Identification and Validation of Diagnostic Biomarkers for Osteoarthritis Coexisting with IVDD and LFH via Bioinformatics and Machine Learning

INTRODUCTION: Osteoarthritis (OA), intervertebral disc degeneration (IVDD), and ligamentum flavum hypertrophy (LFH) frequently manifest concurrently in the aging population. This study aims to identify potential diagnostic genes associated with these three interconnected conditions. METHODS: Utilizing datasets from the Gene Expression Omnibus (GEO) database, we applied Limma and weighted gene co-expression network analysis (WGCNA) to discern pivotal genes. Subsequent enrichment analyses shed light on the functional implications of these identified genes. The utilization of three distinct machine learning algorithms facilitated the identification of hub genes. Evaluation of the predictive capacity of these hub genes was conducted through nomograms and receiver operating characteristic (ROC) curves. Additionally, predictions pertaining to transcription factors, microRNAs, and potential therapeutic drugs were made. Furthermore, the study delved into the exploration of immune cell infiltration in OA. RESULTS: An integrated bioinformatics analysis of OA datasets identified 246 key genes enriched in inflammatory pathways, including MAPK signaling and interleukin signaling. Comparison across OA, IVDD, and LFH datasets identified 9 common differentially expressed genes. Machine learning algorithms subsequently identified ANKH and GADD45B as central hub genes. A diagnostic nomogram constructed using these genes demonstrated strong predictive performance, particularly for OA. Further analyses predicted SREBF1 as a potential co-regulator and quercetin as a drug candidate targeting both hub genes. Immune infiltration analysis revealed altered levels of resting memory CD4 T cells and activated mast cells in OA, correlating with hub gene expression. Finally, RT-qPCR validation in clinical samples confirmed the differential expression patterns of ANKH and GADD45B, supporting their relevance. DISCUSSION: The identification of ANKH and GADD45B as common hub genes offers novel insights into the shared molecular mechanisms underlying co-occurring OA, IVDD, and LFH. The strong predictive performance of the nomogram using these genes underscores their potential as diagnostic biomarkers for OA. The predicted interaction of quercetin with both hub genes, alongside SREBF1 as a co-regulator, suggests new therapeutic targets for these genes. Moreover, the findings on altered activated mast cell levels in OA correlating with hub gene expression highlight their potential role in OA pathogenesis and as therapeutic targets. These findings, supported by initial RT-qPCR validation, provide a foundation for further experimental studies to confirm their clinical utility. CONCLUSION: This study identifies ANKH and GADD45B as promising diagnostic genes for distinguishing OA co-occurring with IVDD and LFH. Furthermore, our findings underscore the significance of MCs in the context of OA, providing insights into their potential role in the pathogenesis of the condition.