O P I N I O N A jury found Wilber Ulises Molina guilty of aggravated sexual assault and assessed his punishment at 55 years of confinement. Molina appeals, contending: (1) the trial court violated his constitutional right to confront the witnesses against him by allowing an analyst to testify based on DNA testing that was performed by others at an out-of-state laboratory; the evidence is legally insufficient to support his conviction because there is no admissible evidence that he sexually assaulted the complainant; and the prosecutor misstated the law and made improper and prejudicial statements about matters outside the record during closing arguments. Finding no reversible error, we affirm. BACKGROUND In 2000, when the complainant was 23 years old, four men abducted her, at least three of whom then sexually assaulted her at gunpoint. She was blindfolded during the assaults. More than one of the assailants ejaculated while sexually assaulting the complainant. To her knowledge, none of her assailants used a condom. Nor did any of the complainant’s assailants make any effort to remove their semen from her or her clothing after they were done assaulting her. The complainant’s assailants abandoned her afterward. She then sought help and summoned law enforcement. A police officer took her to a hospital, where a nurse completed a sexual-assault kit and took the complainant’s clothes, including her undergarments, to preserve any evidence of the assaults. None of the complainant’s assailants were identified for more than a decade and a half. In 2017, however, Molina voluntarily provided a cheek swab to the Houston Police Department for DNA analysis. A grand jury subsequently indicted Molina for aggravated sexual assault after a comparison of Molina’s DNA profile with a DNA profile generated from semen found in the complainant’s undergarments matched. Molina pleaded not guilty. At trial, the complainant testified that she would not be able to identify any of her abductors. No other witnesses could identify the complainant’s abductors either. The DNA evidence was the sole link connecting Molina to the crime. Motion to Exclude Molina had moved to exclude the DNA evidence, contending that its introduction would violate his constitutional right to confront the witnesses against him. He argued that this was so because the complainant’s undergarments were tested for DNA by an out-of-state laboratory and neither the analyst who performed the test nor any other employee from that out-of-state lab would be testifying. The trial court deferred its ruling pending the testimony of the state’s expert. Evidentiary Hearing The state presented Lloyd Halsell, Operations Coordinator for the Houston Forensic Science Center, as its DNA expert. The trial court then held an evidentiary hearing about the DNA evidence outside the presence of the jury. Halsell testified that the Houston laboratory did not process any DNA evidence in 2003 due to quality-assurance concerns. The sexual-assault kit at issue therefore was sent for processing to Reliagene, an independent laboratory in New Orleans. Reliagene processed the kit and issued a report of its findings the following year. No one at the Houston Forensic Science Center independently processed this evidence. Nor did Halsell supervise Reliagene’s processing of the sexual-assault kit. Halsell explained that processing of the type performed by Reliagene in 2004 consists of physical examination of the evidence to determine if there is any biological material present, the extraction of any DNA from this material, and the application of techniques necessary to generate a profile from the DNA. The processing of evidence differs from its analysis, which entails examination of the data accumulated by processing to generate a DNA profile, if possible, that can then be used for comparison with profiles from other samples. Halsell acknowledged that each laboratory has different standards and protocols, and that he did not know what standards and protocols Reliagene used. Halsell testified, however, that he knew Reliagene was accredited with respect to maintaining the proper quality-assurance standards. He also testified that the paperwork accompanying the processed evidence indicated that Reliagene had applied proper standards to preserve it from contamination and to maintain a proper chain of custody. In 2017, the Houston Forensic Science Center received a cheek swab taken from Molina. The Center processed this swab for DNA. Halsell then analyzed the underlying data generated by