3+ meristem letter + clinical liver cancer SCI, hurry up and learn!

Click [Medical prescription] Follow our clinical partners to post articles with various difficulties, how to break it? Today, I will interpret the latest article published in Cancer Medicine (IF=3.
491) in April 2021 for clinical children's shoes.

Using public databases combined with clinical practice, a model based on the expression of Hippo-related genes in adjacent tissues of liver cancer was constructed to predict the prognosis of patients with liver cancer.

Come look, your SCI article is on your way! Title: Normal tissue adjacent to tumor expression profile analysis developed and validated a prognostic model based on Hippo-related genes in hepatocellular carcinoma HCC) is the most common malignant disease in the world.

Although the diagnosis and treatment of liver cancer have been greatly improved in recent years, there is still a lack of accurate methods for predicting the prognosis of patients.

There is evidence that Hippo signaling in tissues adjacent to liver cancer plays an important role in the occurrence of liver cancer.

In this study, the author aims to construct a model based on the expression of Hippo-related genes (HRGs) in tissues adjacent to liver cancer to predict the prognosis of patients with liver cancer.

Methods: The gene expression data and clinical information of normal tissue adjacent to liver cancer (PNTAH) came from GEO and TCGA databases; Hippo-related genes were determined from 4 typical Hippo-related signal pathways; single-factor Cox regression analysis was used to screen survival-related Hippo-related Gene; LASSO and multivariate Cox regression analysis were used to construct a prognostic model; ROC analysis was used to confirm the true and false positive rate of the model.

Results: A prognostic model was constructed based on the expression levels of five Hippo-related genes (NF2, MYC, BIRC3, CSNK1E, and MINK1) in normal tissues adjacent to liver cancer; the mortality of liver cancer patients increased with the increase in the risk score determined by the model; , Found that the risk score is an independent risk factor for patient survival; ROC analysis shows that the prognostic model has better predictive value than other conventional clinical indicators; the TCGA-LIHC cohort also confirmed the reliability of the prognostic model; generating a nomogram to predict the survival rate of patients , The prediction and exploration of this model in liver cancer tissues shows that it has the specificity of normal tissues adjacent to liver cancer.

Conclusion: The authors established and validated a prognostic model based on the expression levels of five Hippo-related genes in normal tissues adjacent to liver cancer.
This model should help predict the prognosis of liver cancer patients.

Research ideas and results 1.
Data acquisition and sorting Download the GSE14520 data set from GEO database, including 232 liver cancer samples, 232 normal tissue samples adjacent to liver cancer and corresponding clinical data; GSE102079 data set, including 14 normal liver samples and GSE112790 The expression profile of 15 normal liver samples in the dataset.

From the TCGAGTEx cohort downloaded from UCSC Xena, 110 normal liver specimens of GTEx were selected; 50 normal tissue specimens adjacent to liver cancer; 371 liver cancer specimens and corresponding clinical data of TCGA liver cancer (TCGA-LIHC) for further analysis.

A prognosis model was constructed using GSE14520's 232 adjacent liver cancer samples corresponding to normal tissue samples, and 50 liver cancer adjacent samples of TCGA liver cancer corresponded to normal tissue samples for verification.

2.
Differential gene expression patterns of normal liver samples, normal tissue samples adjacent to liver cancer and liver cancer samples.
A total of 29 normal liver samples with mRNA expression profiles, 232 normal tissue samples adjacent to liver cancer and 232 liver cancer samples were obtained from the GEO database.

Table 1 shows the clinical characteristics of 232 patients with liver cancer, including age, gender, tumor stage, tumor size, recurrence status, survival time, and survival status.

After removing the batch effect, PCA was performed to analyze the expression patterns of different samples.
The first principal component (PCA1) and the second principal component (PCA2) were 25.
66% and 5.
97% of the data variation, respectively.

The mRNA expression profile is scattered in three different clusters, which reveals that there are surprising differences in the genes of normal liver samples, normal tissue samples adjacent to liver cancer, and liver cancer samples in the GEO cohort (Figure 2A), and in TCGA-GTEx The same discrepancy results also appeared in the cohort (Figure 2B).

