Automated iterative STOCSY on X by peak intensity variable defined by the chemical shift driver_peak.
Iterates multiple rounds of STOCSY initially from a given driver peak of interest.
In subsequent rounds from all peaks correlating above a certain threshold to driver/s in the previous round.
Highly correlating (putatively structural) peaks are grouped together and the results automatically plotted in the istocsy_iplot interactive plot (showing node-to-node associations alongside the corresponding spectral data)
model = csm_istocsy( spectra, driver_peak, 'peak_inds', peak_inds, 'istocsy_cutoff', istocsy_cutoff, 'structural_cutoff', structural_cutoff, 'n_rounds', n_rounds, 'save_name', save_name, 'plot_method', plot_method )
| Variable | Type | Default Value | Description |
|---|---|---|---|
| *spectra | csm_spectra | None | csm_spectra object containing spectral matrix. |
| *driver_peak | 1*1 | None | Chemical shift value of target intensity variable. |
| peak_inds | 1*n | [] | Peak picked indices, if not included peak list generated through detection at zero crossings of a smoothed spectral derivative calculated using a Savitzky-Golay third order polynomial filter of the mean spectrum |
| istocsy_cutoff | 1*1 | None | Correlation threshold above which( |r| > istocsy_cutoff |
| structural_cutoff | 1*1 | 0.95 | For each set of newly identified peaks at each round, groups those which correlate with absolute value > structural_cutoff |
| n_rounds | 1*1 | 10 | Number of iterations |
| save_name | str | None | Name for saving results and plots. |
| plot_method | str | 'all' | Colour scale for plot. 'all' colours from -1 to 1, else colours on scale from min to max corr |
| Variable | Type | Description |
|---|---|---|
| csm_istocsy | csm_wrapper | Object with some stored inputs, the outputs and auditInfo. |
| csm_istocsy.output.results | m*2 | Results of which peaks where detected in each round. Columns correspond to round | index or representative peak. |
| csm_istocsy.output.sets | m*n | 'structural' or highly related sets, each row contains the indices of highly related peaks (represented by one node in the interactive plot). |
| csm_istocsy.output.connections | m*n | 'structural' or highly related sets, each row contains the indices of highly related peaks (represented by one node in the interactive plot). |
| csm_istocsy.output.correlates | m*n | Node connectivities; for each row, the first number corresponds to the driver node, and subsequent values to those connected nodes (where all numbers relate to row indices in the results matrix). |
| csm_istocsy.output.peaks_plot | m*n | Plotting peaks information, each row corresponds to a row in connections, and all peaks from the same structural set have the same index. |
| csm_istocsy.output.all_peaks_plot | 1*n | All identified peaks. |
| csm_istocsy.output.round_peaks_plot | m*n | Peaks identified in each round -( rows = n_rounds ). |
| csm_istocsy.output.plot_xy | m*2 | x and y locations for each node in the interactive plot. |
Sands CJ, Coen M, Ebbels TM, Holmes E, Lindon JC, Nicholson JK.
Data-driven approach for metabolite relationship recovery in biological 1H NMR data sets using iterative statistical total correlation spectroscopy.
Analytical Chemistry 2011 Mar 15;83(6):2075-82
Copyright Imperial College London 2019