Reliagene in 2004 and the Center in 2017. He examined the data to ensure that it was adequate for comparison. While Halsell reviewed and considered Reliagene’s report, he testified that his report was not based solely on Reliagene’s and that his analysis was independent of Reliagene’s. Halsell stated that he reviewed the computer-generated data compiled by Reliagene and that his own report was based on this data. He relied on this computer-generated data in forming his expert opinion in this case. Halsell opined that Reliagene’s data was scientifically reliable. He based this opinion on Reliagene’s paperwork, which documented that it had performed the steps that the Center uses to ensure reliability. His confidence in the reliability of the data was bolstered by his ability to independently analyze the data and generate a DNA profile. Halsell testified that the generation of a DNA profile would have been less likely—”we would not expect a profile to be generated”—if Reliagene had not gathered the underlying data in a scientifically reliable way. After hearing Halsell’s testimony, the trial court ruled that it was admissible. The trial court, however, excluded Reliagene’s report. Halsell’s Trial Testimony Halsell testified about the DNA evidence before the jury. He opined that, based on his comparison of the 2004 and 2017 DNA profiles, Molina could not be excluded as a possible contributor of the DNA in the complainant’s undergarments. In other words, Molina’s DNA profile matched the DNA profile obtained from the complainant’s undergarments. As to the first sample obtained from her undergarments, the probability that a random, unrelated Hispanic male would be included as a possible DNA contributor was 1 in 26 trillion. As to the second sample, the probability was 1 in 3.9 quadrillion. For reference, earth’s population is about seven billion. Jury Verdict The jury found Molina guilty as charged. It assessed his punishment at 55 years’ incarceration. The trial court entered a judgment of conviction in conformity with the jury’s verdict. DISCUSSION I. Confrontation Clause Molina contends that the trial court violated his constitutional right to confront the witnesses against him by allowing Halsell to testify based in part on the DNA testing performed by Reliagene, an independent, out-of-state laboratory. A. Standard of review and applicable law The Confrontation Clause of the Sixth Amendment to the United States Constitution gives a criminal defendant the right to cross-examine the witnesses against him. See generally Crawford v. Washington., 541 U.S. 36 (2004). Thus, testimonial statements of witnesses who do not take the stand at trial cannot be admitted into evidence unless the absent witness is both unavailable and the defendant had a prior opportunity to cross-examine the witness. Id. at 59. But this constitutional rule of exclusion applies only to statements that are testimonial in nature; thus, whether an absent witness’s statement is testimonial is a threshold issue for the court to decide. See id. at 68; Woods v. State, 152 S.W.3d 105, 113 (Tex. Crim. App. 2004). We review de novo a trial court’s constitutional legal rulings, including whether an absent witness’s statement is testimonial and thus barred from evidence. Wall v. State, 184 S.W.3d 730, 742 (Tex. Crim. App. 2006). B. Analysis The question before us is whether the Confrontation Clause bars a testifying expert from relying on computer-generated data gathered by employees of a different laboratory who processed physical evidence for DNA unless those employees also testify. Three decisions inform our analysis—Williams v. Illinois, 567 U.S. 50 (2012); Paredes v. State, 462 S.W.3d 510 (Tex. Crim. App. 2015); and Garrett v. State, 518 S.W.3d 546 (Tex. App.—Houston [1st Dist.] 2017, pet. ref’d). Based on these decisions, we hold that computer-generated DNA data from another lab is not testimonial, and the Confrontation Clause thus does not bar a testifying expert from relying on it even though the persons who accumulated the data do not take the stand and are not subject to cross-examination. Williams v. Illinois In Williams, the Supreme Court faced an issue very like the one before us. In a rape prosecution, the state’s expert testified that a DNA profile produced by an out-of-state laboratory matched a DNA profile produced by the state’s crime lab using a blood sample from the defendant. 