Fig. 5From: PMF-CPI: assessing drug selectivity with a pretrained multi-functional model for compound–protein interactionsFine-tuning results of PMF-CPI on datasets about drug selectivity. Different fine-tuning set scales are used with missing entries from 70% to 95%. MSE and Pearson’s correlation coefficient are shown for regression sets a AR, b JAK, and c CYP. d AUC and AUPR are used for CYP classification. All results are from five-fold cross-validation with the standard deviation as error barsBack to article page