Democratizing Cheminformatics: Interpretable Chemical Grouping Using an Automated KNIME Workflow

Thank you for your interest in this MoViz chemical grouping workflow! If you are interested in running the workflow locally, either to make edits or just see how it works, follow the directions below to install it on your computer.

Installation

Prerequisites

Before you can use the workflow, you will need to install its dependencies with Anaconda. If you don't have it already, download miniconda3. To install the dependencies:

  1. Download the environment.yml file from the repository.
  2. Open the Anaconda/Miniconda terminal. image
  3. Navigate to the folder containing environment.yml using cd /path/to/environment.yml.
  4. Run the following command: conda env create -f environment.yml. Press y to confirm the install if prompted. image
  5. Wait for the packages to install (this may take a while).

PaDEL-descriptor also requires Java JRE version 6 or above, which you can get here.

Running the workflow

To set up the workflow on your computer:

  1. Clone or download the repository from GitHub.
  2. In KNIME Analytics Platform, select your Local Space. image
  3. In your local space, select Import Workflow image
  4. Browse to the workflow file you downloaded; it will have the .knwf extension.
  5. KNIME may prompt you to install extensions; follow the on-screen instructions to do so. You may need to restart KNIME when done. image image
  6. Once KNIME restarts, open File > Preferences. In the left panel, navigate to KNIME > Conda. Click Browse or enter the path to your Anaconda/Miniconda installation (On Windows, this will often be in your AppData or User folder). image
  7. Exit Preferences and run the workflow with your data!

Note: If you used the environment.yml file to create an Anaconda environment, the environment should have the default name "movizenv". If you choose to rename the environment, you will need to update that in the workflow:

  1. Control-click on the first component in the workflow ("Upload").
  2. Double click on the node titled "Conda environment propagation"
  3. Select your environment from the dropdown.
  4. Click "OK" and run the workflow. github2