Identification of diagnostic signature and immune infiltration for ischemic cardiomyopathy based on cuproptosis-related genes through bioinformatics analysis and experimental validation

Research context and source limits

Limited evidence note. Only the source material identified below has been reviewed; this is not a complete Paper GEO V2 assessment.

Full-article-supported candidate-gene and mouse-expression evidence. Unresolved human cohort accounting and model-validation questions prevent a complete V2 upgrade.

Five candidate genes and experimental follow-up

The article selects DLST, ATP7B, FDX1, SLC31A1 and DLAT after a retrospective public-transcriptomic analysis. The full text is now available for review, but its classification performance is not treated here as independently validated clinical diagnosis.

Source: Publisher-layout PDF pp. 2–3, Methods 2.1–2.5; pp. 5–8, Results 3.3 and Figures 4–5.

Mouse expression evidence

Methods 2.9 describes male C57BL/6 mice assigned to sham control or left-anterior-descending coronary artery ligation, six per group. Figure 8 reports echocardiography with n=6 and Western blot with n=4; these are separate reported assay counts, not additional animal cohorts. The myocardial Western blot and immunohistochemistry show higher DLST and ATP7B and lower FDX1, SLC31A1 and DLAT in the ligation group. This establishes reported expression differences in the tested model, not that these genes cause cuproptosis or diagnose disease in patients.

Source: Publisher-layout PDF pp. 3–4, Methods 2.9–2.14; p. 8, Results 3.6; p. 11, Figure 8C–E and caption (echocardiography n=6; Western blot n=4).

Computational chemical hypotheses

Chemical–gene associations were retrieved from CTD and examined with molecular docking. These analyses generate hypotheses; no compound-treatment experiment or human treatment benefit is established by this article.

Source: Publisher-layout PDF p. 3, Methods 2.8; pp. 8 and 10, Results 3.5 and Figure 7; pp. 3–4, animal experimental methods.

Why diagnostic validation remains unresolved

Figure 1 labels the training inputs as 181 IC samples from GSE5406 and 91 from GSE57338, but labels their merged IC group as 203. Those counts do not reconcile. Moreover, GSE57338 is a SubSeries of the stated validation accession GSE57345; the article does not provide a verified sample-level exclusion map. Figure 5 describes five-fold cross-validation within its evaluation datasets. These records do not establish an untouched external test of a frozen diagnostic model.

Source: Official GEO Series GSE57338, relation “SubSeries of: GSE57345”; compare publisher PDF p. 3 Figure 1, p. 2 Methods 2.1 and p. 8 Figure 5 caption.

Evidence limits

• Human sample accounting remains unresolved: Figure 1 input IC counts 181 and 91 do not equal the displayed combined 203. No corrected denominator or deduplicated participant total is inferred. The GSE42955 AUC is 0.943 in the abstract but 0.933 in Results 3.3 and Figure 5D. Neither number is selected as a reconciled performance estimate.
• Training GSE57338 is part of validation SuperSeries GSE57345. A GSM-level membership and overlap audit and the frozen-model evaluation procedure are needed before claiming independent external validation. The article alternates between disease identification and cluster prediction. Methods 2.4, Results 3.3 and the Figure 5 caption do not resolve the exact target and partitioning of every model analysis.
• Animal expression differences do not demonstrate causal cuproptosis activity, clinical diagnostic utility or treatment efficacy. The post-ligation assessment interval and the relationship between the six allocated animals and four Western-blot observations require clarification.
• Supplementary Figures S1–S7 and Tables S1–S3 were not obtained in this review. No supplementary-only values or docking-based therapeutic recommendation is adopted.
• The author-account PDF matches the formal article title, DOI, 18-author order, journal, volume and article number. It is retained privately for audit; no open redistribution licence is asserted.

Provenance

• Formal article and publisher-layout PDF — Thirteen-page publisher-layout PDF obtained through the authorized ResearchGate author account. International Immunopharmacology 138 (2024), 112574; Methods 2.1–2.14, Results 3.3–3.6 and Figures 1, 4, 5, 7, 8.
• Author-account source record — ResearchGate Private Full-text; Ge Zhang upload dated 16 July 2024. Private full-text 1; file 1-s2.0-S1567576924010956-main.pdf, accessed 27 October 2026. File access requires authorized account permissions.
• Indexed publication identity — PubMed record. PMID 38971104; ordered authors, DOI, journal, volume and electronic publication date 6 July 2024.
• Official dataset relationship — GEO Series record. GSE57338 is a SubSeries of GSE57345; previously saved official SOFT Series response verified during this review.

Source review: 2026-10-27.

Verify publication identity

These records identify the same original work; database inclusion is not evidence of scientific validity or clinical benefit.