Workflow 1: A KNIME workflow illustrating data input, data cleaning, manipulation of pKa values to identify the most acidic pKa for compounds with multiple acidic groups, decision tree analysis, and the development of a preliminary machine learning model. The input file for this workflow is titled "Example input."

Workflow 2: A KNIME workflow showing data input, feature elimination through a linear correlation routine to remove autocorrelated features, feature list generation, and the application of a genetic algorithm within a feature selection loop to count the selected features. The input file for this workflow is titled "Example input."

Workflow 3: A KNIME workflow integrating threshold variations and using the output from workflow 2 as input for a machine learning model. The input file for this workflow is titled "Features after GA and LC."

For more details, please refer to https://doi.org/10.1016/j.aichem.2024.100063