How the "Intensity Threshold" feature works for screening analyses reprocessing with SCIEX OS ?

For research use only. Not for use in diagnostic procedures.

Answer

Identification of a compound during reprocessing of a screening analysis uses the fragments resulting from the fragmentation of compound by tandem mass spectrometry which are compared to a database of fragmentation spectra.

To ensure proper compound identification, clean mass spectra are essential. The acquisition of noisy spectra makes a big difference in results. It is therefore necessary to specify an threshold for all fragments during reprocessing by SCIEX OS to indicate which fragments are considered background noise or not.

Any fragments with an intensity lower than the value indicated by the Threshold will not be considered by SCIEX OS when comparing experimentally acquired fragments with fragments in the mass spectrum database.

The threshold applied is determined from the intensity of the most intense fragment multiplied by the Intensity Threshold. By default, the Intensity Threshold corresponds to 5% of the intensity of the most intense fragment.

It is important to know that SCIEX OS does not use the peak apex intensity of the mass spectrum to determine the intensity of the most intense fragment and the intensity of the fragment which will be compared to the threshold value. Background calculations are performed using Centroided Spectra and the algorithm that considers Sum Intensity above 50% as an intensity value, which is more reliable than considering only one data point with the highest intensity for calculation.

Intensity sum above 50%: For each peak, the Y-value is the sum of the portion of the intensities comprising the peak which are above 50% of the peak apex intensity. This value is useful because it does not depend only on the intensity of a single data point, as the Intensity and Height metrics do, and it is not influenced by the edges of the peak which might be noisy or which might have interference.

The Intensity Threshold value can greatly impact the quality of the identification. If the Threshold is too high, as in the example below, certain key fragments for identification will no longer be taken into account by the SCIEX OS identification algorithm. If the Threshold is too low, background noise will be considered as fragments of the compound to be identified, even this is not really the case.

Example: Identification of a compound from these fragmentation spectra with two different Intensity Thresholds (5% vs 4%)

• If Intensity Threshold = 4% then Threshold = 21802 x 0.04 = 872 → The fragment at m/z = 799.2175 is considered by the SCIEX OS identification algorithm because its "Sum Intensity above 50%" is higher than the Threshold. SCIEX OS will identify the target compound as Vitisin B with a Library score of 99.4 out of 100.
• If Intensity Threshold = 5% then Threshold = 21802 x 0.05 =1090.1 → The fragment at m/z = 799.2175 is not considered by the SCIEX OS identification algorithm because its Sum Intensity above 50% is lower than the Threshold. SCIEX OS will identify the target compound as Vitsin B with a Library score of 22.6 out of 100 (Less fragment available for identification).