Resolving human neuronal herpesvirus reactivation via petabase-scale association studies

Name: Verified Rigor Record — Resolving human neuronal herpesvirus reactivation via petabase-scale association studies
Item: Resolving human neuronal herpesvirus reactivation via petabase-scale association studies
Rating: 100
Author: Αlpha¹

Gutierrez, J. C.; Chen, Y.; Babaian, A.; Dhindsa, R. S.; Lareau, C. A.

Paper source

Read PDF

Αlpha¹ Rigor Review

Rigor Review Verified

Critical

High-priority

Action items to resolve · Critical / High

Confidence: highEngine: 7.6.0Source: full text

This Alpha1 Rigor Review uses AI Rigor Reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.

The paper is a well-executed computational reanalysis with a strong scientific premise, appropriate methods, and transparent reporting. The main weaknesses are incomplete reporting of biological variables (donor demographics not systematically provided) and absence of effect sizes, confidence intervals, and assumption verification for the statistical comparisons. Minor copyedit issues (typos, inconsistencies) should also be corrected.

Three independent rigor evaluations were synthesized. All three were full-text reviews of the same published manuscript. The paper is a computational reanalysis, so several sub-criteria were correctly marked as not applicable (e.g., randomization, blinding, power analysis, replicates). The citation check found no retracted or missing references. Statistics verification checked 2 reported tests (both consistent) — all other statistical claims are unverified. The copyedit pass flagged 14 minor issues.

12 major claims checked against the paper's own evidence: all adequately supported.

