WO2018134346A1 - Mass spectrometry with improved dynamic range - Google Patents

Abstract

The present invention relates to a method of mass spectrometry (MS) comprising: (a) partitioning an m/z range of interest in two or more regions; (b) defining at least two sets of regions, each set of regions consisting of at least one region as defined in step (a), provided that all sets of regions, when taken together, cover said m/z range in its entirety; (c) ionizing a sample of interest and storing (a) defined number(s) of ions for each region of a given set of regions, thereby obtaining one or more ion populations; (d) combining, if applicable, the more than one ion populations of (c); (e) analyzing the one ion population obtained in step (c) or the combined ion populations obtained in step (d), respectively, in a detector, thereby obtaining a partial mass spectrum; and (f) repeating steps (c) to (e) for each set of regions.

Description

Mass spectrometry with improved dynamic range

The present invention relates to a method of mass spectrometry (MS) comprising: (a) partitioning an mass-to-charge (m/z) range of interest in two or more regions; (b) defining at least two sets of regions, each set of regions consisting of at least one region as defined in step (a), provided that all sets of regions, when taken together, cover said m/z range in its entirety; (c) ionizing a sample of interest and storing (a) defined number(s) of ions for each region of a given set of regions, thereby obtaining one or more ion populations; (d) combining, if applicable, the more than one ion populations of (c); (e) analyzing the one ion population obtained in step (c) or the combined ion populations obtained in step (d), respectively, in a detector, thereby obtaining a partial mass spectrum; and (f) repeating steps (c) to (e) for each set of regions.

In this specification, a number of documents including patent applications and manufacturer's manuals are cited. The disclosure of these documents, while not considered relevant for the patentability of this invention, is herewith incorporated by reference in its entirety. More specifically, all referenced documents are incorporated by reference to the same extent as if each individual document was specifically and individually indicated to be incorporated by reference. Mass spectrometers that trap ions are an important segment of the overall market and they dominate the proteomics market in particular. Examples for this type of mass spectrometers include, not exclusively, quadrupole-ion trap, quadrupole-Fourier transform ion cyclotron resonance, linear ion trap-Orbitrap and quadrupole-Orbitrap instruments. They are very powerful but an inherent limitation is that the number of ions that can be simultaneously stored in an ion trapping device is limited by space charge effects. Exceeding the space charge limit has deleterious effects on the quantitative accuracy, as well as mass resolution and mass accuracy when the trapping device is employed as a mass analyzer. To achieve an optimal performance of the ion trap, it is therefore necessary to control the number of stored ions. One common technique to control the ion population in the trapping device is often referred to as automatic gain control (AGC), in which information about the incoming ion current is used as an estimate to restrict the ion injection time. In other words, if more ions are predicted to enter the ion trap per time interval, the ion accumulation time is shortened and vice versa. Knowledge about the incoming ion current can be obtained from a preceding analytical scan or from a dedicated scan with a known and short ion injection time. Further detail is established in the art and disclosed, for example in US 5,107,109, US 5,559,325, or US 20040217272. An art-established method for mass spectrometry, in which ions from a broad m/z range are mass analyzed simultaneously, is referred to as 'full scan' or 'MS scan'. Applying AGC, or any other method to control the trapped ion population, to perform a full scan on a mass spectrometer that employs ion trapping in at least one step of mass analysis, limits the time in which ions from the m/z range of interest are injected into the ion trap. It is important to notice that typically ions from the entire m/z range of interest enter the ion trap simultaneously and the same ion injection time is applied for all ions from the entire m/z range As a consequence, the total ion injection time for full scans is predominantly determined by highly abundant species. This limits the dynamic range of the mass analysis as the applied ion injection time is insufficient to detect much lower abundant species with adequate signal-to-noise ratios. This is due to the fact that the signal-to-noise ratio at which a given ion is detected, depends on the absolute number of ions of said species in the ion trap rather than its abundance in the initial sample mixture. The dynamic range is understood as the difference in ion abundance between the most and least abundant, detectable ion species. In addition to compressing the dynamic range, short ion injection times reduce the usage of the total ion current generated in the ion source in cases where the time for mass analysis is much longer than the ion injection time. In the analysis of complex compound mixtures, as exemplified by shotgun proteomics, the median ion injection times for full scans are typically <1 ms; less than 0.8 % of the required transient time for a spectrum at a mass resolution of 60,000 at m/z 200 with the high-field Orbitrap cell (128 ms); see, for example, Scheltema, R. A. et al. The Q Exactive HF, a Benchtop Mass Spectrometer with a Pre-filter, High Performance Quadrupole and an Ultra-High Field Orbitrap Analyzer. Mol. Cell. Proteomics M 14.043489- (2014). doi:10.1074/mcp.M1 14.043489; and Kelstrup, C. D. et al. Rapid and Deep Proteomes by Faster Sequencing on a Benchtop Quadrupole Ultra-High- Field Orbitrap Mass Spectrometer. J. Proteome Res. (2014). doi: 10.1021 /pr500985w. Thus, typically more than 99% of the ions are not used for analysis in MS scans.

