Machine‑learning pipeline links air‑pollutant targets to hepatocellular carcinoma risk and prognosis

A multi-omics study identified 43 genes linking 10 air pollutants with hepatocellular carcinoma. A 7-gene prognostic signature and two diagnostic markers were derived via 101 machine‑learning combinations; the hub gene PSMB5 was highly expressed in regulatory T cells and positiv…

• Design: Integrated network toxicology, transcriptome‑wide gene‑expression, and 101‑algorithm machine‑learning pipeline — entirely in silico; no external or prospective clinical validation.
• Key secondary finding: Molecular docking suggested potential binding affinity between the identified protein targets and the air pollutants.
• Primary limitation: Purely computational/retrospective study — the gene signatures have not been validated in independent patient cohorts, so clinical utility remains unproven.

RadPigeon summaries are original and for information only. They are not clinical advice.