3.
Identification and functional enrichment of Hippo-related genes To determine Hippo-related genes, as shown in Table 2, the author downloaded 4 typical Hippo-related pathways from MsigDB, namely GO_Hippo_SIGNALING, KEGG_Hippo_SIGNALING_pathway, RETOME_SIGNALING_BY_Hippo and WP_HIPPOYAP_SIGNALING_pathway.

After removing the duplication and deletion probes in the expression profile, a total of 76 genes were identified as Hippo-related genes.

The heat maps of these 76 Hippo-related genes indicate that Hippo signaling may be regulated differently in normal liver samples, normal tissue samples adjacent to liver cancer, and liver cancer samples (Figure 3A).

Table 3 shows the GO function enrichment analysis of 76 Hippo-related genes.

At the same time, these Hippo-related genes are related to the Hippo signaling pathway, the regulation of the Hippo signaling pathway, the regulation of the canonical Wnt signaling pathway, the stress-activated MAPK cascade, and the stress-activated protein kinase signaling cascade (Figure 3B); KEGG pathway analysis The above genes are mainly enriched in the Hippo signaling pathway, tight junction and MAPK signaling pathway (Figure 3C).

4.
Use the expression profile of normal tissue samples adjacent to liver cancer in the GEO cohort to construct a prognostic model based on Hippo-related genes.
Univariate Cox regression analysis was used to determine the expression of prognostic Hippo-related genes in normal tissue samples adjacent to liver cancer.
The standard was p<0.
05.

14 genes were found to be significantly related to the patient's OS in the corresponding normal tissue samples next to liver cancer (Figure 4A).

At the same time, the hazard ratio (HR) and 95% confidence interval (CI) are estimated and displayed.
Among them, NF2 is significantly related to survival rate, and the HR is the highest, suggesting that the expression of NF2 in normal tissue samples adjacent to liver cancer is of great significance in liver cancer patients.

Since the HR of NF2 is much higher than that of other genes, the forest plot for univariate regression analysis is not included here.

In addition, these 14 genes were found to be differentially expressed in normal liver samples, normal tissue samples adjacent to liver cancer, and liver cancer samples (Figure 4B).

Based on the prognostic-related Hippo-related genes in normal tissue samples adjacent to liver cancer, LASSO and multivariate Cox regression analysis were performed to construct a model based on Hippo-related genes (Figure 4C, D).

Finally, five genes, NF2, MYC, CSNK1E, BIRC3, and MINK1, were determined to construct a prognostic model (Table 4).

5.
Validation of the model constructed by normal tissue samples adjacent to liver cancer.
To evaluate the performance of the prognostic model in predicting the clinical outcome of patients, the risk score of each liver cancer patient is calculated according to the expression profile of normal tissue samples adjacent to liver cancer, and the patients are divided into groups based on the median risk score In the high-risk group or low-risk group, 22 liver cancer patients with missing clinical data were excluded, with a median risk score of 1.
0278; the survival curve showed that the survival rate of the high-risk group was lower than that of the low-risk group (Figure 5A); the ROC curve had an AUC of 0.
750, which proved The prognostic model performed well for survival prediction (Figure 5B); as shown in Figure 5C–E, it shows the distribution of risk scores in normal tissue samples adjacent to GSE14520 liver cancer, the survival status of each patient, and a heat map of 5 gene expression profiles.
The results showed that as the risk score increased, survival time shortened and mortality increased.

6.
Use the TCGA-LIHC adjacent normal tissue sample expression profile to further verify the prognosis model.
The authors further analyzed the correlation between the risk score in TCGA-LIHC and the survival of liver cancer patients to verify the performance of the constructed model.

According to the median risk score calculated by the prognostic model constructed using the TCGA-LIHC adjacent normal tissue sample expression profile for liver cancer, these patients were divided into low-risk groups or high-risk groups.

The median risk score of the TCGA-LIHC cohort was 0.
0028.