The ion current, in particular in the analysis of complex compound mixtures, is typically concentrated on a few mass-to-charge channels rather than evenly distributed across the full scan. According to the considerations above, the described limitations of full scans in terms of dynamic range and sensitivity can be partially overcome by focusing the analysis on a narrow m/z range, which is a fraction of the initial m/z range of interest. This is commonly applied in targeted quantification assays, for instance 'single ion monitoring', in which a quadrupole mass filter selects a narrow mass window prior to the trapping device and thus excludes high abundant interferences. This maximizes the ion injection time, and thus signal- to-noise ratio, for the targeted compound of interest; see, for example, Gallien et al., Targeted proteomic quantification on quadrupole-Orbitrap mass spectrometer. Mol. Cell. Proteomics, 2012, 11: 1709-1723. doi: 10.1074/mcp.O112.019802. Similarly, Goodlett et al. (Analytical Chemistry, 2009, 81 (15), 6481-6488. doi: 10.1021/ac900888s) demonstrated that cycling through adjacent narrow precursor mass windows increases the dynamic range of data-independent analyses on a linear ion trap-Orbitrap instrument. However, these methods are limited by the sensitivity, selectivity, and multiply the total acquisition time per sample.

The latter disadvantage, multiplication of the total acquisition time, has been partially addressed by a mass spectrometry method comprising the sequential selection of multiple narrow m/z ranges followed by their combined mass analysis. In WO 2006/129083 a method is described in which AGC is applied to control the ion injection times for said multiple ion packets without exceeding the overall space charge limit of the ion trapping device. However, this method only increases the signal-to-noise ratios of a few selected peaks that have to be known beforehand, not of the whole spectrum.

WO 2014/200987 describes MS involving gas-phase enrichment using notched isolation waveforms. The method described in this document involves simultaneous entry of all ions to be analyzed into the ion trap. A mass filter which would allow partitioning of the m/z range is not applied.

Southam et al. (Anal. Chem. 79, 4595-4602 (2007)) describes MS using a spectral stitching method. This document is silent about combining ion populations corresponding to regions of a set of interspaced regions.

In view of the shortcomings of the art-established methods, the technical problem underlying the present invention can be seen in the provision of an improved method of mass spectrometry. Improvements preferably address the acquisition of full scans and include better signal-to-noise ratio, more efficient use of the incoming ion beam and increased dynamic range.

This technical problem is solved by the subject-matter of the attached claims.