567 U.S. at 56, 59 (Alito, J., plurality op.). The expert relied on her own comparison of the two DNA profiles in opining that the defendant could not be excluded as a possible contributor. Id. at 61-62. The out- of-state lab’s report was not admitted into evidence and the state’s expert did not read from it on the stand or identify it as a source of her opinions. Id. at 62. The expert did not conduct or observe the work the out-of-state lab did to generate its DNA profile. Id. The defendant objected based on the Confrontation Clause. Id. In a 5-4 decision, the Court rejected the defendant’s constitutional challenge. But a majority of the Court did not agree on a rationale. Writing for himself and three others, Justice Alito concluded that the expert’s reliance on another lab’s DNA profile was not testimonial either because she merely informed the factfinder of the basis for her opinion, rather than vouching for the profile’s accuracy, or because the other lab made the profile before the defendant was a suspect. Id. at 70-75, 81-84. Writing for herself and three others, Justice Kagan rejected both of these positions. Id. at 125-38 (Kagan, J., dissenting). Justice Thomas, writing for himself, agreed with the dissent’s criticism of the plurality opinion, but nonetheless thought the out- of-state lab’s DNA profile was not testimonial because its report containing the data on which the state’s expert relied lacked the formality and solemnity necessary to render its contents testimonial. Id. at 103 (Thomas, J., concurring in the judgment). As the Court of Criminal Appeals observed, the Supreme Court’s decision in Williams is too fractured to serve as precedent. See Paredes, 462 S.W.3d at 516 (no rule can be derived from Williams); see also Vasquez v. State, 389 S.W.3d 361, 370 (Tex. Crim. App. 2012) (plurality opinions lack precedential value). Williams establishes only that the Supreme Court has yet to resolve the issue before us. Paredes v. State In Paredes, the Court of Criminal Appeals decided whether “the admission of a supervising DNA analyst’s opinion regarding a DNA match” violates the Confrontation Clause “when that opinion is based upon computer-generated data obtained through batch DNA testing.” 462 S.W.3d at 511. Under the circumstances before the Court, it held that the admission of this opinion was constitutional. Id. Paredes involved a murder prosecution, in which the state’s DNA expert was the director of the laboratory that tested the evidence. Id. at 512. Three different lab analysts processed the evidence for DNA. Id. The testifying expert had supervised their work and analyzed the resulting DNA profiles to ascertain whether there was a match. Id. She acknowledged that she did not actually watch the analysts perform their work even though she relied on their raw data. Id. at 512-13. She testified that had there been a problem with their work, her analysis would have produced no result, rather than producing an incorrect result. Id. at 512. None of the three analysts testified at trial. Id. at 513. Nor was their raw data admitted into evidence. Id. Whether the expert wrote her own report was not clear from the record; however, if she did, it was not admitted into evidence either. Id. The defendant objected that he was constitutionally entitled to cross-examine the analysts. Id. After discussing the Supreme Court’s Confrontation Clause jurisprudence and its own decision in Burch v. State, 401 S.W.3d 634 (Tex. Crim. App. 2013), the Court of Criminal Appeals concluded that “several general principles are clear, assuming a defendant was afforded no prior opportunity to cross-examine.” Paredes, 462 S.W.3d at 517. In summary, the Court observed that: the Confrontation Clause renders inadmissible a lab report created solely by an analyst who does not testify at trial; the Confrontation Clause likewise renders inadmissible expert testimony explaining a report solely created by a non-testifying analyst; and an expert may testify based on DNA analysis performed by non-testifying analysts, but only to the extent of the expert’s opinions and conclusions. Id. at 517-18. With these general principles in mind, the Court held that the lab director’s testimony was admissible because she “was more than a surrogate for a non-testifying analyst’s report.” Id. at 518. In reaching its holding, the Court relied in significant part on the expert’s independent analysis of computer-generated data, specifically the DNA profiles. See id. at 518-19. The Court held that prior decisions were distinguishable “because the testifying expert in this case relied upon raw, computer-generated data in reaching her conclusion rather than another