partialAbstract, Results paragraphs 2-3Reviewer 2
HSV-1 reactivation is detectable in healthy but not pathological post-mortem human brain tissue.
The claim is supported for the GTEx vs. ROSMAP comparison, but the single AD-protected donor with reactivation complicates the 'not pathological' part.
Evidence: GTEx donors (healthy) show HSV-1 expression; ROSMAP donors (AD pathology) show minimal expression. However, the PSEN1 E280A/APOE3 Christchurch donor had AD pathology but showed reactivation.
“identifying recurrent herpes simplex virus 1 (HSV-1) reactivation in healthy but not pathological post-mortem human brain tissue.”
partialResults, paragraph 6Reviewer 2
The absence of HSV-1 in AD post-mortem brains is due to survivor bias: neurons that reactivate HSV-1 are cleared before autopsy.
The claim is a plausible interpretation consistent with the data, but it is not directly tested or proven. It is presented as a hypothesis.
Evidence: No HSV-1 detected in 499 AD donors' nuclei; only detected in the Christchurch donor where RORB+ neurons were preserved. This is correlative evidence.
“We interpret this pattern as evidence of survivor bias: the neurons that reactivate HSV-1 in AD are cleared before autopsy, rendering viral transcripts undetectable in standard post-mortem cohorts.”
partialDiscussionReviewer 2
RORB+ neurons are selectively vulnerable in AD and may be lost due to HSV-1 reactivation.
The paper shows RORB+ neurons are vulnerable in AD (citing prior work) and that HSV-1 reactivation occurs in these neurons in one protected donor, but the causal link is inferred, not demonstrated.
Evidence: Prior studies (refs 16-18) show RORB+ neuron loss in AD. This study shows HSV-1 in RORB+ neurons in one donor. The connection is correlative.
“our data support a model in which RORB+ neurons in the hippocampus and frontal cortex serve as reservoirs of latent, ultimately reactivated HSV-1. Because these neurons are selectively vulnerable in AD, they are likely lost following reactivation.”
partialResults, paragraph 9Reviewer 3
RORB+ neurons serve as reservoirs of latent, ultimately reactivated HSV-1, and their selective loss in AD explains the absence of detectable HSV-1 in post-mortem AD brains (survivor bias).
The claim of survivor bias is a plausible interpretation, but it is not directly tested: the paper does not demonstrate that RORB+ neurons are lost after reactivation in vivo (only that they are depleted in AD and that reactivation is detected in a protected brain where they are preserved). The model is logically consistent but remains a hypothesis.
Evidence: Results paragraph 9: 'We interpret this pattern as evidence of survivor bias: the neurons that reactivate HSV-1 in AD are cleared before autopsy.' The paper shows that reactivation occurs in RORB+ neurons (supported) and that these neurons are depleted in AD (cited from prior work), but no causal link is experimentally demonstrated.
“We interpret this pattern as evidence of survivor bias: the neurons that reactivate HSV-1 in AD are cleared before autopsy, rendering viral transcripts undetectable in standard post-mortem cohorts”
supportedIntroduction, paragraph 1Reviewer 1
Mounting evidence implicates herpesvirus reactivation in Alzheimer's disease etiology.
The claim is supported by multiple cited references in the introduction.
Evidence: Cited references 4, 5, 6, 7, 8 in the introduction.
“Mounting evidence implicates herpesvirus reactivation in the etiology of Alzheimer’s disease”
supportedResults, paragraph 2 and 6Reviewer 1
HSV-1 reactivation is detected in healthy but not pathological post-mortem human brain tissue.
The claim is supported by the GTEx (healthy donors) and ROSMAP (AD) results showing abundant HSV-1 reads in GTEx and minimal in ROSMAP.
Evidence: Results, Figure 1b: 'top 4 GTEx (all exceeding 10,000 HSV-1 reads) and top ROSMAP library (4 HSV-1 reads)'. Table 1 shows no HSV-1 in AD atlases.
“we observed minimal HSV-1 expression from ROSMAP samples, where the top library had only four assigned HSV-1 reads ()”
supportedResults, paragraph 6Reviewer 1
Integrative single-nucleus analyses resolve direct evidence of HSV-1 expression in RORB+ glutamatergic neurons.
The claim is supported by the single-nucleus RNA-seq analysis of the Christchurch donor, showing HSV-1 expression in a RORB+ neuronal cluster.
Evidence: Results, paragraph 6: 'localized the signal to a specific population of glutamatergic neurons, including 105 HSV-1 super-expressor cells... marked by RORB expression'. Figure 2e-h.
“localized the signal to a specific population of glutamatergic neurons, including 105 HSV-1 super-expressor cells (; ,). Notably, this population is marked by RORB expression”
supportedResults, paragraph 6Reviewer 1
HSV-1 reactivation is essentially absent from post-mortem AD brain samples, consistent with survivor bias.
The claim is supported by the null results in ROSMAP and the AD atlases, and the interpretation as survivor bias is discussed.
Evidence: Results, paragraph 6: 'we did not detect a single high-confidence HSV-1-derived UMI from >185 billion reads' from AD cohorts. Table 1b shows 0 HSV-1 reads.
“we did not detect a single high-confidence HSV-1-derived UMI from >185 billion reads”
supportedDiscussion, paragraph 1Reviewer 1
RORB+ excitatory neurons are selectively vulnerable in AD and may be lost after HSV-1 reactivation.
The claim is supported by references to prior literature (16-18, 19) and the observation that these neurons are preserved in the protected donor.
Evidence: Introduction and Discussion: references 16-18 show RORB+ neurons are depleted in AD. The Christchurch donor had preserved RORB+ neurons.
“Because these neurons are selectively vulnerable in AD - , they are likely lost following reactivation”
supportedResults, paragraph 5Reviewer 2
HSV-1 expression is localized to RORB+ glutamatergic neurons.
Single-nucleus RNA-seq data from the Christchurch donor clearly show HSV-1 expression in a RORB+ glutamatergic neuron population.
Evidence: UMAP embedding and cell type annotation show HSV-1 expression concentrated in RORB+ glutamatergic neurons (Figure 2e-h).
“localized the signal to a specific population of glutamatergic neurons, including 105 HSV-1 super-expressor cells... marked by RORB expression.”
supportedResults, paragraph 5; MethodsReviewer 2
The HSV-1 strain in the Christchurch donor is not attenuated.
Variant analysis found no evidence of attenuating mutations, supporting the claim.
Evidence: 68 high-confidence variants were identified, and ESM predictions showed no outliers in loss-of-function alleles.
“we detected no deleterious variants among high-confidence mutations in the viral genome.”
supportedResults, paragraphs 1-8Reviewer 3
HSV-1 reactivation is detectable in the healthy human brain but not in pathological post-mortem AD brain tissue.
The claim is supported by the data: GTEx bulk RNA-seq shows seven donors with HSV-1 reactivation, while ROSMAP (AD-enriched) shows essentially no HSV-1 expression. The null result is further confirmed in >2.6 million nuclei from 499 AD donors across multiple atlases.
Evidence: Table 1 and results text: 7 GTEx donors with reactivation; top ROSMAP library has only 4 HSV-1 reads; AD atlases show 0 HSV-1 UMIs from >185 billion reads.
“we observed minimal HSV-1 expression from ROSMAP samples, where the top library had only four assigned HSV-1 reads”

Recomputed 2 tests: 2 consistent, 0 inconsistent, 2 via agent-written checks.

