In the GAPRO core, the parameter w (Table  1 ), which in the original version of GAPRO (Hiller et al.

All local maxima of the spectrum are identified, and those with an intensity larger or equal to ηR_min (Table  1 ) are considered to be peaks.

  1 c, 2 ), and therefore we recommend to use 1 digital point as the input value for Δ ν_min and r_min (Table  1 ).

For example, all peaks with an average sensitivity below a user-defined threshold value of R_ave can be deleted, or all peaks with more than n violations of the threshold R_single (Table  1 ) can be deleted, where n and R_single are user-defined threshold parameters.

Signals were identified using the new GAPRO peak picker with R_min  = 4.0, and the following set of GAPRO parameters: S_min  = 6, Δ ν_min  = 5.0 Hz, r_min  = 15.0 Hz and k  = 200 (Table  1 ).

Among the input variables of the geometric algorithm GAPRO, three parameters have a dominant effect on the result of the spectral analysis (Table  1 ).

The potential impact is illustrated by its use, with n  = 0 and R_single  = 0 (Table  1 ), to validate the results obtained with S_min values of 3 and 4 (Fig.  9 ).

The other GAPRO parameters (Table  1 ) were R_min  = 5.0, Δ ν_min  = 5.0 Hz, r_min  = 15.0 Hz and k  = 400.

^c The GAPRO parameter S_min is defined in Table  1 .

The GAPRO analysis of the projection spectra with the commonly used parameters (Tables  1 , 4 ) often reports only one of these two peaks, since all projections with projection angles that do not resolve the HA dimension do not contribute to the support, S , for the second glycine peak.