Accordingly, in a first aspect, the present invention relates to a method of mass spectrometry (MS) comprising: (a) partitioning an m/z range of interest in two or more regions; (b) defining at least two sets of regions, each set of regions consisting of at least one region as defined in step (a), provided that all sets of regions, when taken together, cover said m/z range in its entirety; (c) ionizing a sample of interest and storing (a) defined number(s) of ions for each region of a given set of regions, thereby obtaining one or more ion populations; (d) combining, if applicable, the more than one ion populations of (c); (e) analyzing the one ion population obtained in step (c) or the combined ion populations obtained in step (d), respectively, in a detector, thereby obtaining a partial mass spectrum; and (f) repeating steps (c) to (e) for each set of regions. The term "mass spectrometry", abbreviated "MS", has its art-established meaning. In particular, it relates to an analytical method that ionizes chemical compounds and sorts the obtained ions based on their mass-to-charge (m/z) ratio. A common unit of the mass-to- charge ratio is Thomson (Th). The information acquired in a mass spectrometer can be depicted and stored as a mass spectrum. A mass spectrum is a diagram or table, or, in computational terms, an array, where for any given m/z value of interest, the ion abundance is given. Assuming that peaks in the mass spectrometer would not coincide or overlap, a non zero abundance at a given m/z value is a measure of the abundance of a specific ion species. In other words, as used herein, the term "abundance" is generally understood as designating the amount of a specific chemical species.

It will become apparent in the following that, in particular for the purpose of the present invention, the number (as opposed to abundance) of ions acquired or stored in an ion store is an important parameter. When reference is made to a "number of ions", this is generally understood as not being specific for a particular chemical species. Rather, it is the total number of ions, irrespective of the underlying chemical nature, which has been acquired in a given time interval or is being stored in an ion store.

Step (a) of the method in accordance with the first aspect of the present invention provides for partitioning or dividing a given m/z range of interest in two or more regions. It is understood that the m/z range of interest is preferably contiguous. Furthermore, it is understood that said regions in their entirety cover the entire m/z range of interest. In accordance with step (b) of the method of the first aspect, the regions resulting from the partitioning in accordance with step (a) are assigned to sets of regions. Each of the regions is assigned to a set of regions. As a consequence, all sets of regions, when taken together, cover said m/z range in its entirety.

Regions and set of regions are illustrated in an exemplary manner in Figure 1. The three graphs being designated first boxcar scan, second boxcar scan and third boxcar scan each depict a set of regions, wherein each set of regions consists of four regions. The particular shape of the ion transmission function depicted in Figure 1 is preferred, but not limiting. More specifically, in Figure 1 , each region is defined by a boxcar function. As is established in the art, a boxcar function is a function which is zero or the entire real-line except for a single interval where it is equal to a constant. In the exemplary implementation depicted in Figure 1 , said constant is 100% ion transmission. For comparison purposes, the uppermost part of Figure 1 depicts the standard acquisition scheme of a full scan in a mass spectrometer. In accordance with step (c), a sample of interest is ionized and ions for each region of a given set of regions is stored. Ions corresponding to different regions within said given set of regions may be stored in distinct ion stores. Preferred is that they are stored in the same ion store or, equivalently, accumulated in said same ion store. Preferably ions for each region of a given set of regions are selected sequentially (timewise) and stored.

Preferred methods of ionizing are known in the art and include for example electrospray ionization, nano-electrospray ionization, chemical ionization, atmospheric pressure chemical ionization, atmospheric pressure photoionization and matrix-assisted laser desorption ionization.

