Peak width extraction

Hi there,

I am trying to assign an HNcoCA spectrum. Is there any way to extract peak width by pick peaking? Maybe I can have the H, C and N widths for each peak in the peak table or something like that.
Thank you for your time and help.


Hi Mahdi,

if you do a normal area peak pick (Ctrl + Shift + left-drag (or Cmd + Shift + left-drag on a Mac)), then the program will automatically try and fit the peak (similarly if you picked your peaks with the Pick & Assign module). In the peakList table you should then see an LW (line width) parameter for each dimension. This is the width at the half height in Hz.

You can check how good the fit is in the SpectrumDisplay by opening the Settings and selecting to show the Peaks as “Filled Line Widths” rather than a “Cross”. Also remember that you can switch the axes, e.g with shortcut YZ, so you can open a new SpectrumDisplay in which the C rather than the N is in the z-dimension. That way you can also check the fit in the third dimension, too.

You can refit fit selected peaks with the RP or you can fit several partially overlapped peaks together with RG (these commands are also in the right hand mouse menu for selected peaks).



A bit more info

The fit commands are ‘refit peak(s) singular’ and ‘refit peak(s) group’ they can be accessed on selected peaks by the right click menu

or when not clicking on the peaks but with selected peaks by right clicking on the spectrum widget and using the ‘selected peaks’ menu item


Awosome, Thank you Vicky and Gary.
Later I just need to convert Hz to ppm.


If you like, then this little macro will do the calculation for you and put the answer in the peak comment box. Simply make sure you’ve selected the correct PeakList at the top:

pl = get('PL:hsqc.1')

for pk in pl.peaks:
    comment =[]
    for lw, freq in zip (pk.lineWidths, pl.spectrum.spectrometerFrequencies):
        lwppm = (lw / freq)
    pk.comment =  str(comment)

See our Running Macros documentation on how to run a macro.

Best wishes,

Yes, it worked, many tnx.