that were made under circumstances which would lead
3
The report was included in the record for appellate purposes and was not admitted at
trial.
Molina–7
an objective witness reasonably to believe that the statement would be available for use at
a later trial.” This includes some forensic analyses. Paredes, 462 S.W.3d at 514 (citing
Crawford, 541 U.S. at 52).
a. Burch
In Burch v. State, 401 S.W.3d at 635, Burch and a companion were arrested by a
police officer who saw them in possession of drugs and drug paraphernalia. Burch was
indicted for possession of cocaine with intent to deliver. At trial, the State sought to
introduce a one-page laboratory report asserting that Burch was in possession of cocaine.
The report was signed by the analyst who performed the testing as well as her supervisor.
The State called the supervisor to testify but not the analyst who performed the analysis
because that analyst no longer worked for the laboratory. The supervisor testified that she
“basically double-checked everything that was done,” but there was no evidence that she
participated in the testing or observed it, and she could not confirm that the non-testifying
analyst reached the correct result—that the sample seized from Burch was cocaine. Burch
objected on Confrontation Clause grounds. The trial court overruled his objection and
admitted the report, the physical evidence, and the supervisor’s testimony that the
substance was cocaine. We concluded that Burch was entitled to confront the
non-testifying analyst because her report contained testimonial statements (that Burch was
in possession of cocaine of a particular amount), and the supervisor was only a surrogate
witness for the non-testifying analyst.
Molina–8
b. Paredes
In Paredes, 462 S.W.3d at 99, this Court revisited the question of how the
Confrontation Clause applies to forensic reports. In that case, Paredes and other gang
members broke into an apartment to steal money and drugs. During the robbery, two
occupants were shot and killed. After Paredes and the other gang members left, Paredes
gave another gang member the shirt that he had been wearing, which had blood on it, and
asked her to wash it. Instead of washing it, the other gang member notified the police,
who seized the shirt. The shirt was subsequently sent to a forensic laboratory called
Identigene for DNA testing. Identigene used an assembly-line batch process to “generate
[the] raw DNA data.” The State called the director of Identigene to testify but not the
three analysts who generated the raw DNA data. Paredes objected and argued that he was
entitled to confront the three non-testifying analysts and that the director’s proposed
testimony that the blood on Paredes’s shirt likely came from one of the victims would
violate the Confrontation Clause because she was only a surrogate for the three
non-testifying analysts. The trial court overruled his objection and allowed the director to
testify. The computer-generated DNA data was not admitted into evidence. Paredes was
subsequently convicted of capital murder.
We held that there was no Confrontation Clause violation, reasoning that the DNA
profile developed by the non-testifying analysts was not testimonial because the
computer-generated data stood for nothing without further analysis and that the director
Molina–9
was more than just a surrogate because she “performed the crucial analysis determining
the DNA match and testified to her own conclusions.” We also noted that the reports the
director relied on were not offered into evidence and that the same potential human-error
problems present in Burch were not at play in Paredes because the director could verify
that her conclusions were properly generated. We further noted that the director also
“testified about the safety measures in place at Identigene to detect . . . errors” and that no
profile would be generated “if part of the analysis were done improperly . . . .”
COURT OF APPEALS
a. The Majority
The court of appeals concluded that Paredes and its decision in Garrett controlled.
Garrett v. State, 518 S.W.3d 546, 547 (Tex. App.—Houston [1st Dist.] 2017, pet. ref’d).
In Garrett, the State’s DNA expert, who worked at the Houston Forensic Science Center,
testified that he compared two DNA profiles generated by other analysts in his laboratory
and that the results implicated Garrett. Id. The question was whether admission of the
testifying expert’s testimony violated the Confrontation Clause. Id. The court of appeals,
relying on Paredes, concluded that it did not. It reasoned that the computer-generated data
was non-testimonial, and it noted that the computer-generated DNA data was not
included in the testifying expert’s report and that the underlying reports were not admitted
into evidence. Id. at 555. Garrett nonetheless argued that the Confrontation Clause was
violated because, unlike in Paredes, the testifying expert in his case did not supervise the
Molina–10
analysts who generated the DNA data. Id. at 555–56. The court of appeals disagreed. Id.
at 556. It acknowledged the distinction but explained that the difference was irrelevant
because, as in Paredes, the testifying expert performed the crucial, independent
DNA-comparison analysis, and Garrett was able to cross-examine him about the analysis
and report. Id.