Step (d) provides for combining the ion populations of (c) for those implementations where at least one set of regions consists of more than one region. To the extent a set of regions consists of only one region (which is not preferred), step (d) is dispensable. The method of the invention also includes (less preferred) implementations where ions from different regions of a set of regions are stored in different ion stores and subsequently combined in step (d). A mass spectrometer that could be used for this implementation is disclosed in US 2010-0314538-A1. Once combined ion populations corresponding to all regions of one given set of regions are obtained, the method in accordance with the first aspect proceeds, in accordance with step (e), to analyzing them in the detector of the mass spectrometer. This yields a mass spectrum, more specifically a partial mass spectrum given that one set of regions does not correspond to the entire m/z range of interest. In order to achieve complete coverage, it is necessary to perform step (f) which consists of repeating steps (c) to (e) for the remaining sets of regions. Said mass analyzing is also referred to as "scan" herein. To the extent the ion transmission function used for defining regions is a boxcar function, scans in accordance with the invention are also referred to as "boxcar scans". Given that preference is given to using a single ion store, a preferred implementation of the methods in accordance with the first aspect provides for accumulating the ion populations originating from the regions of a set of regions in a single ion store. Such preferred implementation is as follows: a method of mass spectrometry (MS) comprising: (a) partitioning an m/z range of interest in two or more regions; (b) defining at least two sets of regions, each set of regions consisting of at least one region as defined in step (a), provided that all sets of regions, when taken together, cover said m/z range in its entirety; (c) ionizing a sample of interest and storing (a) defined number(s) of ions for each region of a given set of regions, thereby obtaining one or more ion populations; (d) accumulating, if applicable, the more than one ion populations of (c); (e) analyzing the one ion population obtained in step (c) or the accumulated ion populations obtained in step (d), respectively, in a detector, thereby obtaining a partial mass spectrum; and (f) repeating steps (c) to (e) for each set of regions.

Furthermore, it is preferred each set of regions consists of at least two regions. As a consequence, a further preferred implementation of the method in accordance with the first aspect is as follows: a method of mass spectrometry (MS) comprising: (a) partitioning an m/z range of interest in four or more regions; (b) defining at least two sets of regions, each set of regions consisting of at least two regions as defined in step (a), provided that all sets of regions, when taken together, cover said m/z range in its entirety; (c) ionizing a sample of interest and storing (a) defined number(s) of ions for each region of a given set of regions, thereby obtaining two or more ion populations; (d) accumulating the two more ion populations of (c); (e) analyzing the accumulated ion populations obtained in step (d) in a detector, thereby obtaining a partial mass spectrum; and (f) repeating steps (c) to (e) for each set of regions. In a preferred embodiment, the method in accordance with the first aspect further comprises (g) combining the mass spectra obtained in steps (e) and (f), thereby obtaining a combined mass spectrum of said m/z range of interest.

The method in accordance with the first aspect provides for improved signal-to-noise ratio (S/N), greater dynamic range, and more efficient use of the incoming ion beam. In fact, when compared to state-of-the-art MS, a much larger proportion (up to 10-fold or more) of the incoming ion current is used for mass analysis and quantification. In more detail:

In a standard full scan, as art-established in mass spectrometry, the total number of ions which eventually end up in the detector is largely dominated by a few highly abundant species. Species of interest which are of low abundance either appear with a bad signal-to- noise ratio or escape detection altogether. Given that the total number of ions which can be handled by ion traps and detectors is limited (see the explanation in the background section herein above), it follows that allowing more ions to enter the ion store and the detector is not an option.

Owing to steps (a) and (b) of the method of the first aspect, less ions originating from highly abundant species and more ions originating from low abundant species enter the ion store and eventually the detector. As a consequence, low abundant species are detected at a better signal-to-noise ratio or, as shown in Figure 2, low abundant species which were absent in the standard full scan, appear at all in the spectrum. Signal-to-noise values (S/N) are indicated for a number of selected peaks in Figure 2. In both parts of Figure 2, the total ion injection time for each scan is given at the top of each figure. Owing to partitioning and defining sets of regions, the total injection time can be dramatically increased. This can be done without running risk of space charge effects because, owing to the design of the method of the invention, more ion injection time is spent with less abundant species. As a consequence, and as stated above, more efficient use is made of the incoming ion beam.

