Date: May 23, 2026

“Shifted Science” DNA Reinterpretation Does Not, Without Trial-Stage Materiality, Excuse Procedural Default of Brady/Napue Claims on Federal Habeas

1. Introduction

Case: Herod v. Guerrero, No. 25-40247 (5th Cir. May 20, 2026) (per curiam) (unpublished).
Parties: Richard Anthony Herod (state prisoner; federal habeas petitioner; Plaintiff—Appellee) vs. Eric Guerrero, Director, Texas Department of Criminal Justice, Correctional Institutions Division (Respondent; Defendant—Appellant).
Background: Herod was convicted in Texas state court (2012) of aggravated sexual assault and aggravated robbery arising from a home invasion in Texas City. A key trial feature was expert DNA mixture testimony regarding a white t-shirt used to blindfold a victim. Years later, Texas DPS issued a supplemental report (2017) stating that under updated DNA-mixture interpretation protocols, Herod was excluded as a contributor to that mixture.

Core issues on appeal: Whether Herod could overcome state procedural default of due process claims under Brady v. Maryland and Napue v. Illinois by showing prejudice (i.e., materiality/actual substantial disadvantage) from (a) allegedly suppressed evidence about the DNA interpretation and (b) allegedly false DNA testimony/statistics.

2. Summary of the Opinion

The Fifth Circuit reversed the district court’s grant of § 2254 habeas relief on Herod’s Brady and Napue theories. Although the State conceded “cause” and all agreed the claims were procedurally defaulted (the Texas Court of Criminal Appeals dismissed the second state habeas application on procedural grounds), the panel held Herod failed to establish “prejudice” sufficient to excuse the default. The court emphasized that the DNA evidence—especially as presented at trial—was of limited inculpatory value, was already substantially contested by the defense expert, and was overshadowed by other inculpatory proof (eyewitness identification, accomplice/girlfriend testimony, and cell-phone/cell-tower evidence). The case was remanded for proceedings on the remainder of Herod’s habeas petition.

In a concurrence, Judge Engelhardt (joined by Judge Wilson) added that, even if default were excused, one of Herod’s Brady theories is independently barred by Teague v. Lane because it would require a “new constitutional rule” obligating disclosure of a “fact” (Herod’s exclusion) that was not known until years after trial, when science/protocols changed.

3. Analysis

A. Procedural posture, AEDPA frame, and the controlling “gateway” test

The opinion situates the dispute within AEDPA’s structure: federal habeas relief under 28 U.S.C. § 2254 is available only for custody “in violation of the Constitution.” Ordinarily, § 2254(d) deference applies to claims “adjudicated on the merits” in state court, but the panel relied on Nelson v. Davis to note that where the state court did not reach the merits (here, a procedural dismissal), federal review is barred unless the petitioner satisfies the “cause and prejudice” gateway.

The court adopted the Fifth Circuit’s established mapping of Brady to the gateway: per Prible v. Lumpkin, Brady suppression corresponds to “cause” and Brady materiality corresponds to “prejudice.” The State conceded cause; the decisive question became whether the asserted Brady/Napue problems created the kind of “actual and substantial disadvantage” required by Smith v. Quarterman, beyond the “possibility of prejudice” rejected in Hughes v. Quarterman, and amounting to the “pervasive actual prejudice”/fundamental unfairness described in Murray v. Carrier.

