Potential mechanisms for predicting comorbidity between multiple myeloma and femoral head necrosis based on multiple bioinformatics

doi:10.21203/rs.3.rs-3792368/v1

Download PDF

Research Article

Potential mechanisms for predicting comorbidity between multiple myeloma and femoral head necrosis based on multiple bioinformatics

https://doi.org/10.21203/rs.3.rs-3792368/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective: To explore Multiple myeloma (MM) through multiple bioinformatics analysis

The potential mechanism of comorbidity with Osteonecrosis of the femoral head (ONFH).

Method:According to the inclusion criteria, download the MM and ONFH high-throughput chip datasets from the GEO database and preprocess them separately. Using weighted gene co expression network analysis (WGCNA) to construct co expression networks on MM self datasets, further screening modules and genes related to MM clinical characteristics, enriching and analyzing potential co disease genes of NAFLD and IS, and preliminarily screening key genes for MM and ONFH co disease through Cytoscape. Validate the ability of gene expression and performance evaluation through two disease related datasets. And evaluate the relationship between the difference and consistency of the two in the immune microenvironment.

Results: Through screening 418 co pathogenic genes with immunity, we found that the biological process of the two kinds of diseases in the ribosome synthesis process was consistent, especially the protein synthesis. Further screening key genes through PPI, and analyzing the dataset of the validation queue, the average area (ROC) of the 5 genes under the operating characteristics of the subjects was between 0.8 and above. Comparing the analysis of the two groups of data in the immune microenvironment, it is found that the infiltration of plasma cell is consistent with the comorbidity of diseases, and it can be used as the relevant immune target for subsequent targeting.

Conclusion: MM and ONFH share common pathogenic genes, which in turn mediate differential changes in related signaling channels and immune cells, affecting the high incidence of OA and MDS and the phenomenon of these two diseases. In terms of biological mechanism, MM and ONFH co disease may be mainly associated with RPS19, RPL35, RPL24, RPL36, EIF3G as key genes, as well as plasma cell as key immune infiltration, which can serve as the central mechanism for the development of the two diseases. This study provides ideas and references for further research.

Multiple myeloma

Osteonecrosis of the femoral head

Bioinformatics

Mechanisms

Multiple myeloma (MM) is a plasma cell malignancy characterized by the presence of lytic bone disease resulting in severe bone pain, pathological fractures, spinal cord compression and hypercalcemia, with up to 90% of myeloma patients developing osteolytic lesions during the course of the disease^{[1, 2]}. Osteonecrosis of the femoral head (ONFH) is a progressive debilitating lesion associated with many diseases, including trauma, steroid use, alcohol abuse and hypercholesterolemia, including some common clinical manifestations such as pain or limitation of motion^{[3, 4]}. ONFH is one of the devastating complications of systemic steroid use, and long-term steroid use leads to a hyperlipidemic state in most patients and exposes them to risk of osteoporosis and osteonecrosis^[5]. One study noted an overall incidence of ONFH of 6.7% in MM patients receiving dexamethasone hormone therapy^[6]. Although MM and ONFH are two different systemic diseases, the human body is a complex and unified whole, and clinical evidence suggests that there is an important link between MM and ONFH^[7].

Early diagnosis and treatment of multiple myeloma bone disease is clinically important to improve patient prognosis^[8]. Although imaging provides visual evidence of MM, it does not allow for early and quantitative accurate evaluation. In recent years, a large number of clinicopathological sections have confirmed that ONFH in the repair response area of femoral head necrosis is an inflammatory response and that malignant lesions of bone marrow may affect femoral head necrosis, therefore, it is important to study the mechanism of multiple myeloma in the occurrence of femoral head necrosis. However, there are few reports on the relational nature of the two. It is important to clarify the potential link between the two for intervention and treatment of MM and ONFH. Bioinformatics is a potential genomic analysis technique based on high-throughput sequencing technology, which can search for potential patterns between diseases from the perspective of genetic molecules. Therefore, the aim of this study is to integrate clinical sequencing results of ONFH and MM patients, combined with transcriptomic data of patients, to analyze and reveal the relationship between different microenvironmental cell populations and patients' prognosis and clinical characteristics, it was conducted to screen and identify common disease genes in MM and ONFH based on bioinformatics to provide theoretical support for intervention and treatment of MM and ONFH co-morbidities.

Data download and pre-processing

Downloaded the ONFH related datasets GSE123568, GSE74089 and MM related datasets GSE6477, GSE13591 from the GEO database through the “GEOquery” package^[9]. The GSE123568 dataset, from Homo sapiens, contains 40 samples, including 10 normal samples and 30 ONFH samples. GSE74089 dataset, from Homo sapiens, contains 8 samples, including 4 normal samples and 4 ONFH patients. GSE6477 dataset is from Homo sapiens, contains 140 samples, including 15 normal samples and 125 MM patients. GSE13591 dataset is from Homo sapiens, contains 138 samples, including 5 normal samples and 133 MM patients. We normalize the data and standardize the data through the “lima” package. Datas were normalized, filtered, filled with missing values, and merged with probes for special cases such as probes that were not expressed in any of the samples, missing values, or genes that corresponded to multiple probes.

Differential gene and function enrichment

Differentially expressed gene (DEG) was obtained by using the “limma”package of R, setting P < 0.05 and the absolute value of Fold change ≥ 0.5 as the screening conditions. “ggplot2” and “pheatmap” are R languages that can visualize the genes in normal and tumor groups. The obtained DEG were analyzed by GO^[10] and KEGG^[11] enrichment using “cluster-Profiler”, “org.Hs.eg.db” and enrichplot packages, with P < 0.05 as the screening criterion; and visualized using “ggplot2” package.

Weighted gene co-expression (WGCNA) network construction and correlation module screening

In this study, the following analyses will be conducted on the respective data sets of MM: genes with more than 25% inter-sample variation in the respective comprehensive data sets will be imported into the WGCNA data package^[12], abnormal samples will be excluded, soft thresholds and adjacency matrices will be calculated by using the co-expression similarity of the pick-Soft threshold function, the adjacency matrix will be transformed into a topological overlap matrix, and genes with similar expression profiles will be clustered into the gene module by hierarchical clustering and dynamic tree cutting functions to detect modules and cluster genes with similar expression profiles into gene modules, calculate module membership and Gene Significance for the modules most relevant to MM clinical features, and retain the genes in the modules for subsequent analysis.

Protein-protein interaction (PPI)

In this study, we used STRING database^[13] to construct protein-protein interaction network related to DEGs from mRNA genes screened by differential analysis, and used Cytoscape^[14] to visualize PPI network model. Closely connected local regions in PPI network. By using the algorithm of MCC in the plug-in of cytoHubba, we selected the top 10 DEGs as the hub genes .

Validation of diagnostic markers

The value of the obtained genes as diagnostic markers was verified in the cohort of the training and validation sets by drawing receiver operating characteristic (ROC) curves for the above-mentioned intersection sets^[15], and the diagnostic value was determined by using P < 0.05 as the threshold. The area under the ROC curve (AUC) was calculated, and the value of AUC ranged from 0 to 1. At the same time, we validated the final gene expression in the MM and ONFH related datasets

Analysis of immune cell infiltration

Using CIBERSORT^[16] analysis tool to analyze the composition of immune cells in microenvironment from gene expression data. After uploading gene expression matrix data to CIBERSORT website, select LM22 marker gene matrix and set the permutation times to 1000. Wilcoxon test was used for the difference of cell infiltration level, and P-value < 0.05 was considered as difference.

Analysis of differentially expressed genes in ONFH

A total of 384 DEGs were obtained from the data set GSE123568, of which 119 were highly expressed in ONFH group and 265 were low expressed in the disease group (Figure 1A-B).

Differences in MM data and WGCNA analysis

According to the GSE6477 database, 2450 DEGs that meet the conditions were obtained through differential analysis, including 1251 upregulated genes and 1199 downregulated genes. The DEGs were displayed using volcano and heat maps (see Figure 2A-B). Construct a co expression network of 12548 genes using R package "WGCNA" and cluster CSCC samples. The optimal soft threshold was determined based on the scale-free fitting index R2 when it first reached 0.9 β= 4 (Figure 2C-D), at this point, the average connection level of the network is relatively high and can contain sufficient information. A hierarchical clustering tree was constructed using dynamic hybrid cutting technology. Each leaf on the tree represents a gene, and genes with similar expression data stand up to form a branch of the tree, representing a gene module, and a total of 16 modules were generated (Figure 2D-F). The brown templatehas the highest correlation (Figure 2 F).

Screening of co-morbid genes and enrichment analysis

A total of 418 potential co-morbid genes between MM and ONFH were obtained by intersecting the genes with the highest correlation between the differential genes of ONFH and the clinical features of MM (Figure 3A).GO analysis revealed that the biological processes were mostly focused on ribonucleoprotein complex biogenesis, cytoplasmic translation, ribonucleoprotein complex assembly, and further molecular functions such as DNA−binding transcription activator activity, RNA polymerase II−specific, structural constituent of ribosome(Figure 3B). KEGG enrichment analysis revealed that after excluding non-specific pathways, it was mainly associated with PI3K-Akt, MAPK, Proteoglycans in cancer, Proteoglycans in cancer and Signaling pathways regulating pluripotency of stem cells and other signaling pathways (Figure 3C).

PPI network construction and preliminary screening of key co-disease genes

The PPI network of NAFLD and IS co-disease genes was visualized by Cytoscape 3.8.0 software, and the results showed that there were 47 nodes and 105 edges in the PPI network of NAFLD and IS co-disease genes, as shown in Figure 4A. Figure 3A shows. The key co-disease genes were screened by centrality analysis of the PPI network using the cytoHubba plug-in, and 10 potential key co-disease genes were initially screened in this study (Figure 4B). Further we used the MCODE algorithm to model the interaction network based on the results of PPI network analysis to find densely linked protein clusters in the network and annotate the biological functions of each cluster, and a total of 13 subclusters were analyzed, the first two of which are shown below (Figure 4C).

Analysis of Hub Genes Related to MM and ONFH

Due to the screening of the top 10 Hub gene through PPI, each Hub gene showed significant differential expression in both MM and ONFH. No differences were found in RPL14, RPL18, RPL15, and RPL35A in MM, while all other genes showed upregulation in MM (Figure 5). The diagnostic ROC curve of the Hub gene showed that except for RPL12, 9 genes had good diagnostic value in MM, as shown in Figure 6. In the ONFH validation group, it was found that there was no difference between the RPL14 and RPL12 groups. Except for the EIF3G change, which was down regulated, all other genes were up regulated (Figure 7). The diagnostic ROC curve of the Hub gene shows that except for RPL18, RPL14, RPL15, and RPL10A, six genes have good diagnostic value in ONFH, as shown in Figure 8.

Differential analysis of immune infiltration in MM and ONFH

As shown in Figure 8, the difference in immune infiltration cells between MM and ONFH groups was analyzed using violin plots. Plasma cells in the ONFH group were significantly reduced, while Monocytes were significantly increased; In MM, B cells naive and Plasma cells significantly decreased, while in contrast, B cells memory, T cells CD8, T cells follicular helper, T cells regulation (Tregs), Dendritic cells activated, and Mast cells activated significantly increased. The review found that the two groups of immune cells with the same changes were Plasma cells, and the trend of change was significantly reduced.

Compared with previous bioinformatics studies on disease co-morbidity, this study has the following advantages: first, we guessed that MM and ONFH are correlated based on clinical experience, and we used R language multiple data packages to normalize the data by strict background calibration, normalization, elimination of batch effects and removal of outliers to ensure the reliability of the results as much as possible; the main advantage of WGCNA is that it can identify highly correlated genes and cluster them into gene modules for subsequent analysis. The main advantage of WGCNA is that it can identify highly correlated genes and cluster them into gene modules, and the resulting modules can show the correlation coefficients with clinical features for subsequent analysis. In addition, this study combined multiple biomolecules to analyze the co-morbidity mechanism from multiple bioinformatics perspectives in order to explore the potential mechanisms of MM and ONFH co-morbidity more comprehensively and deeply.

Through gene bank search, it was found that both MM and ONFH, as two complex malignant diseases in clinical practice, have a large gene network and a large number of overlapping predicted gene targets, indicating that they have a close relationship contact. Through degree of freedom screening, it was found that RPL14, RPL24, RPL36, RPL18, RPL12, RPL15, EIF3G, RPL10A, RPS19, and RPL35A have high degrees of freedom and are located at the core of the target interaction network, possibly related to MM and ONFH genes. Furthermore, by analyzing the expression of core genes and comparing their ROC performance, RPL24, RPL36, EIF3G, RPS19, and RPL35 were identified as candidate comorbidity genes.

RPS19, RPL35, RPL24, and RPL36 are genes of the gene family essential for pre-rRNA processing and maturation of the 40S ribosomal subunit. It has been reported that increased levels of ribosomal proteins are closely associated with the development of human cancers. Of course, under stress stress conditions, the opposite phenomenon occurs, leading to reduced ribosomal activity and decreased protein synthesis. Thus, the ribosome biogenesis is a key factor in the control of cell growth and proliferation, and any deregulation in this process Any deregulation in this process may lead to abnormal cell proliferation or pathological states such as cancer and metabolic disorders. At the same time, deletion in breast cancer cells will reduce the growth and vitality of cells, inhibit the release of elF 6 of 60S subunit, thus inhibiting the expression of hat dependent protein required for the survival and proliferation of breast cancer cells, and also inhibit the assembly of 80S ribosome and polymer. Research has shown that the upregulation of RPL24 transcription may be related to the occurrence of human colorectal tumors^[17]. And it was found that RPS19 is overexpressed in bone marrow mesenchymal stem cells, and mutations in this gene cause Diamond-Blackfan anemia^[18]. There are reports that RPS19 is significantly overexpressed in colorectal cancer tissues and cells, and a decrease in RPS19 expression can inhibit colorectal cancer cell growth and reduce tumorigenicity in nude mice^[19]. Other studies have found that targeted up regulation of RPL15 expression in pancreatic cancer can inhibit the growth of pancreatic cancer cells ^[20]. EIF3G is one of the subunits of translation initiation factor 3 in eukaryotic cell and is necessary for protein translation initiation^[21].However, these genes have not been reported in the study of MM and ONFH co-morbidities, and there are bioinformatics results that suggest that RPS19, RPL35, and RPL24 are core genes involved in the translational silencing and protein modification processes in the pathogenesis of multiple myeloma^[22].

GO and KEGG analyses were able to clarify the network-like relationships among the targets, and GO analysis revealed that the biological processes were mostly focused on ribonucleoprotein complex biogenesis, cytoplasmic translation, ribonucleoprotein complex assembly, and further molecular functions such as DNA − binding transcription activator activity, RNA polymerase II − specific, structural constituent of ribosome. KEGG enrichment analysis revealed that after excluding non-specific pathways, it was mainly associated with PI3K-Akt, MAPK, Proteoglycans in cancer, Proteoglycans in cancer and Signaling pathways regulating pluripotency of stem cells and other signaling pathways. It has been demonstrated that high levels of phosphorylated GSK-3β inhibit osteogenic differentiation of stem cells and even accelerate osteoblast apoptosis in the establishment of animal models^[23].Zhang et al. ^[24]found that atorvata action resulted in upregulation of miRNA-186, which assisted TLR4 in co-mediating the MAPK/NF-κB pathway to inhibit the process of ischemic necrosis of the femoral head, leading to apoptosis promoting autophagy. In addition, the PI3K/Akt signaling pathway plays a crucial role in regulating MM. Research has shown that in MM cells, formosanin C inhibits cell apoptosis and autophagy through the PI3K/AKT/mTOR signaling pathway, suggesting that changes in the malignant biological characteristics of MM cells depend to some extent on the activation of the PI3K/AKT/mTOR signaling pathway^[25]. ROS/p38 MAPK can participate in the regulation of tumor growth and apoptosis processes, playing a key role in anticancer therapy. The accumulation of ROS in large quantities can lead to energy deficiency in MM cells, induce apoptosis, promote p38 MAPK protein phosphorylation, induce cell autophagy, and significantly promote MM cell apoptosis, which has a significant anticancer effect on MM^[26].

Monoclonal gammopathy of undetermined significance (MGUS) as one of the early stages of MM, It has been observed that the expression levels of disaccharide chain proteoglycan, plasmin activator inhibitor 1 and periosteum element in bone marrow ECM of MGUS and MM patients increase, and the expression of fibronectin 2, protein lysine 6-oxidase and chitinase 3-like protein 1 decreases. These differentially expressed ECM proteins can promote the adhesion, invasion, migration and proliferation of MM malignant plasma cell ^[27]. The extracellular matrix is composed of collagen, fibrinogen, proteoglycans and other proteins involved in cell adhesion, migration and differentiation. The interaction between bone marrow ECM and cells plays an important role in the homing, proliferation, and differentiation of bone marrow cells. However, there is currently no report on the role of proteoglycans in the process of femoral head necrosis. This result shows that there is a certain correlation between the synthesis and metabolic process of proteoglycans and the occurrence of femoral head necrosis, which needs further experimental verification.

Furthermore, through immune infiltration analysis, it was found that many immune system components, such as B cells memory, T cells CD8, T cells follicular helper, T cells regulation (Tregs), Dendritic cells activated, Mast cells, B cells naive, and Plasma cells, are involved in the immune supervision and killing effects of malignant transformed osteomyeloma cells, and can cause functional disorders under the influence of myeloma cells, Even promoting and maintaining the growth and survival of myeloma, inducing the generation of drug resistance, and leading to recurrent disease^{[28, 29]}. After the occurrence of osteonecrosis, the necrotic tissue and cells in the femoral head release various cytokines, which tend to activate inflammatory reactions with a large number of inflammatory cells, thereby disrupting the original immune balance, leading to greater bone destruction than osteogenesis, ultimately leading to irreversible collapse of the femoral head. More and more studies have shown that MM and ONFH are related immune inflammatory diseases^{[30, 31]}.

MM and ONFH both have complex pathological processes. In this study, the cross sections of two groups of large gene networks were excavated through bioinformatics methods, and the functional analysis of target genes found that they interact in inflammatory pathways and glycoprotein metabolism. The exploration of the biological functions and synergistic relationships of numerous core targets has provided possible directions for studying the mechanism of action and prevention and treatment between MM and ONFH to a certain extent, but the specific relationship still requires further research on clinical tissue information and basic experiments.

This article is an analysis of bioinformatics and does not involve ethics.

Funding

No funding .

Availability of data and materials

This study obtained raw data from the corresponding author.

Author Contribution

Jie Li, Jing Dong and Ming Li, Hongbo Zhu wrote the main manuscript text and Peicheng Xin prepared figures. All authors reviewed the manuscript.

BUMMA N, DHAKAL B, FRASER R, et al. Impact of bortezomib-based versus lenalidomide maintenance therapy on outcomes of patients with high-risk multiple myeloma [J]. Cancer, 2023.
CAVO M. CAR T-cell therapy for triple-class exposed relapsed/refractory multiple myeloma [J]. Haematologica, 2023.
EDMOND T, HAMEED D, OLSEN D, et al. Surgical Technique: Core Decompression of Femoral Head Osteonecrosis with Bone Marrow Aspirate Concentrate [J]. Surg Technol Int, 2023, 42.
FAN Y, LIU X, ZHONG Y, et al. Evaluation of the predictive values of collapse and necrotic lesion boundary for osteonecrosis of the femoral head prognosis [J]. Front Endocrinol (Lausanne), 2023, 14: 1137786.
OZGUL H A, AKIN I B, MUTLU U, et al. Diagnostic value of machine learning-based computed tomography texture analysis for differentiating multiple myeloma from osteolytic metastatic bone lesions in the peripheral skeleton [J]. Skeletal Radiol, 2023.
WU X, GENG C, SUN W, et al. Incidence and Risk Factors of Osteonecrosis of Femoral Head in Multiple Myeloma Patients Undergoing Dexamethasone-Based Regimens [J]. Biomed Res Int, 2020, 2020: 7126982.
ALLEGRA A, MANIA M, D'ASCOLA A, et al. Altered Long Noncoding RNA Expression Profile in Multiple Myeloma Patients with Bisphosphonate-Induced Osteonecrosis of the Jaw [J]. Biomed Res Int, 2020, 2020: 9879876.
JAGANNATH S, JOSEPH N, CRIVERA C, et al. Component Costs of CAR-T Therapy in Addition to Treatment Acquisition Costs in Patients with Multiple Myeloma [J]. Oncol Ther, 2023.
DAVIS S, MELTZER P S. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor [J]. Bioinformatics, 2007, 23(14): 1846–7.
GENE ONTOLOGY C, BLAKE J A, DOLAN M, et al. Gene Ontology annotations and resources [J]. Nucleic Acids Res, 2013, 41(Database issue): D530-5.
KANEHISA M, FURUMICHI M, TANABE M, et al. KEGG: new perspectives on genomes, pathways, diseases and drugs [J]. Nucleic Acids Res, 2017, 45(D1): D353-D61.
ZHANG J, LU K, XIANG Y, et al. Weighted frequent gene co-expression network mining to identify genes involved in genome stability [J]. PLoS Comput Biol, 2012, 8(8): e1002656.
SZKLARCZYK D, KIRSCH R, KOUTROULI M, et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest [J]. Nucleic Acids Res, 2023, 51(D1): D638-D46.
HOLMAS S, PUIG R R, ACENCIO M L, et al. The Cytoscape BioGateway App: explorative network building from the BioGateway triple store [J]. Bioinformatics, 2019, 36(6): 1966–7.
WU L, JIN Q, CHEN J, et al. Diagnostic Accuracy of Chest Computed Tomography Scans for Suspected Patients With COVID-19: Receiver Operating Characteristic Curve Analysis [J]. JMIR Public Health Surveill, 2020, 6(4): e19424.
GUAN M, JIAO Y, ZHOU L. Immune Infiltration Analysis with the CIBERSORT Method in Lung Cancer [J]. Dis Markers, 2022, 2022: 3186427.
KNIGHT J R, VLAHOV N, GAY D M, et al. Rpl24(Bst) mutation suppresses colorectal cancer by promoting eEF2 phosphorylation via eEF2K [J]. Elife, 2021, 10.
HIREGANGE D G, RIVALTA A, YONATH A, et al. Mutations in RPS19 may affect ribosome function and biogenesis in Diamond Blackfan anemia [J]. FEBS Open Bio, 2022, 12(7): 1419–34.
ZHOU X, LIU Z, HE T, et al. DDX10 promotes the proliferation and metastasis of colorectal cancer cells via splicing RPL35 [J]. Cancer Cell Int, 2022, 22(1): 58.
PROVOST E, BAILEY J M, ALDRUGH S, et al. The tumor suppressor rpl36 restrains KRAS(G12V)-induced pancreatic cancer [J]. Zebrafish, 2014, 11(6): 551–9.
ZHANG X, ZHANG J, ZHAO W, et al. Long non-coding RNA LINC02446 suppresses the proliferation and metastasis of bladder cancer cells by binding with EIF3G and regulating the mTOR signalling pathway [J]. Cancer Gene Ther, 2021, 28(12): 1376–89.
YAN H, ZHENG G, QU J, et al. Identification of key candidate genes and pathways in multiple myeloma by integrated bioinformatics analysis [J]. J Cell Physiol, 2019, 234(12): 23785–97.
SUN W, LI M, ZHANG Y, et al. Total flavonoids of rhizoma drynariae ameliorates bone formation and mineralization in BMP-Smad signaling pathway induced large tibial defect rats [J]. Biomed Pharmacother, 2021, 138: 111480.
ZHANG Y, MA L, LU E, et al. Atorvastatin Upregulates microRNA-186 and Inhibits the TLR4-Mediated MAPKs/NF-kappaB Pathway to Relieve Steroid-Induced Avascular Necrosis of the Femoral Head [J]. Front Pharmacol, 2021, 12: 583975.
CHEN P, WU S, DONG X, et al. Formosanin C induces autophagy-mediated apoptosis in multiple myeloma cells through the PI3K/AKT/mTOR signaling pathway [J]. Hematology, 2022, 27(1): 977–86.
WU X, LIU Y, ZHANG E, et al. Dihydroartemisinin Modulates Apoptosis and Autophagy in Multiple Myeloma through the P38/MAPK and Wnt/beta-Catenin Signaling Pathways [J]. Oxid Med Cell Longev, 2020, 2020: 6096391.
REN Z, SPAARGAREN M, PALS S T. Syndecan-1 and stromal heparan sulfate proteoglycans: key moderators of plasma cell biology and myeloma pathogenesis [J]. Blood, 2021, 137(13): 1713–8.
BRUGNARA C. Immune Therapies in the Treatment of Multiple Myeloma [J]. Am J Hematol, 2023.
GIANNOTTA C, CASTELLA B, TRIPOLI E, et al. Immune dysfunctions affecting bone marrow Vgamma9Vdelta2 T cells in multiple myeloma: Role of immune checkpoints and disease status [J]. Front Immunol, 2022, 13: 1073227.
MA J, GE J, GAO F, et al. The Role of Immune Regulatory Cells in Nontraumatic Osteonecrosis of the Femoral Head: A Retrospective Clinical Study [J]. Biomed Res Int, 2019, 2019: 1302015.
TIAN L, WEN Q, DANG X, et al. Immune response associated with Toll-like receptor 4 signaling pathway leads to steroid-induced femoral head osteonecrosis [J]. BMC Musculoskelet Disord, 2014, 15: 18.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Potential mechanisms for predicting comorbidity between multiple myeloma and femoral head necrosis based on multiple bioinformatics

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Data download and pre-processing

Differential gene and function enrichment

Weighted gene co-expression (WGCNA) network construction and correlation module screening

Protein-protein interaction (PPI)

Validation of diagnostic markers

Analysis of immune cell infiltration

Results

Discussion

Declarations

References

Additional Declarations

Status:

Version 1