Machine learning identifies lactate metabolism biomarkers and deciphers immune infiltration landscapes in Parkinson's disease

机器学习可识别乳酸代谢生物标志物并解析帕金森病中的免疫浸润图谱

阅读:2

Abstract

Parkinson's disease (PD), the second most common neurodegenerative disorder worldwide, is characterized by irreversible neuronal loss. Its progressive motor and non-motor symptoms-including resting tremor, postural instability, and autonomic dysfunction-substantially impair quality of life and impose significant socioeconomic burdens. Current diagnosis relies primarily on subjective motor assessments due to the absence of objective biomarkers, often delaying detection until advanced stages of neurodegeneration. Therefore, identifying PD-specific biomarkers is critical for early diagnosis, targeted interventions, and disease management. Recent evidence suggests a strong association between PD and dysregulated lactate metabolism, involving altered lactate levels, lactylation-dependent epigenetic modifications, and abnormal expression of genes related to lactate metabolism, all contributing to disease pathogenesis. To identify novel biomarkers associated with lactate metabolism for early PD detection, we integrated two GEO datasets as training cohorts. Differential expression analysis identified 119 differentially expressed genes (DEGs), of which 33 overlapped with 5,193 lactate metabolism-related genes. Using machine learning algorithms, including the least absolute shrinkage and selection operator (LASSO) and support vector machine (SVM), we prioritized four candidate biomarkers: KCNJ6, PDK4, LRP2, and DENR. External validation confirmed their diagnostic utility. A nomogram model constructed with these markers demonstrated robust predictive performance for PD. Immune infiltration analysis further revealed distinct immune cell profiles in the substantia nigra of PD patients and their significant correlations with the identified biomarkers. Collectively, these genes associated with lactate metabolism represent promising diagnostic indicators, offering valuable insights for both PD research and clinical application.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。