Clustering toolbox

The Clustering Toolbox available here was designed as a user interface that allows for clustering analysis of untargeted metabolomics data. This user interface contains 23 functionalities spanning from hierarchical clustering to clustering comparision to KEGG pathways analysis. Most functionalities within this UI contain user options allowing for customization of analyses. Furthermore, we provide a start-up window that allows the user to define the number of threads (e.g., processes they would like to use during analysis allowing for rapid clustering analysis). We recommend the user use at least one less thread than available on the computer, with the optimal being half of the available threads.

Ensemble clustering combined with Clustering Optimization or ECCO is the main functionality within the clustering toolbox. This form of ensemble clustering is discussed in our pre-print pre-print and can be adapted for your preferences. ECCO is built for ensemble clustering solutions of agglomerative hierarcical clustering algorithms.

The Clustering Toolbox was built to mimic the pre-processing steps taken during untargeted metabolomics data analysis. We acknowledge that we do not provide all of the available pre-processing steps and recommend pre-processing prior to submission to UI and selecting 'None' and 'None' when prompted to select a transformation and scaling for your data.

1. If you do not have python installed, install at python.org and make sure to install pip to your command line.
2. Otherwise, proceed to next steps.

1. git clone https://github.com/hisl6802/ECCO

   • or Download zip file from the Code <> button above.

2. Recommended: Create a Conda environment for running the UI (see example below)

   • conda create ECCO_env
   • conda activate ECCO_env

3. After activating the environment use pip to install packages for UI

   • pip install -r requirements.txt

4. Start UI

   • python JuneLabClusteringGUI.py
   • or python3 JuneLabClusteringGUI.py

NOTE: You may need to install openpyxl if you get file won't read errors.

The ExampleFiles directory contains two example input files. Please see further documentation for other input files.

Please submit an Issue in the Issues tab, and I will address as quickly as possible.