Turning to Appellant’s case, the court of appeals recognized that Paredes and
Garrett were distinguishable, but it nonetheless concluded that they controlled. The issue
was not where the computer-generated DNA data was developed, it explained, because
the data was non-testimonial regardless of where it was generated. The real issue was
whether Halsell performed the crucial DNA-comparison analysis, which he did. The court
of appeals also noted that, as in Paredes and Garrett, no reports written by or
computer-generated data developed by a non-testifying analyst were admitted into
evidence.
The court of appeals was similarly unconvinced by Appellant’s argument that
Halsell lacked personal knowledge about the quality control protocols at Reliagene. It
pointed out that the testifying expert in Paredes also did not have personal knowledge
about the analysts who developed the computer-generated DNA data even though she
supervised them and could testify about the quality control practices of the laboratory in
general. The court of appeals found more convincing the director’s testimony in Paredes
that, if there was a technical mistake processing evidence, no DNA profile would have
Molina–11
been generated rather than the wrong one, and it noted that Halsell gave similar
testimony. Halsell testified that “he found Reliagene’s computer-generated data to be
reliable because he was able to generate a DNA profile based on his independent analysis
of the data” and that, “[i]f Reliagene had not gathered this data in a scientifically reliable
manner,” he “would not expect a profile to be generated.”
b. The Dissent
Justice Countiss dissented. According to her, the majority erred in following
Paredes because, in that case, the director’s testimony about the quality control protocols
used in her laboratory provided indicia of reliability that the evidence was correctly
processed, but Halsell could not provide such testimony. She argued that, unlike in
Paredes, Halsell did not rely solely on the computer-generated DNA data. Rather, she
claimed, “Halsell made clear that his testimony and his own report and conclusions were
reliant upon Reliagene’s independently created work product . . . .” She also argued that
Halsell did not limit himself to testifying about his comparison of the DNA profiles; he
also “certified that the analysis performed by an unknown ReliaGene analyst was accurate
despite his admitted lack of personal knowledge of ReliaGene’s procedures and
processes.”
PARTIES’S ARGUMENTS ON DISCRETIONARY REVIEW
The issue here is whether Burch or Paredes controls. Appellant argues that Burch
controls because Halsell was only a surrogate witness for testimonial statements in the
Molina–12
Reliagene report, and he asserts that he was unable to “explore the types of corruption and
missteps the Confrontation Clause was designed to protect against” because Halsell could
not be cross-examined about Reliagene’s quality control protocols. See Burch, 401
S.W.3d at 637–38. Appellant acknowledges that the Reliagene report was not admitted at
trial, but he contends that Halsell “testified directly from the excluded Reliagene report.”
The State responds that Halsell was not a surrogate because he performed his own
independent analysis and that the raw DNA data developed at Reliagene was not
testimonial because it “stand[s] for nothing on [its] own.” It also asserts that there is no
evidence that Halsell testified directly from the Reliagene report, and it points out that
Appellant provided no record citations to support his claim.4
ANALYSIS
We agree with the court of appeals’s analysis. The Reliagene report is not
testimonial. It reflects that presumptive tests were performed on some of the items of
evidence, that an epithelial-cell fraction was recovered from the vaginal swab(s), that
epithelial-cell and sperm-cell fractions were extracted from the two undergarment
cuttings, that quality control protocols had been employed, and it includes data about the
partial DNA profile. None of those things are “inherently inculpatory or [were] created
for use against [Appellant].” They stand for nothing on their own without additional
4
The State is correct. Appellant included no record citations in his brief. See TEX. R. APP.
38.1(I) (“The brief must contain a clear and concise argument for the contentions made, with
appropriate citations to authorities and to the record.”).
Molina–13
analysis, and even if they had some intrinsic inculpatory value, there was no suspect until
years later when Appellant voluntarily provided a buccal swab. Burch, 401 S.W.3d at 639
(citing Williams, 132 S. Ct. at 2228).