Partitioning in accordance with step (a) of the method of the first aspect may be data- dependent or data-independent. An example of data-independent partitioning is a simple division of the m/z range of interest by the number of regions. If the number of regions is n, the size of each region will then be the n-th part of the m/z range of interest. While overall a significant number of low abundant species will benefit from such data-independent partition scheme, it cannot be excluded that in a given region there is a highly abundant species and a very low abundant species, the low abundant species being of particular interest. Owing to the presence of the highly abundant species in the same region, the low abundant species might still escape detection. One approach of dealing with this situation is to increase the number of regions. As a consequence of increasing the number of regions, the likelihood of a low abundant species of interest co-occurring with a highly abundant species in the same region will decrease. Alternatively or in addition, partitioning may be effected such that highly abundant species become assigned to regions which are narrow on the m/z axis, the consequence being that the abundant species is practically the only species present in such a tailored region. Low abundant species of interest, even if they would be close to the mentioned highly abundant species on the m/z axis would end up in adjacent regions, where more ion injection time would be dedicated to them.

Owing to the design of the method in accordance with the first aspect, the total ion injection time spent will in general exceed by far the injection time used in a standard full scan, assuming that in either case the same total number of ions (such as 3 x 10⁶) would be analyzed. Yet, this can be easily accommodated within the normal duty cycle of the mass spectrometer. Typically 64, 128 or 256 milliseconds are spent for the acquisition of a full mass spectrum. These values refer to the current generation of Orbitrap mass analyzers. The invention also applies to other detectors, for example linear ion traps, 3D ion traps, time- of-flight or FT-ICR mass analyzers. Given that the median ion injection time for art- established full scans is typically below 1 millisecond, it becomes apparent that there is room for spending about two orders of magnitude more time with ion injection, in particular if ion injection follows the procedure in accordance with the first aspect of the present invention. Owing to the fact that in (almost) any case more ion injection time is dedicated to less abundant species, a raw combined mass spectrum obtained in step (g) will generally show low abundance species artificially overrepresented. Given that the ion injection time spent in each region is known, this can be accounted for. Said accounting for is subject of a preferred embodiment and of a separate aspect drawn to a computational method which are also part of this invention and will be disclosed and discussed further below.

In a further preferred embodiment of the method of the first aspect, (i) said defined number of ions is the same for each of said regions; and/or (ii) the total number of ions is the same for each set of regions.

In a further preferred embodiment, the sum of the defined numbers of ions, said sum being over all regions in a given set of regions, (i) does not exceed the total ion capacity of said ion store and/or said detector; and/or (ii) is less or equal about 5 x 10⁵, preferably less or equal about 10⁶, and more preferably less or equal about 3 x 10⁶.

Related thereto, in another preferred embodiment, said defined number of ions is the same for each region and (i) does not exceed the total ion capacity of said ion store and/or of said detector divided by the number of regions per set of regions; and/or (ii) is less or equal about 10⁵.

The above preferred embodiments account for inherent limitations of ion stores and detectors in that beyond the space charge limit, accuracy and precision decrease. A consensus value at present and valid, for example, for the mass spectrometer shown in Figure 5, is three million ions. In a given scan in accordance with the present invention, that value should not be exceeded. Accordingly, the above preferred embodiment limits the number of ions originating from all regions of a given set of regions to less or equal about 3 x 10⁶.

The maximum ion injection time may be used as an additional parameter to restrict the analysis time in cases where the target number of ions is not reached within a reasonable time and waiting for the total ion number criterion as defined above being met would lower throughput of the spectrometer. Optionally, one may apply a double criterion of a maximum for the sum of the defined numbers of ions plus a maximum ion injection time, the latter being, e.g. between about 50 ms and about 200 ms. An exemplary double criterion would be to fill the C-trap with a maximum of 3 x 10⁶ ions in maximum of 100 ms. Such double criterion ensures a high duty cycle of the Orbitrap mass analysis.

In a further preferred embodiment, said sets of regions are interleaved on the m/z axis.

In other words, in accordance with this preferred embodiment, two regions which are adjacent on the m/z axis do not belong to the same set of regions. An example of these interleaved sets of regions is shown in Figure 1 . In case of two sets of regions, regions belonging to the first set and regions belonging to the second set alternate along the m/z axis in accordance with this embodiment.

In a further preferred embodiment, (i) the number of sets of regions is 2, 3, 4, 5 or more, preferably 2, more preferably 3; and/or (ii) the number of regions per set of regions is independently for each set of regions 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15 or more, preferably 12.

