NMR Workshops

NMRFx Analyst: Processing, Analysis and Deep Learning from Small to Big (Molecules and Magnets)

NMRFx Analyst: Processing, Analysis and Deep Learning from Small to Big (Molecules and Magnets)

The topic of our round-table workshop on October 31, 2024, at 12:00 PM ET USA was on NMRFx Analyst: Processing, Analysis and Deep Learning from Small to Big (Molecules and Magnets) The Speaker was be Bruce Johnson Sr. Research Director – Computational Sciences Structural Biology Initiative CUNY Advanced Science Research Center.

NMRFx Analyst is a cross-platform NMR analysis program that has features for everything from NMR processing to advanced analysis for both small and large molecules. This workshop will be a live demonstration of a variety of the features present in NMRFx. I’ll first show basic processing on one-dimensional and multi-dimensional datasets, including using NUS algorithms. Then I’ll highlight some of the following features. Molecular structures can be read in for small molecules, RNA and proteins and chemical shift prediction performed. NMRFx has a facilitated assignment method that starts with predicted shifts and allows for a rapid interactive assignment for small molecules, peptides and RNA. Two-dimensional (flat) small molecules can be converted to energy minimized 3D structures and full torsion angle dynamics can be done on protein and RNAs. A recent advance is the incorporation of a Deep Learning Library. The widely used Tensorflow package is incorporated directly into NMRFx. This allows for training of models outside of NMRFx using Keras, and then incorporation of the trained models into NMRFx. We’ll show some early examples using this technology for predicting the secondary structure of RNA and then chemical shifts and spectra from the RNA sequence as well as predicting the secondary structure of proteins using assigned chemical shifts.

 

Leave a Reply

Your email address will not be published. Required fields are marked *