There is a significant difference in the survival rate between the high and low groups (Figure 6A); in addition, the AUC calculated by the ROC curve is 0.
775, indicating that the model can predict the survival of liver cancer patients well (Figure 6B); Figure 6C–E also shows the different risk groups The distribution of risk scores, the number of patients examined, and a heat map of the prognostic Hippo-related genes.

7.
Independence evaluation and nomogram drawing of prognostic model The author conducted multivariate and univariate Cox analysis to explore whether the risk score or other routine clinical parameters of liver cancer patients in the GEO cohort are independent risk factors for OS.

Univariate Cox analysis showed that advanced tumor, larger tumor volume, and risk score were risk factors for OS (Figure 7A).

After multivariate analysis and correction of clinical parameters, only risk score and tumor stage are independent prognostic factors for patients (Figure 7B); ROC curve is drawn to compare the predictive value of risk score and other clinical indicators.

The results show that among the routine clinical parameters, tumor staging factors are the most predictive (Figure 7C).

But the predictive value of risk score (AUC = 0.
750) is higher than tumor stage.

The authors further analyzed the correlation between the risk score calculated by the liver cancer patient model and the clinical parameters (Figure 7D–H), and found that the high risk score was significantly associated with larger tumor size, advanced cancer and more recurrence.

In order to show the relationship between individual predictors and survival rates, the author developed a nomogram model based on the GEO cohort data and converted it into a scale within a certain range (Figure 8A).
Each parameter in the figure is age.
, Gender, tumor stage, tumor size, and risk score correspond to one point.

Add the points of all parameters to get the total point, which is used to determine the 1-year, 3-year, and 5-year overall survival rate.
The C index is 0.
755.
The calibration curve for 1-year, 3-year and 5-year survival prediction has good linearity ( Figure 8B), which means that the nomogram has a better calibration.

8.
Use liver cancer tissue expression profile to verify the specificity and prognostic model of 5 Hippo-related genes.
In order to study whether the expression of 5 Hippo-related genes in liver cancer tissues also has prognostic value, 232 liver cancer tissues were extracted from the GSE14520 data set Expression profile.

According to the median expression of each Hippo-related gene, liver cancer patients were divided into high expression group and low expression group.

Survival analysis showed that for each Hippo-related gene, there was no statistical difference between the high-expression group and the low-expression group; in addition, the expression value of Hippo-related genes in liver cancer tissues was used to calculate the risk score based on the formula derived from the normal tissue samples adjacent to the liver cancer.
According to the median risk score, liver cancer patients were divided into high-risk groups or low-risk groups.
The results also showed that there was no difference in survival rates between high-risk groups and low-risk groups; similar results were obtained in the analysis of liver cancer tissue expression profiles by TCGA-LIHC.

The above results indicate that the selected 5 Hippo-related genes and the constructed model have the specificity of normal tissues adjacent to liver cancer.

Conclusion In this study, the author constructed and verified a prognostic model based on the expression profile of normal tissue adjacent to liver cancer, which can predict the survival time of liver cancer patients.

The differentially expressed Hippo-related genes may provide a new perspective for the elucidation of the molecular mechanism of liver cancer.

The END call for papers "Medical Formula" is now officially open to fans! The content must be originally published and related to scientific research.
Once adopted, a generous reward (300-2000 yuan) will be given.
Please stamp for details.

"Medical side" has always been committed to serving "medical people", pushing the most cutting-edge and most valuable original clinical and scientific research articles to clinicians and scientific researchers.

The medical department has launched "Laboratory Basics", "SCI Writing Skills", "Document Intensive Reading and Analysis", "Easy Learning of Medical English", "National Natural Science Foundation of China", "Clinical Data Mining", "Gene Data Mining", "R Language Tutorial", "Medical Statistics", "Minimally Invasive Animal Experiment Training" and many other special courses, if you need to know the detailed tweets of the courses, you can follow the "Medical Party" public account and click "Exquisite Topics" to enter the Tencent classroom: https://medfun .
ke.
qq.
com NetEase Cloud Classroom: http://study.
163.
com/u/ykt1467466791112 Customer Service Tel: 15821255568 Customer Service WeChat: yixuefang1234