We also conclude that Halsell was not merely a surrogate witness for the
non-testifying analysts. The Reliagene report was only a basis for Halsell’s independent
analysis of the two DNA profiles and his conclusions that the unknown epithelial-cell and
the sperm-cell fractions likely came from Appellant. Also, while Appellant claims that
Halsell testified directly from the Reliagene report in front of the jury, and thereby put its
contents at issue, we have found no such evidence in the record. We also note that the
computer-generated DNA data was not included in Halsell’s report, and the Reliagene
report was not admitted into evidence.
Finally, like the court of appeals, we find Appellant’s argument about Halsell’s
lack of personal knowledge to be unpersuasive. As the court of appeals correctly noted,
Halsell gave testimony similar to the director in Paredes, and their testimony shows that
an important control in the process is that, if there is an error processing evidence, no
profile suitable for comparison would be generated, and that in no case would a
processing error result in another person’s profile being mistakenly generated. As a result,
we conclude that Halsell’s lack of personal knowledge about the quality control protocols
in place at Reliagene does not change the outcome.
Justice Countiss believed that “Halsell did not just rely on raw computer-generated
Molina–14
data from ReliaGene in order to reach his conclusion . . . .” and that Halsell “certified that
the analysis performed by an unknown ReliaGene analyst was accurate despite his
admitted lack of personal knowledge of ReliaGene’s procedures and processes.” We
disagree. While Halsell said that he relied on the Reliagene report and Reliagene case file,
the report included the computer-generated DNA data in written form, and the case file
included worksheets and other raw computer data generated by Reliagene to support the
DNA profile it developed. Halsell did not rely on an independent analysis by a Reliagene
analyst; he “checked their homework.” That is, he used the underlying, raw data to
determine if he could develop the same profile as the analysts at Reliagene, which he did.5
5
The following exchange took place during the evidentiary hearing,
[STATE:] When you say “process,” what’s the difference between the processing
of a sample and the actual analysis of the sample?
[HALSELL:] So, processing would be the physical looking at the evidence to
determine, is there anything to test and then to actually extract the DNA from an
item and send it to all the techniques that are necessary to generate that DNA
profile. Analysis, in that sense, is to look at the data that is used -- you look at the
data that’s used to generate that DNA profile and compare it to, one, determine, is
it interpretable? Is it something we should do a comparison on? And then if it is, to
then look at any known references that we may have and do a comparison with
those references.
[STATE:] Now, when you say “data,” are you referring to, like, the lab report, or
are you referring to computer-generated data?
[HALSELL:] In this instance, I actually looked at printouts of computer-generated
data, the original computer-generated data.
The State also asked Halsell what he relied on in forming his opinion and conclusions, and he
responded simply, “The computer-generated data.” Later, defense counsel asked Halsell whether
he compared the known profile from the buccal swab to the “report from Reliagene Technology,”
Molina–15
Also, we do not understand Halsell to have certified that Reliagene followed proper
quality control protocols when it processed the evidence. He testified that, based on
Reliagene's case file and statements contained therein, Reliagene appeared to use some of
the same controls as his laboratory did and that he thought that the data was reliable
because he was able to generate the same DNA profile.6
The court of appeals was correct that this case is more like Paredes than Burch.
CONCLUSION
Because we agree with the court of appeals that there was no Confrontation Clause
violation, we affirm its judgment.
Delivered: October 20, 2021
Publish
and Halsell answered that the known profile was “compared to the data that supported [the
Reliagene] report.”
6
In response to the prosecutor asking Halsell during the evidentiary hearing whether there
was anything about the data itself that “lets you know that there were certain standards or control
or other things in place to ensure the data is reliable from a scientific perspective,” Halsell said
that,
When I say “review of the data,” and truly what I guess I mean is, I reviewed the
case file, under the controls, meaning any reagent blanks that are associated with
the extractions. So, those are samples that do not contain DNA and are meant to
test the reagents to show that they’re clean during processing. During one step, a
positive control was introduced to show that when you’re copying the DNA and
making numerous, numerous copies of it, that step was correct. And, so, I also
reviewed that to ensure that that worked appropriately, to show that step
functioned the way it should have. And, ultimately, all of those have to be
acceptable for the data to be acceptable and relied upon.