Algorithms peak detection

In a subsequent development, a novel, real-time peak detection algorithm known as intelligent automated LC/MS/MS (INTAMS) was developed for the analysis of samples generated by in vitro systems.27 INTAMS also requires two separate chromatographic runs for each sample. It allows the user to detect the two most abundant ions of all components and also any predetermined metabolite precursor ions. In the second chromatographic run, INTAMS conducts automatic product ion scanning of molecular ions of metabolites detected in the first full scan analysis. 

When dealing with noise, one should consider use of a peak detection algorithm optimized for CE. This algorithm, which is readily available on some CE systems, will allow a much lower signal-to-noise ratio and an improvement of reproducibility by a factor 1.5—5, and thereby, of the LOD. 

Peak detection algorithms use a predetermined signal threshold level above the basehne, once the threshold is exceeded peak monitoring occurs. Two methods may be used to identify the start and end of a peak (Figure 8.12). 

Yang, C., He, Z.Y., Yu, W.C. (2009) Comparison of Public Peak Detection Algorithms for MALDI Mass Spectrometry Data Analysis. BMC Bioinformatics 10 4. 

Comolli L, Micieli A. Proc. of Optical Fiber Sensing 2011, Ottawa, Canada.Numerical comparison of peak detection algorithms for the response of FBG in non-homogeneous strain fields. 

Yang C, He Z, Yu W. Comparison of public peak detection algorithms for MALDl mass spectrometry data analysis. BMC Bioinformatics. 2009 10 4. doi 10.1186/1471-2105-10-4. 

For an automated compound detection several software packages using different peak detection algorithms are available and usually offered by the MS manufacturer (Hogenboom et al. 2009). The capabilities and limitations of the elemental composition prediction from accurate mass measurements by using heuristic rules (Kind and Fiehn 2006, 2007) or profile MS data after peak shape calibration (Erve et al. 2009) have been comprehensively investigated. 

A typical MALDI spectrum of a bacterial sample has a number of peaks that vary greatly in intensity superimposed on a relatively noisy baseline. This can be problematic for many peak detection routines. Therefore methods that eliminate the need for peak detection also eliminate problems associated with poor peak detection performance. Full-spectrum identification algorithms use the (usually smoothed) spectral data without first performing peak detection. 

Peak purity tests are used to demonstrate that an observed chromatographic peak is attributable to a single component. Mass spectrometry is the most sensitive and accurate technique to use for peak purity evaluation because of the specific information derived from the analysis. However, a good number of HPLC methods use mobile phase conditions that are incompatible with mass spectrometry detection. In this case, PDA spectrophotometers using peak purity algorithms may be used to support the specificity of the method. Almost all commercially available diode array detectors are equipped with proprietary software that will perform these calculations. Although this technique is more universal in application to HPLC methods, the data provided is neither particularly... 

Despite these advancements in chromatographic data processing, peak detection and integration algorithms were crude, the user interface was cumbersome, and there was very little flexibility in the types of reports that these systems could generate. 

Hastings CA, Norton SM, Roy S. New algorithms for processing and peak detection in liquid chromatogra-phy/mass spectrometry data. Rapid Commun Mass Spectrom 2002 16 (5) 462-7. 

Electronic integration zero 0.1 very good (depending on the algorithm for curve smoothing, peak detection, and baseline drift) suited... 

On the other hand, the approach to increase the error tolerance (instead of increase the NOE precision) by a calculation algorithm was studied by Kuszewski et al.90 Usually, the experimental error and peak detection error (due to resonance overlap) are uniformly treated by the margin of error (Section 8.1). The method to separate the two types of errors by statistical analysis may become mandatory in the future. 

Figure 5.9. Diagrams illustrating the use of traditional integration and ApexTrack integration algorithm (Waters EmPower) based on second derivative peak detection. Note that traditional integration can typically handle complex chromatograms well after some manipulations. Diagram courtesy of Waters Corporation.

Fig. 10 Comparison of the efficiency of CLEAN (b) and SSA (c) shown on a simulated signal containing six components of relative amplitudes 1 2 4 8 16 32 and equal decay rates of 20 s . Additionally, white Gaussian noise of tr = 0.02 was present. Both algorithms started from the same initial nuFT spectmm (a), and the same threshold for peak detection equal to 5fr, was used. Spectral width of 4 kHz and max. evolution time of 70 ms were set. Seventy out of 280 points were sampled, yielding a relative sampling density of 0.25...

---