Rigor

Eight Dimensions7 pass

12 major claims4 partial8 supported

14 copyedit issues flagged: mostly consistency, typo, clarity.

Checked 33 references: 30 verified — 3 not checked.

Reproducibility

3 data/code links checked; 3 live.

datahttps://logan-search.org/LIVEHTTP 200
Resolved page looks like data.
datahttp://synapse.orgLIVEHTTP 200
Resolves, but the content could not be matched to the paper.
codeGitHubLIVEHTTP 200
https://github.com/clareaulab/ad-hsv-mapping
Resolves to GitHub (code repository).

Alpha¹ Rigor Action ItemsFinal pass before journal submission

This is a post-publication audit. The published paper is robust in its main findings and methodology. The key rigor gaps — incomplete biological variable reporting and absent effect sizes/CI — do not invalidate the primary conclusion (detection of HSV-1 reactivation in one brain) but weaken the formal statistical claims about covariate associations. An informed reader should weigh these reporting gaps as limitations. Minor typographical errors, if corrected, would improve clarity but do not affect the scientific validity. No issues warrant a retraction or formal erratum, but the authors could strengthen the work by adding a supplementary demographics table and effect size estimates.

1.
HIGHrigor
Add a supplementary table systematically reporting donor demographics (age, sex, race/ethnicity, comorbidities) for all cohorts analyzed (GTEx, ROSMAP, Christchurch donor, AD atlases).
Donor metadata are currently reported only for GTEx, leaving demographic information incomplete for other key datasets, which limits assessment of potential confounding variables.
2.
HIGHstatistics
Report effect sizes (e.g., odds ratio for Fisher's test, mean difference with 95% CI for the t-test) for all statistical comparisons in the Methods/Results.
Effect sizes and confidence intervals are standard for quantifying the magnitude and precision of associations and are currently absent, reducing the interpretability of the reported p-values.
3.
HIGHstatistics
Add a statement in the Methods verifying test assumptions (e.g., normality for the t-test, expected cell counts for Fisher's test) or justify their use without formal verification.
Statistical test assumptions are not reported; while the t-test on age is robust, the paper should justify its use or note that assumptions were checked and met.
4.
MEDIUMreporting
Add a brief statement in the Methods describing how outlier libraries or cells were handled in the single-nucleus analysis beyond the defined threshold for super-expressor cells.
Outlier handling is not reported; this is a minor gap for a computational reanalysis but would improve reproducibility.
5.
MEDIUMcopyedit
Fix the typo 'rDNA depletion' to 'rRNA depletion' in the Methods, GTEx and ROSMAP reanalyses section.
This is an internal inconsistency: 'rDNA depletion' is used once where 'rRNA depletion' is used elsewhere, which could confuse readers.
6.
MEDIUMcopyedit
Remove the duplicate URL in the Data Availability section (LoganSearch URL appears twice).
The URL is duplicated, which is a minor formatting error.
7.
MEDIUMcopyedit
Remove the author ORCID string from the Abstract text.
The ORCID appears to be a paste error and is not standard in abstract text.
8.
MEDIUMcopyedit
Rephrase 'the one brain in which we detected... is also the one brain in which' to avoid repetition (e.g., 'The only brain in which we detected HSV-1 reactivation is also the only brain in which the vulnerable RORB+ neurons were preserved').
The current phrasing is repetitive and slightly awkward, though scientifically clear.
9.
MEDIUMcopyedit
Fix grammatical error: change 'reinforcing the interpretation of that the variable presence' to 'reinforcing the interpretation that the variable presence' in Methods, GTEx and ROSMAP reanalyses.
Extra word 'of' is a grammatical mistake.
10.
LOWcopyedit
Standardize hyphenation of 'post-mortem' vs 'postmortem' throughout the manuscript.
Both forms are used inconsistently; standard biomedical usage is 'postmortem'.
11.
LOWreporting
Consider adding a reporting guideline checklist (e.g., STROBE for reanalyses) as a supplementary file to improve transparency, though this is not a formal requirement for computational studies.
While not mandatory, a reporting guideline would demonstrate commitment to transparency.

Full text (28,306 chars)Analyzed: Jul 6, 2026

Αlpha¹ assesses methodological rigor, not the importance of the findings. See how we calibrate →