Scholarlink

Background <p>Major Depressive Disorder (MDD) represents a grave mental affliction characterised by intricate pathological mechanisms and an elevated susceptibility to neurodegeneration. This study aims to integrate machine learning and bioinformatics analysis to identify potential biomarkers related to the pathogenesis of MDD and to elucidate the genes that have a causal association with this disease.</p> Methods <p>The training dataset was formed by merging two GEO datasets (GSE52790 and GSE98793), while the validation dataset included GSE44593 and GSE54564. Bioinformatics methods such as WGCNA, PPI, SMR, and 113 machine learning models (113 different algorithms) were employed in this study.</p> Results <p>In the course of this study, a total of 20 hub genes were identified. Through machine learning screening, the Lasso + RF model surfaced as the preeminent diagnostic model, registering an AUC value of 0.807 for the SERPING1 gene. Colocalisation analysis unveiled a pronounced upregulation in the expression levels of the pivotal gene, SERPING1, among patients afflicted with MDD. Moreover, Summary-data-based Mendelian Randomisation (SMR) analysis elucidated a substantial positive correlation between SERPING1 and an incresaed risk of MDD, thereby suggested its prospective function in impeding the pathological progression of MDD via the modulation of pertinent pathways.</p> Conclusions <p>By amalgamating bioinformatics analysis, machine learning, colocalisation analysis, and SMR analysis, this study designated SERPING1 as a potential cardinal gene and biomarker implicated in MDD pathogenesis. These discoveries accentuate the clinical applicability of SERPING1 as a diagnostic biomarker for MDD.</p> Clinical trial number <p>Not applicable.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrative analysis via bioinformatics and machine learning identifies SERPING1 as a biomarker candidate for major depressive disorder

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月7日

查看全文
  • Xiaokui Yuan,
  • Tong Wang

摘要

Background

Major Depressive Disorder (MDD) represents a grave mental affliction characterised by intricate pathological mechanisms and an elevated susceptibility to neurodegeneration. This study aims to integrate machine learning and bioinformatics analysis to identify potential biomarkers related to the pathogenesis of MDD and to elucidate the genes that have a causal association with this disease.

Methods

The training dataset was formed by merging two GEO datasets (GSE52790 and GSE98793), while the validation dataset included GSE44593 and GSE54564. Bioinformatics methods such as WGCNA, PPI, SMR, and 113 machine learning models (113 different algorithms) were employed in this study.

Results

In the course of this study, a total of 20 hub genes were identified. Through machine learning screening, the Lasso + RF model surfaced as the preeminent diagnostic model, registering an AUC value of 0.807 for the SERPING1 gene. Colocalisation analysis unveiled a pronounced upregulation in the expression levels of the pivotal gene, SERPING1, among patients afflicted with MDD. Moreover, Summary-data-based Mendelian Randomisation (SMR) analysis elucidated a substantial positive correlation between SERPING1 and an incresaed risk of MDD, thereby suggested its prospective function in impeding the pathological progression of MDD via the modulation of pertinent pathways.

Conclusions

By amalgamating bioinformatics analysis, machine learning, colocalisation analysis, and SMR analysis, this study designated SERPING1 as a potential cardinal gene and biomarker implicated in MDD pathogenesis. These discoveries accentuate the clinical applicability of SERPING1 as a diagnostic biomarker for MDD.

Clinical trial number

Not applicable.