laboratory analyst’s report.” Id. at 518. The expert analyzed this computer-generated data, which was “the crucial analysis determining the DNA match” and she “testified to her own conclusions” rather than any conclusions made by the analysts. Id. Without her independent analysis of the DNA profiles, they stood “for nothing on their own.” Id. at 519. The raw data on which the expert relied therefore was not testimonial in nature because it “did not come from a witness capable of being cross-examined” but “from a computer.” Id. The Court also noted that the analysts’ lab reports that the expert relied on for the raw data were not admitted into evidence. Id. at 518. Thus, the expert was not a mere surrogate for otherwise inadmissible testimonial lab reports. Id. The Court rejected the defendant’s contention that the potential for human error—mishandling of samples or misreporting of results by the analysts—changed the constitutional calculus. Id. The lab director testified about the measures in place at the laboratory to detect errors. Id. Moreover, she testified that errors would have resulted in no DNA profile rather than an incorrect profile. Id. Garrett v. State In Garrett, this court rejected a defendant’s contention that the trial court erred in admitting the testimony and report of a DNA analyst in violation of the Confrontation Clause. See 518 S.W.3d at 547. The defendant argued that admission of this evidence violated his right to confront the witnesses against him because two other analysts who processed the evidence for DNA did not testify. Id. at 547, 549. Garrett involved a murder prosecution, in which the state’s DNA expert testified based on his analysis of DNA profiles generated by two other analysts in the Houston Forensic Science Center. See id. at 548, 554. These analysts performed tests that ultimately resulted in computer-generated DNA profiles, which the testifying expert then analyzed to determine whether there was a DNA match. Id. at 555. The court therefore held that Paredes was dispositive because the expert “independently analyzed the raw DNA data and offered his own opinion concerning the comparison of the DNA profiles,” notwithstanding the fact that the expert did not supervise the work of the two non-testifying analysts. Id. at 555-56. Our Case Molina contends that Paredes and Garrett are distinguishable because the experts in those cases at least testified about DNA that was processed in their own labs. In contrast, Molina argues, Halsell had no personal knowledge about the DNA data generated by Reliagene. Given Halsell’s lack of personal knowledge about Reliagene’s data and the absence of the Reliagene analyst who processed the evidence for DNA at trial, Molina maintains he had no means “to explore the types of misconduct and mistakes against which the Confrontation Clause was designed to protect.” We disagree. While neither Paredes nor Garrett involved situations in which an expert relied on data produced by analysts from another laboratory, both cases are materially indistinguishable from this one in that the testifying experts in those cases opined based on their own independent analyses of computer-generated data derived by other analysts who actually processed some of the physical evidence for DNA. See Paredes, 462 S.W.3d at 518-19; Garrett, 518 S.W.3d at 555-56. Parades and Garrett hold that computer-generated DNA data is not testimonial and therefore is not subject to the Confrontation Clause’s cross-examination requirement. See Paredes, 462 S.W.3d at 519; Garrett, 518 S.W.3d at 555. That the data was produced by analysts at a different laboratory does not render it testimonial in nature. As in Paredes and Garrett, the state likewise did not offer into evidence any reports written or raw data compiled by a non-testifying analyst. See Paredes, 462 S.W.3d at 513, 518; Garrett, 518 S.W.3d at 554. Nor did Halsell act as surrogate for a non-testifying analyst’s report by trying to explain the contents of this inadmissible evidence to the jury; instead, he offered his own opinions and conclusions based on his analysis of the underlying data. Halsell’s testimony therefore conformed to the three general principles articulated by the Court of Criminal Appeals in Paredes. See 462 S.W.3d at 517-18. Halsell’s lack of personal knowledge as to Reliagene’s laboratory practices does not alter the analysis. While the supervising analyst in Paredes was from the same lab and therefore familiar with its practices, “she did not physically watch each of the three analysts conduct the DNA testing process.” Id. at 512-13, 518. Thus, the supervising analyst in that case also lacked the personal knowledge necessary to testify as to whether the three analysts mishandled the evidence, made technical mistakes in processing the evidence, or engaged in scientific misconduct. She did, however, testify that if their work had been flawed, the tests would not have generated a DNA profile rather than generating an incorrect one. Id. at 512, 518. Halsell’s testimony was similar. He stated in part 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. If Reliagene had not gathered this data in a scientifically reliable manner, Halsell opined, “we would not expect a profile to be generated.” Thus, like Paredes, this case “does not present the human- error problem” that Molina raises on appeal. See Paredes, 462 S.W.3d at 518. We hold that the computer-generated data on which Halsell relied for his opinion was not testimonial in nature and that Molina’s ability to cross-examine Halsell as to his analysis of the data therefore satisfied the Confrontation Clause. II. Legal Sufficiency Molina contends that the only evidence identifying him as one of the complainant’s assailants is Halsell’s testimony as to the DNA evidence. Since the Confrontation Clause bars that testimony, Molina reasons, there is not legally sufficient evidence of his guilt. Having rejected Molina’s confrontation claim, we also reject his legal- sufficiency challenge. Halsell testified that Molina could not be excluded as a possible contributor of the DNA recovered from the complainant’s undergarments. Statistically, it was exceedingly improbable that the DNA was someone else’s. Standing alone, Halsell’s testimony is legally sufficient to prove beyond a reasonable doubt that Molina was one of the complainant’s assailants under the circumstances of this case. See King v. State, 91 S.W.3d 375, 378-81 (Tex. App.—Texarkana 2002, pet. ref’d) (DNA evidence was sufficient to establish defendant’s identity as rapist even without additional evidence of identity); Roberson v. State, 16 S.W.3d 156, 166-72 (Tex. App.—Austin 2000, pet. ref’d) (same); see also Hinojosa v. State, 4 S.W.3d 240, 245 (Tex. Crim. App. 1999) (DNA analysis showing probability of almost 1 in 20 million that semen found in body of woman who was raped and murdered came from someone other than defendant was impressive evidence supporting jury’s guilty verdict). Molina does not challenge the sufficiency of the evidence in any other respect. Accordingly, we hold that legally sufficient evidence supports the jury’s determination that Molina sexually assaulted the complainant. III. Closing Arguments Molina contends that the trial court erred in overruling three objections his counsel made during the state’s closing argument. The trial court should have sustained these objections, Molina argues, because in the first instance the state misstated the law and in the second two instances the state introduced matters outside the record. A. Standard of review and applicable law We review a trial court’s rulings on objections as to the proper scope of closing arguments for an abuse of discretion. See Milton v. State, 572 S.W.3d 234, 240 (Tex. Crim. App. 2019); Vasquez v. State, 484 S.W.3d 526, 531 (Tex. App.— Houston [1st Dist.] 2016, no pet.). Counsel cannot argue contrary to the law, of course. See Davis v. State, 329 S.W.3d 798, 825 (Tex. Crim. App. 2010). But counsel are entitled to correctly argue the law, even if the law is not included in the jury charge. State v. Renteria, 977 S.W.2d 606, 608 (Tex. Crim. App. 1998); Vasquez, 484 S.W.3d at 531. As to the facts, while counsel are bound by the record, they are afforded wide latitude in argument and may draw all reasonable, fair, and legitimate inferences from the evidence. Sterling v. State, 830 S.W.2d 114, 120 (Tex. Crim. App. 1992); Williams v. State, 417 S.W.3d 162, 174 (Tex. App.—Houston [1st Dist.] 2013, pet. ref’d). Improper jury argument is not a basis for reversal unless, when viewed in light of the record as a whole, it is extreme or manifestly improper, violates a mandatory statute, or injects new facts adverse to the defendant into the trial. Borjan v. State, 787 S.W.2d 53, 57 (Tex. Crim. App. 1990) (per curiam); Stout v. State, 426 S.W.3d 214, 220 (Tex. App.