Abstract
BACKGROUND: Monoclonal immunoglobulins are commonly found in multiple myeloma (MM), a prevalent hematologic malignancy that is currently incurable. In recent years, the association between systemic lupus erythematosus (SLE), an autoimmune disease, and MM has garnered increasing attention. However, there remains a lack of in-depth research regarding the interactions between these two conditions and their potential pathogenic mechanisms. Therefore, in order to improve the identification of MM associated with SLE, this work attempts to clarify the pathogenic pathways that are shared by MM and SLE and to develop corresponding diagnostic models. METHODS: This study employs a comprehensive bioinformatics analysis combined with machine learning techniques to extract relevant data from public databases. We used GO and KEGG pathway analyses to investigate the functionalities and pathway enrichments of the DEGs that we found in MM and SLE populations. Furthermore, we used the STRING database to build a PPI network for the intersecting genes and the cytoHubba plugin in Cytoscape software to identify important genes with biological significance. To establish a diagnostic model for SLE-related MM, we compared 113 combinations of 12 machine learning algorithms, ultimately determining the optimal model. RESULTS: Our analysis identified 63 intersecting genes, with 31 exhibiting upregulated expression and 32 showing downregulated expression. The selection of key genes indicated that nine genes met the criteria of having both Degree and MCC values exceeding 3, among which seven (CDH1, IL4, AURKB, HGF, H2BC9, AREG, TJP1) have previously been confirmed to have direct associations with MM. Notably, H2BC5 was identified as a specific gene associated with SLE-related MM. Our findings revealed that elevated expression of H2BC5 is significantly correlated with an increased risk of MM, suggesting its potential critical role in the diagnosis and identification of this malignancy. CONCLUSION: A new molecular framework for the early diagnosis of MM, especially in SLE patients, is established by this study. Our findings highlight H2BC5 as a possible biomarker that merits more research into how it contributes to the development of MM. By identifying the shared pathogenic mechanisms between SLE and MM, our research offers new perspectives for future clinical interventions and personalized therapies.