Estimate Background

Estimate background toggles optional estimation and subtraction of local background signal in the vicinity of the candidate spectrum. For each candidate spectrum, a linear regression is performed to determine the scaling factor by which the theoretical spectrum needs to be multiplied to approximate the matched spectrum. Linear regression is performed either with a forced zero intercept or a variable intercept. In case variable intercept is used, the intercept determined from linear regression is approximately equal to the local background noise adding a systematic intensity bias to experimental peaks. Then the intercept is subtracted from matched experimental peaks before scoring matched spectra and showing plots. “Estimate background” allows for a variable intercept if it is on and forces a zero intercept otherwise. Background subtraction can make matches more definitive in some cases, but can also lead to missed matched when background noise is too strong.