ridge tracking

Ridge tracking is the method used in analyzing time series NMR data. Quantifying changing peaks through time is like tracking ridges

It is implemented by both clustering and curvature techniques.

The main function on the new curvature based approach is here. For more information, please refer to readme and tutorial. Here we provide a short guideline and for more information please go to the tutorial or contact the author(yue.wu@uga.edu).

short tutorial

1. pre-step: Clone the Toolbox locally and add to the path
2. pre-step: Download the example data: on Metabolomics Workbench (PR000738). The /ST001103/analysis/multi_sample_data/sampleData.mat is file to load.
3. Use this script as an example. Move to working folder and load the data
4. Run lines 1-21. Check the spectral quality.
5. Run lines 22-45. Click “Delete Ridge(s)”. Select all ridges to delete and hit return/enter (in order to have a clean region for re-tracking).
6. Draw a box among the re-tracking region and click once in the figure.
7. Click “Pick Final Clusters” on the MENU. Select ridges till they turn thicker. Hit return/enter.
8. Click 'OK' and hit space to save the information for selected ridges. Click 'Yes'. Click ridges to be refined and hit return/enter to continue. Draw a box for removing ridge regions, click once, and hit space.

Contact Yue Wu (yue.wu@uga.edu) for questions

Ref:

Wu, Y., Judge, M. T., Arnold, J., Bhandarkar, S. M. & Edison, A. S. RTExtract: Time-series NMR spectra quantification based on 3D surface ridge tracking. Bioinformatics, doi:10.1093/bioinformatics/btaa631 (2020).