—Houston [1st Dist.] 2012, no pet.). Even then, an improper argument that is not constitutional in nature, such as one that misstates the facts, cannot serve as a basis for reversal unless it also affects the defendant’s substantial rights. Freeman v. State, 340 S.W.3d 717, 728 (Tex. Crim. App. 2011). An error affects a defendant’s substantial rights if it has a substantial and injurious influence on the jury’s verdict. Coble v. State, 330 S.W.3d 253, 280 (Tex. Crim. App. 2010). An error does not affect the defendant’s substantial rights if the record as a whole shows that the error either did not influence the jury or influenced it only slightly. Barshaw v. State, 342 S.W.3d 91, 93 (Tex. Crim. App. 2011). B. Analysis 1. Misstatement of the law At trial, there was a dispute as to whether Molina went by the name Carlos, as one of the complainant’s assailants referred to another of her abductors by this name. In closing, the state argued that this dispute did not create reasonable doubt about Molina’s guilt. The state then asserted that the law did not require the complainant to testify that she saw Molina during the assault and could identify him or confirm that another of her attackers referred to Molina as Carlos. Defense counsel objected that this was “a misstatement of the law or the elements” that the state had to prove to secure a conviction. Molina does not explain on appeal how the state misstated the law. We perceive no misstatement. Eyewitness identification is not required, and DNA evidence alone may establish an assailant’s identity in a rape prosecution. See, e.g., Roberson, 16 S.W.3d at 159 (affirming aggravated sexual assault conviction based on DNA even though complainant couldn’t identify defendant and there was no other circumstantial evidence linking him to crime). The state correctly argued the law applicable to the case, which it was entitled to do. See Renteria, 977 S.W.2d at 608; Vasquez, 484 S.W.3d at 531. The trial court therefore did not abuse its discretion in overruling Molina’s objection that the state misstated the law. 2. Misstatement of the facts The state argued in closing that Reliagene did not conduct the DNA analysis of the evidence and that the analysis instead was made by Halsell. It further argued that Halsell’s analysis resulted in a finding that the semen in the complainant’s undergarments was Molina’s “to 3.9 quadrillion odds.” Defense counsel objected that both of these arguments misstated the facts. Molina does not explain on appeal how these arguments misstate the evidence. The state’s contention that Halsell analyzed the DNA evidence, rather than Reliagene, is consistent with Halsell’s testimony distinguishing between Reliagene’s processing of the physical evidence for DNA and his own analysis of the data derived from processing the evidence. The state’s argument that Halsell performed the DNA analysis was not outside the record, and the trial court thus did not abuse its discretion in overruling Molina’s objection that the state misstated the facts. See Sterling, 830 S.W.2d at 120; Williams, 417 S.W.3d at 174. The state’s argument about the semen found in the complainant’s undergarments, in contrast, was improper. The prosecutor argued that Halsell’s analysis “found [Molina's] sperm to 3.9 quadrillion odds in her panties.” While Halsell did calculate a probability of one in 3.9 quadrillion, the state misstated its significance. As a scientific treatise explains with respect to DNA evidence: The random-match probability is the probability that the suspect has the DNA genotype of the crime scene sample if he is not the true source of that sample (and is unrelated to the true source). The tendency to invert or transpose the probability—to go from a one-in-a-million chance if the suspect is not the source to a million-to-one chance that the suspect is the source is known as the fallacy of the transposed conditional. David H. Kaye & George Sensabaugh, Reference Guide on DNA Identification Evidence, in Reference Manual on Scientific Evidence 129, 168 (3d ed. 2011); see also McDaniel v. Brown, 558 U.S. 120, 128 (2010) (per curiam) (referring to fallacy of transposed conditional as “prosecutor’s fallacy” and discussing same). Statistically, this kind of transposition error is not a distinction without a difference. See Kaye & Sensabaugh, supra, at 168-69; Wilson v. State, 185 S.W.3d 481, 489 (Tex. Crim. App. 2006) (Johnson, J., concurring) (discussing erroneous transposition of probability statistics concerning DNA evidence). Thus, in the right case, this error could constitute the injection of new—and mistaken—adverse facts that substantially affect the defendant’s rights and thus require reversal. We therefore must decide whether the state’s misstatement of fact warrants reversal when