B. Precedents cited (and how they drive the result)

Brady v. Maryland, 373 U.S. 83 (1963): Supplies the due process framework for suppression of favorable material evidence. Here, its “materiality” concept becomes the prejudice inquiry for excusing default.
Napue v. Illinois, 360 U.S. 264 (1959): Prohibits the State from obtaining a conviction using false testimony (including leaving false impressions uncorrected). Herod argued the analyst’s “non-exclusion” framing and statistics were false/misleading given later DPS reevaluation.
Prible v. Lumpkin, 43 F.4th 501 (5th Cir. 2022): Controls standards of review (de novo on law; clear error on facts) and articulates the cause/prejudice gateway for defaulted claims, including how Brady can supply both.
Grace v. Hooper, 123 F.4th 800 (5th Cir. 2024): Confirms the appellate court applies the same standards to the state court decision as the district court did.
Nelson v. Davis, 952 F.3d 651 (5th Cir. 2020): Explains that where claims were not adjudicated on the merits in state court, deferential § 2254(d) review does not apply—yet procedural default still blocks merits review absent cause and prejudice.
Canales v. Stephens, 765 F.3d 551 (5th Cir. 2014): Defines “cause” as an external objective factor impeding presentation; relevant mainly because the State conceded cause.
Smith v. Quarterman, 515 F.3d 392 (5th Cir. 2008); Hughes v. Quarterman, 530 F.3d 336 (5th Cir. 2008); Murray v. Carrier, 477 U.S. 478 (1986): Supply the demanding prejudice threshold (actual, substantial, pervasive; not merely possible).
Floyd v. Vannoy, 894 F.3d 143 (5th Cir. 2018) (per curiam): Requires assessing Brady materiality “collectively, not item by item,” which the panel applied to the withheld “scientific debate” and the later DPS exclusion.
United States v. Brumfield, 89 F.4th 506 (5th Cir. 2023): Reinforces that Brady materiality is not a sufficiency-of-the-evidence test; the question is trial fairness and confidence in the verdict.
LaCaze v. Warden La. Corr. Inst. for Women, 645 F.3d 728 (5th Cir. 2011): Supplies the “verdict worthy of confidence” fairness framing.
Rocha v. Thaler, 619 F.3d 387 (5th Cir. 2010): Emphasizes materiality depends heavily on the withheld evidence’s value relative to the State’s other evidence—central to the panel’s conclusion that non-DNA evidence independently anchored guilt.
Miller v. Dretke, 404 F.3d 908 (5th Cir. 2005); Kyles v. Whitley, 514 U.S. 419 (1995): Provide the “reasonable probability” and “undermine confidence” materiality standards the panel used to reject prejudice.
Ex parte Torres, 943 S.W.2d 469 (Tex. Crim. App. 1997) (en banc): Distinguishes CCA “denial” (merits) from “dismissal” (procedural), confirming default.
Teague v. Lane, 489 U.S. 288 (1989): In concurrence, invoked to bar a novel Brady obligation premised on post-trial scientific change (“shifted science”), even if default were excused.
Matthew v. Johnson, 201 F.3d 353 (5th Cir. 2000); Lambrix v. Singletary, 520 U.S. 518 (1997); Graham v. Collins, 506 U.S. 461 (1993); Caspari v. Bohlen, 510 U.S. 383 (1994): Provide the concurrence’s three-step Teague framework and finality date concept.
Blakenship v. Johnson, 118 F.3d 312 (5th Cir. 1997); Jackson v. Johnson, 217 F.3d 360 (5th Cir. 2000); Jones v. Davis, 890 F.3d 559 (5th Cir. 2018): Support the concurrence’s point that courts often apply Teague even if the State raises it late, absent “compelling reasons.”
Hughes v. Dretke, 412 F.3d 582 (5th Cir. 2005); United States v. London, 937 F.3d 502 (5th Cir. 2019); Chaidez v. United States, 568 U.S. 342 (2013): Used by the concurrence to define when a rule is “new” and when applying general standards does (or does not) create a new rule.
United States v. Bagley, 473 U.S. 667 (1985); Giglio v. United States, 405 U.S. 150 (1972); Cobb v. Thaler, 682 F.3d 364 (5th Cir. 2012): Reconfirm that Brady encompasses impeachment material and sets out the familiar elements.
West v. Johnson, 92 F.3d 1385 (5th Cir. 1996): Key to the concurrence: Brady concerns information known to the prosecution but unknown to the defense; it does not compel disclosure of later-emerging scientific conclusions not known at trial.
United States v. Runyan, 290 F.3d 223 (5th Cir. 2002): Used by the concurrence to suggest that publicly available scientific debate may not count as “suppressed.”
Saffle v. Parks, 494 U.S. 484 (1990); Gilmore v. Taylor, 508 U.S. 333 (1993); Whorton v. Bockting, 549 U.S. 406 (2007); Gideon v. Wainwright, 372 U.S. 335 (1963); Edwards v. Vannoy, 593 U.S. 255 (2021): Establish the narrowness (practical unavailability) of Teague exceptions.
Vega v. Johnson, 149 F.3d 354 (5th Cir. 1998): Noted in the concurrence in discussing the relationship between merits uncertainty and “new rule” concerns (while acknowledging Lambrix keeps Teague analytically distinct).

