Figure 2 | Scientific Reports

Figure 2

From: A machine learning approach to identify predictive molecular markers for cisplatin chemosensitivity following surgical resection in ovarian cancer

Figure 2

Correlation between gene expression and cisplatin IC50 in the identified molecular markers. (a) Waterfall plot representing the distribution of cisplatin sensitivity (resistant IC50 > 200 nM, intermediate IC50 > 50 nM, and sensitive IC50 < 50 nM) across ovarian cancer cell lines for CYTH3, ERI1, GALNT3, and S100A14. (b) Box plot representing IC50 values in cell lines grouped by expression quartiles (high = top quartile, low = bottom quartile, and medium = within interquartile range) for CYTH3. Dotted lines represent sensitive and resistant cut-offs of IC50 values for CYTH3, ERI1, GALNT3, and S100A14. (c) Modified ROC curve representing the ability of gene expression to correctly classify sensitive or resistant cell lines (CYTH3 AUC = 0·79, ERI1 AUC = 0·82, GALNT3 AUC = 0·90, and S100A14 AUC = 0·90), p-value comparing distribution of expression between sensitive and resistant cells (Mann–Whitney-Wilcoxon).

Back to article page