Machine learning and clinical validation identify programmed cell death-related diagnostic biomarkers for major depressive disorder

Major depressive disorder (MDD) contributes to a major global mental health burden and is characterized by complex molecular mechanisms. Given the emerging roles of programmed cell death (PCD) in MDD, this study aimed to profile PCD-associated transcriptomic features and identify clinically relevant candidate biomarkers. MDD-associated differentially expressed genes (DEGs) were identified from a human prefrontal cortex transcriptomic dataset. Hub genes associated with PCD dysregulation in MDD were screened using gene set variation analysis (GSVA), weighted gene co-expression network analysis (WGCNA), protein–protein interaction (PPI) network analysis, and further prioritized through multiple machine learning approaches. Clinical blood-based validation was performed to assess the diagnostic relevance of the prioritized hub genes. A binary logistic regression model was constructed to combine these hub genes, and the joint diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. A total of 743 MDD-related DEGs were identified in the human prefrontal cortex, including 453 upregulated and 290 downregulated genes. GSVA revealed extensive alterations in PCD pathway activity in MDD, characterized by increased enrichment of methuosis, autophagy, cuproptosis, and apoptosis, together with decreased enrichment of several other PCD pathways. Subsequent analysis prioritized five candidate genes with diagnostic potential including HADHB, NDUFB7, ACADVL, NDUFAF4, and NDUFS7. Their encoded proteins exhibited markedly reduced abundance in the clinical blood cohort. A binary logistic regression model incorporating these genes achieved favorable diagnosistic power in the clinical cohort (AUC = 0.876; 95% CI: 0.784–0.969). This integrative analysis highlights the potential involvement of PCD dysregulation in MDD and nominates HADHB, NDUFB7, ACADVL, NDUFAF4, and NDUFS7 as promising PCD–mitochondria–depression (PMD) diagnostic signature genes. This multigene diagnostic signature demonstrated robust diagnostic performance, supporting its potential translational value as a blood-based diagnostic aid for MDD.

Authors

Institutions

Publication Details

Journal
BMC Psychiatry
Published
2026-11-13
DOI
https://doi.org/10.1186/s12888-026-08793-5
Primary Topic
Treatment of Major Depression
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Machine learning and clinical validation identify programmed cell death-related diagnostic biomarkers for major depressive disorder

S.W. Ye, Yanyi Jiang, Qiuxiang Tan, Juncai Pu et al.
2 links
BMC Psychiatry
Treatment of Major Depression
article

Machine learning and clinical validation identify programmed cell death-related diagnostic biomarkers for major depressive disorder

S.W. Ye, Yanyi Jiang, Qiuxiang Tan, Juncai Pu, Siwen Gui, Yue Chen, Peng Xie, Zhuocan Li, Renjie Qiao, Wenxia Li, Xiaopeng Chen, Wei Wang, Honghui Xu, Haitao Zhao, Dongfang Wang, Hanping Zhang, Xiang Chen, Xiaogang Zhong, Yiyun Liu
article en

Abstract

Major depressive disorder (MDD) contributes to a major global mental health burden and is characterized by complex molecular mechanisms. Given the emerging roles of programmed cell death (PCD) in MDD, this study aimed to profile PCD-associated transcriptomic features and identify clinically relevant candidate biomarkers. MDD-associated differentially expressed genes (DEGs) were identified from a human prefrontal cortex transcriptomic dataset. Hub genes associated with PCD dysregulation in MDD were screened using gene set variation analysis (GSVA), weighted gene co-expression network analysis (WGCNA), protein–protein interaction (PPI) network analysis, and further prioritized through multiple machine learning approaches. Clinical blood-based validation was performed to assess the diagnostic relevance of the prioritized hub genes. A binary logistic regression model was constructed to combine these hub genes, and the joint diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. A total of 743 MDD-related DEGs were identified in the human prefrontal cortex, including 453 upregulated and 290 downregulated genes. GSVA revealed extensive alterations in PCD pathway activity in MDD, characterized by increased enrichment of methuosis, autophagy, cuproptosis, and apoptosis, together with decreased enrichment of several other PCD pathways. Subsequent analysis prioritized five candidate genes with diagnostic potential including HADHB, NDUFB7, ACADVL, NDUFAF4, and NDUFS7. Their encoded proteins exhibited markedly reduced abundance in the clinical blood cohort. A binary logistic regression model incorporating these genes achieved favorable diagnosistic power in the clinical cohort (AUC = 0.876; 95% CI: 0.784–0.969). This integrative analysis highlights the potential involvement of PCD dysregulation in MDD and nominates HADHB, NDUFB7, ACADVL, NDUFAF4, and NDUFS7 as promising PCD–mitochondria–depression (PMD) diagnostic signature genes. This multigene diagnostic signature demonstrated robust diagnostic performance, supporting its potential translational value as a blood-based diagnostic aid for MDD.

BMC Psychiatry
The Affiliated Yongchuan Hospital of Chongqing Medical University (CN), Chongqing Institute for Brain and Intelligence (CN), Jinfeng Laboratory (CN)
Good health and well-being
Treatment of Major Depression
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon