Public data sent 5 plus, the province of the flower.

Today, I would like to share with you an article published in the journal Molecular Oncology (if: 5.962) in May this year, which is a study on the prognosis and subtype analysis of HPV infected cervical cancer based on methylation data.based on the reclassification and risk stratification of cervical cancer caused by HPV related methylation, HPV ‐ related methylation ‐ based recanalization and risk stratification of cervical cancer (CC) is the main cause of cervical cancer (CC).in this study, HPV related methylation markers were used to identify subtypes of cervical cancer and analyze the prognosis.firstly, DNA methylation profiles were obtained from TCGA to identify different subtypes of HPV by unsupervised clustering.patients with cervical cancer were divided into methylation-h, methylation-m and methylation-l clusters.compared with methylated-m and methylated-l, methylated-h had better overall survival (OS).and GSEA analysis showed that methylation-m and methylation-l groups were enriched in cancer-related hallmarks such as KRAS signal transduction. The mutation, amplification and deletion of Myc, notch, PI3K ‐ Akt were detected in methylation-h.in addition, tumor microenvironment analysis showed that the degree of immune cell infiltration of methylated-h was relatively low.finally, six HPV related methylation sites were used as signature for prognostic analysis, which can predict the prognosis of cervical cancer patients.data collection and processing of dataset 2.1 in TCGA and geo databases, 450k data of 309 cervical cancer samples were downloaded from UCSC Xena.for each CpG locus, the beta value ranges from 0 to 1, and the clinical data of patients are downloaded to extract the HPV infection status.Download TCGA to obtain rnaseq, somatic mutation and CNV data of 306, 289 and 297 cervical cancer patients.somatic mutation data were analyzed using maftools.using gistic 2.0 to detect somatic CNV and identify gene amplification or deletion.2.2 screening HPV related methylation sites, using camp R package to identify differential methylation probes between cervical cancer and normal samples, HPV positive and HPV negative cervical cancer.deltabeta & gt; 0.2 and p.adj & lt; 0.05 were considered as hypermethylation, deltabeta & lt; - 0.2 and p.adj & lt; 0.05 were considered hypomethylation.2.3 unsupervised hierarchical clustering analysis used the β value of HPV related methylation sites with prognostic value to identify subtypes of cervical cancer patients. 2.4 identification and screening of HPV related methylation markers 294 patients with survival data were randomly divided into two data sets (training set and test set). first, univariate Cox regression analysis was performed to identify HPV methylation sites with prognostic value. then, Lasso regression analysis was performed using R-package "glmnet" to further screen methylation sites. furthermore, stepwise multiple Cox regression analysis was used to screen methylation sites. use the following formula: risk score = β value 1 × coefficient 1 + β value of methylation site 2 × coefficient 2 + The β value of methylation site n × coefficient n. according to the median risk score, cervical cancer patients were divided into high-risk group and low-risk group. survival R package was used for survival analysis and timeroc R package was used to calculate AUC. 2.5 gene set enrichment analysis (GSEA) used the "fgsea" R package for GSEA enrichment analysis, and 10000 times of disturbance were conducted for each analysis parameter, and the screening p.adj & lt; 0.05 considered it to be significant. 2.6 single sample gene set enrichment analysis (ssgsea) was used to analyze ssgsea using gsva R package to evaluate the immune infiltration level of 24 different immune cell types marker genes. 2.7 tumor microenvironment analysis uses the estimate method to estimate the scores of stromal cells and immune cells in tumor tissue to evaluate the infiltration of tumor microenvironment cells. results 3.1 HPV related methylation sites were screened between tumor and normal tissues, and 35678 differentially methylated probes were screened using camp (Fig. 1a). 48190 differential methylation probes were screened between HPV positive and HPV negative cervical cancer (Fig. 1b). further intersection, 9249 common HPV related methylation sites were obtained. Figure 1. Probe differential methylation analysis 3.2 the survival results of different cervical cancer methylation clusters were significantly different. Univariate Cox regression analysis was performed using HPV related methylation sites. 191 HPV related methylation sites (P & lt; 0.05) were identified for unsupervised cluster analysis, and 294 cervical cancer patients were divided into three groups (Figure 2a). and there were significant differences in survival among different groups (P = 0.009, figure 2b). PCA analysis was used to compare transcriptome data among the three clusters, and sample segregation was found in the three clusters (Fig. 2C). Figure 2. Three cervical cancer classes identified by hierarchical clustering 3.3 are biological processes related to cervical cancer clusters. Gene set enrichment analysis is performed among different clusters. it was found that there was a positive correlation between KRAS signal transduction and methylation L (Fig. 3a). in addition, IFN - γ reaction was positively correlated with methylation m (Fig. 3b). Figure 3. GSEA results 3.4 mutation and CNV analysis of patients in the three clusters further studied the genomic changes in the three clusters. figure 4A – C shows the top 30 most common mutated genes in the three clusters. and the mutation frequencies of 10 common carcinogenic pathways in the three clusters were calculated (Fig. 4D). The mutation frequency of myc was higher in methylated H group (Fig. 4D). CNV analysis showed that 8q24.21 (myc) was amplified in methylated h cluster (Fig. 5a, b). in addition, the amplification of 11q22.1 (Yap1) in methylation m cluster was associated with cancer (Fig. 5C, d). finally, 11q22.1 amplification (Yap1) and 13q14.2 deletion (RB1) were found in methylation-l (Fig. 5E, f). Figure 4. Mutation comparison among three methylation clusters in cervical cancer Fig. 5 amplification and deletion of methylation h, m and l methylation genes; construction of prognostic model for 3.5 HPV related methylation markers; univariate Cox regression analysis identified 20 methylation sites with prognostic significance. after lasso regression analysis, 11 were identified. stepwise multiple Cox regression was performed for these 11 methylation sites. six methylation sites (cg23170347, cg163776000, cg13759702, cg01727408, cg050080700 and cg07227049) were identified to construct the prognosis model (Table 1). The DNA methylation levels of cg13759702 were associated with high risk, while those of the other five probes were associated with low risk. Table 1. Prognostic related loci information 3.6 the ability to predict HPV associated methylation signals was evaluated. Kaplan Meier survival analysis was performed in the training and test data sets, and cervical cancer patients were divided into high-risk and low-risk groups. there was a significant difference in survival between the two groups (Figure 6a). the AUC of the six DNA methylation markers was 0.899 at 5 years and 0.888 at 3 years (Fig. 6b). similar results were observed in both the test data set and the entire data set (Fig. 6C, 6D). together with the test set, the whole data set also has good prediction performance (Fig. 6e-6f). Fig. 6 the prognostic effect of HPV related methylation markers 3.7 the independence of HPV related methylation signals in OS prediction may have a certain impact on the survival of some clinicopathological variables, including age, pathological stage, clinical stage, etc. to assess the independence of this HPV related methylation signal, patients with cervical cancer were reclassified according to different clinicopathological characteristics (Table 2). the results showed that signature was not associated with age, clinical stage, histological grade, T stage, lymph node metastasis and tumor status. Table 2. Association of clinical features 3.8 comparison of immune status of patients in different clusters. In order to compare the difference of the proportion of 24 immune cells among patients with cervical cancer, the relative proportion of 24 immune cells was estimated using ssgsea. thermography shows tumor infiltration in 294 TCGA samples (Fig. 7). Figure 7. Immune status of patients in different clusters. Compared with the other two clusters, methylation-h cluster showed lower estimate score and higher tumor purity. these results suggest that methylated h has a different immunophenotype than the other two clusters. in conclusion, this work uses HPV related methylation markers to identify subtypes of cervical cancer, analyzes the prognosis of different subtypes, gene mutation differences and differences in immune infiltration level, and studies the classification of cervical cancer from the perspective of methylation, which is not wrong and innovative. if you have analysis needs, you can also add a little sister wechat communication