DRC_cellbased_OutlierDetection

Version 1

This workflow can be used to fit dose-response curves from normalised cell-based assay data (%confluence) using the KNIME HCS extension. The workflow expects triplicates for each of eight test concentrations. This workflow needs R-Server to run in the back-end. Start R and run the following command: library(Rserve); Rserve(args = "--vanilla"). Three types of outliers can be removed: 1 - Outliers from triplicate measurement (standard deviation cut-off can be selected), 2 - inactive and weekly active compounds (% confluence cut-offs can be selected), 3 - toxic concentrations (cut-off for reduction in confluence with stepwise increasing concentration can be selected) Output are two dose-response curve fits per compound for pre and post outlier removal with graphical representation and numerical fit parameters.