C. Legal reasoning: why the panel found no prejudice (and thus no gateway)

The panel’s prejudice analysis turned on a pragmatic reconstruction of what the jury actually heard about the t-shirt DNA and how much the disputed points could have changed the deliberative picture.

The “DNA as presented” was already weak and contested.
The prosecution previewed “DNA profile was present,” but the expert (Browder) did not testify that Herod’s DNA was on the shirt; she testified only that he “could not be excluded” from a mixture. She conceded on cross she could not give “certainty” in mixtures. The key statistic for Herod’s non-exclusion was comparatively unimpressive: 1 in 87 Caucasians could fit that allele combination. Meanwhile, the blindfolded victim (Ronnie) had far stronger mixture consistency on the shirt with multi-million probabilities—highlighting that the shirt was dominated by other contributors.

Crucially, the defense expert (Ketchum) agreed Herod “technically” could not be excluded but opined he “should be excluded” in a realistic/interpretive sense, challenged the marker accounting, and flagged an unknown contributor, thereby already providing the jury a pathway to discount Browder’s interpretation.
The allegedly suppressed “updated exclusion” and “scientific debate” would not have moved the needle given the trial posture.
The panel held that adding (i) the later DPS exclusion under updated protocols, (ii) generalized scientific dispute about mixture interpretation methodology, and (iii) suggestions of “difficulty with statistics” would not reasonably change the verdict because the defense had already delivered the functional message: this is a low-power mixture, and even under the State’s numbers the probative value is limited (1 in 87 among Caucasians).

Put differently, the opinion treats the new information as largely cumulative of the skepticism the defense successfully introduced at trial, rather than as the sort of withheld fact that would reframe the whole case under Kyles v. Whitley.
Other evidence of guilt was substantial and independent of the DNA.
Applying Rocha v. Thaler’s comparative-evidence focus, the panel emphasized: (a) Ronnie identified Herod; (b) Alissia identified her assailant as the “white male”; (c) Holly Kelly (Herod’s girlfriend) testified she drove Herod to Texas City, saw “beanies” and guns, and that he returned with codeine and cash; and (d) cell-phone and cell-tower evidence aligned a phone’s travel with the timeline and route of the crime and supported the victim’s testimony about receiving calls from the assailant’s voice.
Result: no “reasonable probability” of a different outcome; no “actual and substantial disadvantage.”
Under Miller v. Dretke and Kyles v. Whitley, the panel found no reasonable probability that disclosure would have changed the result or undermined confidence in the verdict. Therefore Herod could not satisfy the gateway prejudice requirement under Prible v. Lumpkin, so federal merits review remained barred by procedural default.

D. The concurrence’s additional rule: Teague limits “shifted science” Brady theories

Judge Engelhardt’s concurrence is significant for future forensic-change litigation on habeas. It argues that one Brady theory—framed as suppression of “the fact” Herod was excluded from the mixture—would require a constitutional extension: compelling disclosure of something the State did not know at trial and could not have known until 2017 when protocols changed.

