Figure 2

Partial least squares regression results with three principal components. A) A scatter plot illustrating the measured versus predicted values for each different treatment group across all three cell lines (LNCaP, PC3, MDA-PCa-2b). B) The regression coefficients for all 8 phosphoproteins across the 3 time points. C) The mean absolute value of the regression coefficients, indicating the contribution of that phosphoprotein to the overall model. D) The R-squared value as additional phosphoproteins are added to a 3 principal component partial least squares regression model. The R-squared value of Erk is for a model built on Erk data alone. The R-squared value for RPS6 is for a model built on Erk and RPS6 data. The R-squared value for p38 is for the complete model for all data and is 0.577.