Figure 4

Noise and peak overlap tolerance of the NTF2 model for peak factorization of a synthetic HSQC spectrum. A) HSQC spectrum constructed using four synthetic signals. B) Effects of noise and peak overlap on peak picking. The amount of noise in the spectrum is shown on the x-axis. The y-axis shows the peak separation in number of points. Red circles indicate that peaks were incorrectly picked. Blue circles show that the peaks were correctly picked upon factorization. C) Effects of the amount of noise on the peak position determination. Differences in the peak position in number of points are shown on the y-axis. The x-axis shows the amount of noise in the spectrum. Blue points indicate that the peak was correctly picked. Red points indicate that the peak was obtained with incorrect parameters because of higher noise in the spectrum. D) Effects of the amount of noise on the peak intensity: The x-axis shows the amount of noise present in the spectrum. Differences in the peak intensity of the peak are shown on the y-axis. Blue points show that the peak was distinguishable from the noise. Red points show that the peak was obtained with incorrect parameters upon factorization.