Integrated machine learning and bioinformatic analyses constructed a sulfur metabolism-related breast cancer risk model and identified heat-shock protein A9 as a potential therapeutic target for human breast cancer.

PURPOSE: Oncogenesis and tumor progression have been linked to abnormal metabolism. We aimed to investigate the potential connection between sulfur metabolism-related genes and clinical features of patients with breast cancer. METHODS: Machine learning algorithms were utilized to assess the risk index of sulfur metabolism-related genes in breast cancer. All patients were categorized into high- and low-risk clusters, based on their calculated average risk scores. Kaplan–Meier curves were used to evaluate the patient prognoses in different groups. Enrichment analysis was performed on the differentially expressed genes (DEGs) across these distinct clusters. The effect of the highest-risk gene, HSPA9, on the malignant behavior of tumor cells was appraised through siRNA transfection. RESULTS: A risk model with nine sulfur metabolism-related genes (ACOT2, ACOT4, CHPF, ELOVL2, HLCS, HSPA9, MICAL1, SPOCK2, and TCF7L2) was established, and low-risk groups exhibited better outcomes than high-risk groups. Various biological functions and pathways of the DEGs were observed between the different groups. The high-risk group exhibited a higher immune cell infiltration rate than the low-risk group. Inhibiting HSPA9 expression effectively reduced breast cancer cell proliferation and migration. CONCLUSION: Our genetic risk model provides a novel pattern for prognostic evaluations and individualized therapeutic strategies for breast cancer. Given its association with breast cancer risk, HSPA9 represents an exceptionally promising therapeutic target. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s12672-026-04427-0.