Association between Arsenic Level, Gene Expression in Asian Population, and In Vitro Carcinogenic Bladder Tumor

Figure 7

Bladder cancer prediction model. (a, b) Plot of the ROC curve as an outcome of logistic regression prediction model in multivariate fashion, by AUC ROCs. The 95% confidence intervals (CI) are shown. (a) Prediction outcome of primary tumor on GSE13507 dataset. (b) Reevaluation of prediction ability of 3 genes previously identified on GSE3167 dataset. (c) Survival outcome of three genes using the Human Protein Atlas data. Best expression cut off: based on the FPKM value of each gene, patients were classified into two groups and association between prognosis (survival), and gene expression (FPKM) was examined. The best expression cut-off refers the FPKM value that yields maximal difference with regard to survival between the two groups at the lowest log-rank value. Best expression cut-off was selected based on survival analysis. Median expression refers to the median FPKM value calculated based on the gene expression (FPKM) data from all patients in this dataset. When clicking on this number, the vertical dashed line indicating cut-off, the interactive survival plot, and the Kaplan-Meier curve will be adjusted to show results based on the median expression. Median follow-up time refers to the median time (years) after diagnosis with this type of cancer, based on clinical data from all patients in this dataset. score: Log-rank value for Kaplan-Meier plot showing results from analysis of correlation between mRNA expression level and patient survival. 5-year survival for patients with higher expression than the expression cutoff. 5-year survival for patients with lower expression than the expression cutoff.

(a)

(b)

(c)