Rare copy number variation in Post-traumatic stress disorder using whole genome sequencing and biobank data
Kramer, S., Farajzadeh Valilou, S., Singh, M., Singh, M., Vieno, S., Peterson, R. E., Chatzinakos, C., Webb, B. T., et al.
genetic and genomic medicine
Quick verdict
First multi-ancestry WGS rCNV study of PTSD reports deletion-count burden; findings suggestive but hinge on small CNVs and a non-standard meta-analysis.
Summary
This preprint reports the first multi-ancestry analysis of rare copy number variants (rCNVs; MAF < 1%) in PTSD using whole-genome sequencing. Drawing on the All of Us program, the authors defined cases by electronic health records (at least two OMOP PTSD codes), yielding 1,179/504/418 cases and 33,689/12,216/11,935 controls in European-like, African-like, and Admixed American ancestry groups. Using Firth logistic regression at region, genome-wide, and gene-set levels, plus a weighted Stouffer meta-analysis across ancestries, they found deletion-count burden significantly associated with PTSD for all-rCNV lengths in EUR-like (P = 5.99 × 10⁻⁶, FDR = 3.59 × 10⁻⁵) and AFR-like (P = 0.005) cohorts, while analyses restricted to CNVs ≥ 10 kb — the range comparable to prior microarray work — were null. Cross-ancestry meta-analysis identified 11 gene-sets (FDR < 0.05), led by a developing-brain expression set, plus a few small region-level hits (KLF18, RBBP8NL). Key limitations: no orthogonal validation of very short CNV calls, modest case numbers, controls not screened for trauma exposure, mouse 'negative control' gene-sets among significant meta-analysis hits, and a non-standard absolute-z meta-analysis with unclear null calibration. The authors themselves label the findings preliminary; larger biobank replication is required.
Review complete
Assessment at a glance
5
risk flags identified
● high 1
● medium 3
● low 1
4 strengths · 5 weaknesses · 5 author questions
AI text signal = Binoculars zero-shot detector score over 20 sampled passages (Fast-DetectGPT co-reported) — absence does not prove human authorship, and a flag does not imply misconduct.
Levels reflect verification complexity — not a quality score. Evolve never rates your paper.
Key claims under review
- Genome-wide rare deletion burden is significantly associated with EHR-defined PTSD in EUR-like (P = 5.99 × 10⁻⁶, FDR = 3.59 × 10⁻⁵) and AFR-like (P = 0.005, FDR = 0.03) cohorts when all WGS-detected rCNV lengths are analyzed.
- Cross-ancestry weighted-Stouffer meta-analysis identifies 11 neurodevelopmental-related gene-sets enriched for rCNV count (FDR < 0.05), led by the developing-brain expression set BspanML_lg2rpkm0.93.
- No significant associations survive at the ≥10 kb threshold comparable to prior microarray work, placing the detected signal in small CNVs visible only to sequencing.
- A small number of ancestry-specific region-level signals (1p31.1/KLF18, 20q13.33/RBBP8NL, 13q31.1, 22q12.1, 4q35.2) implicate neurodevelopmentally annotated loci, though the authors deem them preliminary.
Strengths
First WGS-based multi-ancestry rCNV analysis of PTSD
Uses 97,061 AoU participants with GATK-SV calls (2,101 cases / 57,840 controls across EUR-like, AFR-like, AMR-like), extending the single prior array-based study (Maihofer et al. 2022) to the 71% of rare CNVs <10 kb that arrays miss.
Empirically justified ancestry-specific variant handling
Supplementary Fisher's exact tests show ~2.6-fold rCNV frequency divergence between AFR-like and EUR-like cohorts; ancestry-specific MAF filtering, within-ancestry PCs, and Cochran's Q / I² heterogeneity checks (all I² < 37% in Table 2) support the design.
Null results reported with the same rigor as positives
Appropriate sparse-event statistics and transparency
Weaknesses
Non-standard meta-analysis underpins the headline gene-set claims
No orthogonal validation; region hits are tiny CNVs with implausibly large ORs
Negative-control gene-sets appear among significant meta results
EHR phenotype misclassification risk
Inconsistent correction schemes and unstable diagnostics
Full Review Report
$1.99 · one-timeUnlock the complete preliminary analysis Evolve produced for this paper.
