Evaluation of the asymmetric least squares baseline algorithm through the accuracy of statistical peak moments

This manuscript evaluates the performance of the Eilers’ smoothing algorithm by comparing the third central statistical peak moment of simulated peaks to their true value. Peaks were simulated by imposing an exponentially modified Gaussian peak profile onto a random segment of pre-recorded high performance liquid chromatography baseline. Prior to moment calculation the noise was removed from the simulated curves and the signal was subtracted by the approximated baseline. We found that with a signal to noise ratio of 1000 an 11% relative error was observed between the practical and true third statistical moment. The Eilers’ smoothing algorithm was observed to both reduced noise and baseline correct without distorting the shape of the chromatographic peak.

► The third peak moment of simulated noisy peaks were compared with their true values.
► The algorithm gave values of μ′3μ′3 that were within 11% of the true value, S/N1000.
► Eilers’ algorithm reduces the noise and without distorting peak shape.