Exploring the diagnostic and immune infiltration roles of disulfidptosis related genes in pulmonary hypertension

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.

Reported molecular-screening workflow and animal-assay design; diagnostic performance is unresolved.

Research question and reported workflow

The study investigates disulfidptosis-related expression patterns in pulmonary hypertension. Its workflow compares random forest, support-vector machine, generalized linear and XGBoost models, then constructs a nomogram and regulatory-network hypotheses. DSTN, NDUFS1, RPN1, TLN1 and MYH10 are the five nominated genes; their nomination does not establish a clinically validated diagnostic panel.

Source: Methods: Construction of machine learning models; Results: Construction and evaluation of SVM, RF, GLM and XGB machine models

Separate experimental component

The paper also reports a hypoxia-exposure rodent experiment and Western-blot measurements of the nominated proteins. This is a preclinical expression assessment, separate from human diagnostic-model evaluation. It does not establish diagnostic accuracy in patients, demonstrate a treatment effect, or validate predicted drugs from molecular docking.

Source: Methods: Animal; Chemicals and materials; Results: Five DiGs as diagnostic markers were validated in HPH animal model; Figure 8

Why performance remains unverified

The abstract and Methods list different training and validation accessions. The Results also assign validation AUCs to cohorts differently from the Figure 5 caption. Neither accession series nor curve attribution is silently selected here; the cohort-to-analysis mapping requires clarification.

Source: Abstract Methods; Materials and methods: Data source; validation Results; Figure 5D–I caption

Evidence limits

• No disputed AUC is presented as verified.
• Protein-expression checks do not resolve the human cohort identity conflicts.

Provenance

• Article source — Official publisher or PMC full text reviewed locally. Methods: Construction of machine learning models; Results: Construction and evaluation of SVM, RF, GLM and XGB machine models

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.