In a further preferred embodiment, said regions in each set of regions are interspaced on the m/z axis. Accordingly, to the extent a set of regions consists of more than one region, two neighboring regions within a given set of regions are never adjacent on the m/z axis. Any intervening region or intervening regions belong to one or more different sets of regions. This is illustrated in Figure 1.

Preferred analytes comprised in the sample of interest are one, more or all of peptides, polypeptides, proteins, lipids, carbohydrates, nucleic acids, oligonucleotides and metabolites. Particularly preferred are peptides as they are obtained by an enzymatic digest of one or more proteins. Preferred implementations of said "one or more proteins" are entire proteomes. Preferred enzymes for enzymatic digestion include trypsin and Lys-C.

In a further preferred embodiment, said m/z range is from about 100 to about 2000 Th, preferably from about 300 to about 1650 Th, about 400 to about 1200 Th, or about 500 to about 1000 Th.

In a further preferred embodiment, (i) (i-1 ) said regions are of equal size; or (i-2) the size of said regions depends on the m/z value and/or the ion current, preferably in real time; and/or (ii) the number of regions in each set of regions is equal. Owing to the distribution of cleavage sites for a given enzyme (such as trypsin) in the one or more proteins or the proteome to be analyzed, the expected m/z distribution of the obtained tryptic peptides is, at least approximately, known in advance. In other words, it is known in advance how the ion beam will behave, in a quantitative sense, as a function of m/z. In LC- MS experiments, the behavior of the ion beam can also be superimposed by a function of the retention time.

Also, a standard scan of the m/z range of interest may be performed beforehand in order to acquire knowledge of the location of the most abundant species. Such scan is subject of the preferred embodiment; see step (h) as described further below. Once the location of the most abundant species is known, regions may be specifically tailored to those most abundant species such that each region comprising a given abundant species is very narrow on the m/z axis. Such narrow regions may have width of less or equal 10 Th, less or equal 8 Th, less or equal 4 Th, less or equal 3 Th, less or equal 2 Th, less or equal 1.4 Th, or less or equal 1 Th.

Either type of knowledge can be used to implement a dependency of the size of said regions on the m/z value. ^ ₁

Alternatively or in addition, the ion current may be measured in real-time as a means of controlling ion transmission and/or region size. In a further preferred embodiment, adjacent regions overlap, preferably by about 0.1 to about 10 Th, more preferably about 1 Th.

Adjacent regions are regions which are contiguous on the m/z axis. In view of the preferred embodiments disclosed above, such adjacent regions preferably belong to different sets of regions.

In a further preferred embodiment of the method of the first aspect, (i) said partitioning is effected in a mass filter, preferred mass filters being quadrupole and linear ion trap; (ii) said storing is effected in at least one ion store, preferred ion stores being C-trap and linear ion trap; and/or (Hi) said analyzing is effected in at least one detector, preferred detectors being Orbitrap mass analyzer, time-of-flight mass analyzer, 3D ion trap, linear ion trap and FT-ICR; wherein said mass filter, said ion store(s) and said detector(s) are comprised in (a) mass spectrometers). As regards the mass filter, particular preference is given to a quadrupole mass filter. An exemplary mass spectrometer is shown in Figure 5.

In a further preferred embodiment (i) said storing of (c) and (f) and said combining of (d) and (f) is effected in a single ion store; and/or (ii) said analyzing of (e) and (f) is effected in a single detector.

In a further preferred embodiment, filling of the ion store(s) is controlled by an automatic gain control (AGC) algorithm. The AGC algorithm is known in the art and explained in the introductory section.

In a particularly preferred embodiment, each region is defined by an ion transmission function which is a boxcar function of m/z. Preferably, the constant value of the boxcar function is 100% ion transmission.

