Example

A typical contour plot (graphical depiction of the raw data file) from one sample is shown below. Eight consecutive spectra from 20.2 to 20.5 min are extracted. As evident, each of these spectra contains several polypeptides at different charge states. It is impossible to obtain information on charge state and consequently mass and signal intensity on all polypeptides detected using the software supplied with the instrument.

Fig.1 reveals the high information density of typical CE-MS spectra. bottom left: contour plot of a typical CE-MS run. Signal intensity is plotted against m/z and migration time. A section (yellow line) of the CE-MS spectrum is shown enlarged in the figure above. Further magnification of a single peak is shown in the bottom right picture.

Evaluating and deconvoluting a ce-ms spectrum (Fig.2, A and B) with MosaiquesVisu is a three step procedure:

First, ms-peak signatures are identified & extracted using a proprietary detection scheme. Peaks are discriminated by shape of their isotopic envelope, S/N ratio and a-priory knowledge of peptide signatures in TOF-spectra.

MS-peaks are subsequently merged into so-called CEMS-peaks (Fig.2, C) with a finite temporal extend and a well defined location in m/z in migration time. In case of positively resolved isotopic structure, they are assigned a charge state along with a corresponding confidence measure.
The list of merged CEMS-peaks is referred to as CEMS-list and will be made available to you.

In the third step, conjugated peaks are identified using a probabilistic clustering scheme. This specifically tailored algorithm yields a maximum likelihood estimate to the question:

What substances/masses give rise the observed ensemble of cems-peaks in the spectrum.

Fig. 2 Graphical representation of the raw data, CEMS-list and PROT-list: Panel A: Shows the total ion current (upper part) and the summarized spectra (lower part). Panel B: Shows the same spectrum as a two dimensional contour plot of signal intensity over m/z (y-axis) and migration time in min (x-axis).Panel C: Refined CEMS-peaks. Panel D: The deconvoluted spectrum showing the real masses on y-axis.

The outcome of this final evaluation step is a list of substances, localized in mass and migration-time, each bearing a confidence measure for their existence (Fig.2, D). The final list is referred to as PROT-list and will be made available to you.

The output data, CEMS-list and PROT-list can either be downloaded from our server or sent to you via email. Both sets of data can be visualized as contour plot in the Peptex window. Below you see an example of these lists viewed in MS-Excel®.

Below you see an example of these lists viewed in MS-Excel®.

Assoc-Protein	ID tag of polypeptide (derived from mass and migration time)
Assoc-Mass	Mass (based on calculated charge and mass/charge)
Charge	charge based on isotopic distribution and conjugated peaks
Confid.[%]	Confidence of charge assignment
t0[min]	Migration/retention time in spectrum
m/z[Da]	Mass/charge
Counts	Total counts from this peak (amplitude)
SNR	signal to noise ratio

This list is the basis for the right panel containing the information on actual mass of the polypeptides identified.

Label	ID tag of polypeptide (derived from mass and migration time)
t0	Migration/retention time in spectrum
m0	Mass
Counts	Total detector counts of this polypeptide
Prob	Probability for this polypeptide to exist

Using MosaiquesVisu, a typical CE-MS run of a complex sample can be analyzed within several minutes resulting in the peaklist shown above, with an error-rate of less than 2 %. The development of this key element in the use of CE-MS for biomarker discovery enabled the analysis and evaluation of human urine with the aim to pinpoint disease-specific polypeptides.