Report preview
Scope & Claims
This medRxiv preprint (posted 2026-09-17, not peer reviewed) claims the first multi-ancestry analysis of rare copy number variants (rCNVs, MAF < 1%) in PTSD from whole-genome sequencing (WGS). The authors (VCU/KCL/SUNY Downstate) test two hypotheses: that WGS-scale rCNV detection identifies PTSD-associated loci and gene-sets in All of Us (AoU), and that cross-ancestry meta-analysis increases power. Headline claims: (1) genome-wide deletion-count burden is associated with EHR-defined PTSD in EUR-like (P = 5.99×10⁻⁶, FDR = 3.59×10⁻⁵) and AFR-like (P = 0.005, FDR = 0.03) cohorts; (2) meta-analysis identifies 11 gene-sets enriched for rCNV count (FDR < 0.05); (3) array-comparable rCNVs (≥10 kb) yield nothing significant. Cases are narrow EHR phenotypes (≥2 OMOP code 436676); controls have EHR data and zero codes.
The remaining sections cover Methodology Assessment, Statistical Evidence, Results & Reproducibility, Limitations & Risks, and Verdict & Recommendations.
- All 4 strengths & 5 weaknesses, analyzed in detail
- 5 risk flags with severity, section & reviewer questions
- Methods & statistical evidence assessment
- 5 questions a peer reviewer would ask
- Full six-dimension evaluation report & recommendations
One-time purchase · this paper only
One-time purchase, this paper only. AI-generated preliminary analysis — not peer review.
Frequently asked questions
Is PTSD genetic according to this study?
The study adds evidence that rare structural variants contribute: people with EHR-defined PTSD carried a small but statistically significant excess of rare deletions genome-wide in two ancestry groups. The per-variant effect is tiny, and the authors themselves call the findings preliminary.
What are rare copy number variants (rCNVs)?
rCNVs are stretches of DNA that are deleted or duplicated and are present in less than 1% of the population. Because a single CNV can remove or add whole genes, these variants often have larger biological effects than common single-letter DNA changes. Whole-genome sequencing detects much smaller CNVs than the older microarray chips.
Did this study find a specific PTSD gene?
No single gene reached definitive significance. The study found a genome-wide excess of rare deletions, a few small region-level signals (near genes such as KLF18 and RBBP8NL), and 11 gene-sets — related to developing brain expression, synaptic function, and genomic constraint — significant in a cross-ancestry meta-analysis.
Can these results be used for clinical genetic testing for PTSD?
No. This is a non-peer-reviewed preprint based on biobank electronic health records. The effect sizes are very small, the CNV calls lack independent validation, and no clinical utility is established. The authors explicitly state the findings should not guide clinical practice at this stage.
About this review
This is an AI-generated preliminary review by Evolve.Pub, produced from the openly available text of the paper. It is not peer review and does not represent the authors' views. The canonical record of the paper is at the publisher link above.
- The extracted text is heavily truncated mid-document: parts of the Discussion, several Supplementary Methods/Results sections, and the end of the phenotype/annotation methods are missing.
- Figures (Figure 1 overview, Manhattan and QQ plots, Supplementary Figures 1–20) are referenced but not analyzable from text.
- Tables are garbled by PDF extraction (e.g., Table 1 betas split across lines, '0.004 7' = 0.0047); numeric values were reconstructed carefully but minor transcription risk remains.
- Supplementary tables (3–9, 16–19) are cited but absent, so per-gene-set statistics and the exact FDR denominators could not be verified; one apparent citation-claim mismatch (1p31.1/KLF18 supported by a 1p34.2–p34.3 SLC2A1 reference) could not be resolved.
Go deeper on this paper
Wrote a paper? Stress-test it before reviewers do.
Upload your manuscript for the same structured scrutiny — reasoning, statistics, evidence, figures, and reproducibility.