Construction of diagnostic markers for hub lncRNAs in Parkinson’s disease based on chip re-annotation

Abstract BACKGROUND Parkinson’s disease (PD) is a progressive neurodegenerative disease that is also the most common motor disorder and is accompanied by the loss of DA neurons in the brain. Long non-coding RNAs (lncRNAs) have recently been identified as new genetic entities that regulate cellular processes. One of the main functions of lncRNAs is the regulation of the expression of specific genes in multiple steps, including the regulation of transcriptional and post-transcriptional mechanisms and epigenetics. MATERIAL AND METHODS Here we downloaded three sets of expression-spectrum data for PD from the GEO database. The data were re-annotated with R package, which were integrated into a set of expression profiles for the analysis of differentially expressed lncRNAs. Subsequently, lncRNA/mRNA co-expression modules were identified through a weighted co-expression analysis and lncRNAs were expressed based on binding differences. The diagnostic tags of PD were filtered with key modules and were finally used to build the PD diagnostic prediction model. RESULTS Based on lncRNA re-annotation, a total of 1931 lncRNA expression values in the three sets of data were obtained and significant differences in expression (P  0.6) as PD diagnostic markers and as features to construct a SVM classification model. The model had good predictive ability in the training set and verification set (AUC of 0.9928 and 0.464, respectively), which illustrated their potential as diagnostic markers of PD. CONCLUSIONS This study provided new molecular entities for the diagnosis of PD, which may promote the early detection of this disease and the development of personalized therapies.