Using Caspari v. Bohlen, the concurrence fixes finality in June 2014. It then reasons (via West v. Johnson) that Brady is about post-trial discovery of information “known to the prosecution” at the time; imposing a duty to disclose later-generated scientific conclusions would be “new” under Teague v. Lane. The concurrence would bar that theory absent an exception, and it concludes neither exception (substantive decriminalization/punishment limitations; “watershed” procedure) applies, relying on Saffle v. Parks, Gilmore v. Taylor, Whorton v. Bockting, Gideon v. Wainwright, and Edwards v. Vannoy.

The concurrence separately suggests that the “scientific debate” suppression theory is less likely to be Teague-barred because it may merely apply existing “general” disclosure principles (see Chaidez v. United States), though it flags the merits hurdle that publicly available debate may not be “suppressed” under United States v. Runyan.

E. Impact

Procedural-default gateways remain demanding even in forensic-evolution cases. The decision underscores that later forensic reassessments do not automatically translate into habeas “prejudice”; petitioners must show the new information would have materially altered the trial’s fairness picture in light of all evidence.
Mixture-DNA “non-exclusion” evidence is treated as comparatively low weight when the trial record already framed it as probabilistic and contested. Where the defense already exposed the limitations (low random match stats, multiple contributors, interpretive uncertainty), later “updated protocol” exclusions may be deemed cumulative for materiality/prejudice.
Teague may independently block “shifted science” reframings of Brady. The concurrence signals a doctrinal constraint: characterizing later scientific developments as “suppressed facts” risks being treated as an attempt to announce a new constitutional disclosure duty on collateral review.
Practical litigation consequence: Future petitioners will likely need to (i) tie the “shift” to information actually known (or reasonably knowable) to the prosecution at trial, or (ii) reframe claims toward other theories (e.g., newly discovered evidence standards in state court, actual-innocence gateways where available, or ineffective assistance where counsel failed to exploit then-available scientific critiques).

4. Complex Concepts Simplified

Procedural default: When a state court refuses to consider a claim because the prisoner violated a state procedural rule (e.g., successive petition rules). Federal habeas courts usually cannot review that claim unless the prisoner clears a gateway.
Cause and prejudice: The gateway to federal review of defaulted claims. “Cause” is an external reason the claim wasn’t raised properly; “prejudice” is a showing that the error had a substantial, outcome-significant effect that rendered the trial fundamentally unfair.
Brady claim (suppression/materiality): A due process violation occurs if the prosecution suppresses evidence favorable to the defense and the evidence is “material”—meaning there is a reasonable probability the result would have been different had it been disclosed.
Napue claim (false testimony): Due process is violated if the State uses false or misleading testimony (or fails to correct it) in a way that matters to the verdict.
DNA mixture & “cannot be excluded”: In mixed DNA samples from multiple contributors, analysts may be unable to definitively “include” someone. “Cannot be excluded” means the person’s profile is consistent with the observed mixture at enough locations, but it is not the same as saying the person’s DNA is definitively present.
Random match probability (e.g., “1 in 87”): A statistic describing how common a compatible genetic pattern is in a population. A low denominator like 87 suggests a relatively common pattern and therefore weaker evidentiary weight.
Teague non-retroactivity: On federal habeas, courts generally cannot apply “new” constitutional procedural rules to convictions that became final before those rules were announced. The concurrence suggests that requiring Brady disclosure of later-developed scientific conclusions would be such a new rule.

5. Conclusion

Herod v. Guerrero reinforces that, for procedurally defaulted due process claims, the decisive question is often not whether later developments cast doubt on an evidentiary point, but whether those developments would have meaningfully altered the jury’s overall assessment in light of the entire record. The Fifth Circuit held Herod could not show that the later DPS DNA reinterpretation (and related methodological critiques) created a reasonable probability of a different outcome, given the weak and already-contested nature of the mixture testimony and the State’s substantial non-DNA evidence. The concurrence further highlights a structural limit: under Teague v. Lane, federal habeas is generally not the forum to constitutionalize new disclosure obligations based on “shifted science” discovered years after trial.