viewed in the context of the present record. At the outset, we acknowledge that the state did not merely misstate the DNA evidence in passing. The prosecutor did so at length—without objection—earlier in his closing argument: We took her panties. And what did we find? Well, we found a sperm fraction from this defendant, 3.9 quadrillion, 3.9 quadrillion. I’m sure a lot of you probably have never even heard the number quadrillion before. I tried to explain a little with the DNA analyst. The earth has however many billions of people. Then when you saw him trying to write, that’s a lot more zeroes. We’re talking about more earths than you could ever even conceive of. And, again, that’s the odds that you would just randomly—oh, the worst of luck, it is your DNA. Well, this defendant must have the worst luck in the entire—but you can’t say world because I guess it would be about a million worlds. He’s got the worst luck in about a million worlds that we would find his sperm in her panties. In addition, the state’s mischaracterization of Halsell’s testimony in closing argument assumes particular significance in this case because the DNA evidence at issue was the sole evidence linking Molina to the charged offense. The record also shows, however, that the state’s mischaracterization of the DNA evidence did not impact the jury’s verdict. During its deliberations, the jury requested that the court give it a transcript of Halsell’s testimony explaining the 3.9 quadrillion statistic, which indicates that the jury had not yet decided what this statistic proved. See Moore v. State, 658 S.W.2d 312, 315 (Tex. App.—Houston [1st Dist.] 1983, pet. ref’d) (jury note requesting information it thought relevant to its decision indicated that it had not formed an opinion on the matter). Without objection from the state or defense, the court gave the jury a four-page excerpt of Halsell’s testimony on this subject. Thus, the prosecutor’s argument was not the last word on the statistic; the jury had Halsell’s actual testimony before it and did not simply rely on the state’s misstatement. The jury’s request and the trial court’s response show that the state’s misstatement of fact as to the DNA evidence did not substantially affect Molina’s rights. See id. (trial court’s instruction in response to jury note cured error); see also Bargas v. State, 252 S.W.3d 876, 902 (Tex. App.— Houston [14th Dist.] 2008, no pet.) (trial court’s answer to jury note asking question about law omitted from charge rendered its omission from charge harmless); Salinas v. State, 652 S.W.2d 468, 469 (Tex. App.—Corpus Christi 1983, no pet.) (trial court’s instruction to consider only evidence introduced at trial in response to jury note indicating that jury was considering matters outside the record cured any error). Moreover, Molina does not contend on appeal that Halsell himself misstated the significance of the 3.9 quadrillion statistic. Halsell’s testimony, properly understood, was powerful evidence connecting Molina to the crime. See Hinojosa, 4 S.W.3d at 245 (“Despite one in 19,900,000 odds, appellant’s DNA profile matched the semen found in the complainant. Contrary to appellant’s argument, these impressive statistics support the jury’s conclusion that appellant, as opposed to some unidentified ‘suspect’ also sharing the same DNA profile, sexually assaulted, kidnapped, and killed Wright.”); Roberson, 16 S.W.3d at 167-68 (characterizing probability of 1 in 420 billion as “strong evidence”). Halsell testified that the probability that Molina had the same DNA profile recovered from the semen in the complainant’s undergarments—but was neither the actual contributor of that semen nor related to the actual contributor—was one in 3.9 quadrillion. Based on Halsell’s testimony, reasonable jurors could find that Molina was one of the complainant’s assailants beyond a reasonable doubt. See King, 91 S.W.3d at 378-81; Roberson, 16 S.W.3d at 162. We hold that the trial court abused its discretion by not sustaining Molina’s objection to the state’s improper jury argument, but that the state’s improper jury argument and the trial court’s erroneous ruling concerning it did not affect Molina’s substantial rights and therefore are not reversible error. See Freeman, 340 S.W.3d at 728; see also Barshaw, 342 S.W.3d at 93; Coble, 330 S.W.3d at 280. CONCLUSION We affirm the judgment of the trial court. Gordon Goodman Justice Panel consists of Chief Justice Radack and Justices Goodman and Countiss. Justice Countiss, dissenting. Publish. Tex. R. App. P. 47.2(b).