In a preferred embodiment, the boxcar functions defining adjacent regions overlap, preferably by about 0.1 to about 10 Th, more preferably about 1 Th. The term "boxcar function" is defined herein above. Figure 1 displays each scan in accordance with the present invention as a series as boxcar functions. By definition, boxcar functions are rectangular in shape when plotted. A set of regions may accordingly be represented as a series of boxcar functions; see Figure 1.

In a further preferred embodiment, the method in accordance with the first aspect of the invention further comprises (h) analyzing, in the absence of prior partitioning, all ions in said m/z range of interest. This step is preferably performed prior to step (c).

Step (h) is a standard full scan of the given m/z interval of interest. It is depicted in the upper part of Figure 1. Given that this is the art-established standard scan, it will suffer from the deficiencies discussed herein above, i.e. worse signal-to-noise ratio, especially of low abundant ion species, as compared to the method of the present invention. Yet, step (h) can be employed for post-processing of the combined mass spectrum obtained in step (g). Said post-processing is subject of a separate aspect of the present invention, namely the computational method discussed further below. A key aspect of said processing is subject of the following preferred embodiment. It is understood that the standard full scan of step (h) can be equal to or wider than the range covered by boxcar scans. To give an example: full scan: m/z 300-1650; all boxcar scans combined: m/z 400-1000.

Accordingly, in a further preferred embodiment, the method in accordance with the first aspect further comprises determining the effective transmission efficiency of the mass filter by comparing ion abundances observed in steps (e), (f) and/or (g) with those observed in step (h).

To explain further, and as noted above, the combined mass spectrum as obtained in step (g), while improving S/N for low abundant species and improving dynamic range, will generally, from a quantitative perspective, be characterized by an overrepresentation of low abundant species. By comparing ion abundances observed in steps (e) and (f) and/or (g) with those observed in step (h), the combined mass spectrum of step (g) can be readjusted to properly reflect the quantities as they occur in the sample of interest. Such readjustment, notably, does not lead to a deterioration of the signal-to-noise ratio which in accordance with the invention is significantly improved when compared to a standard scan in accordance with step (h). Furthermore, it may occur, but does not have to occur, that the practical implementation of partitioning in accordance with step (a), for example by means of boxcar functions is not perfect. In other words, while the ideal boxcar function has exactly 100% ion transmission in the non-zero window, a real boxcar function may be characterized by edge effects in that close to the boundaries of the boxcar function the ion transmission drops below 100%. Also this can be corrected by the mentioned readjusting, said readjusting using the information from the scan performed in accordance with step (h). As noted further above, step (h) can also be used for obtaining information about the overall shape of the spectrum in the m/z range of interest, in particular about locations of the most abundant species. This information in turn can then be used for a data-dependent partitioning of the m/z range in regions. In an alternative preferred embodiment, a preceding boxcar scan (instead of a conventional full scan) could be used for this purpose. In data- dependent partitioning, m/z regions with low ion current may be combined in one larger region.

In a further preferred embodiment, said sample of interest is the eluate of a chromatography device, said chromatography device preferably being coupled online to said mass spectrometer. In preferred embodiments, chromatography is liquid chromatography (LC) or gas chromatography (GC).

In a further preferred embodiment, said method further comprises, after step (f), (g) or (h), performing fragment ion scans of one or more precursor ion species of interest, wherein said precursor ion species is/are selected from (i) the mass spectrum obtained in step (h); (ii) a partial mass spectrum obtained in step (e); and/or (iii) the combined mass spectrum obtained in step (g).

The term "fragment ion scan" refers to a particular operation mode of the mass spectrometer. In a fragment ion scan, the primary ions obtained upon ionizing in accordance with step (c) are subjected to fragmentation. In accordance with established terminology, said primary ions are also referred to as "precursor ions".

In a particularly preferred embodiment, said fragment ion scans are performed (i) separately for each one of said one or more precursor ion species; or (ii) for more than one precursor ion species or all precursor ion species of interest simultaneously. While in the preceding preferred embodiments, further analysis by performing fragment ion scans is driven by knowledge and focuses on specific precursor ion species of interest, in the alternative, fragment ion scans may be performed in a data-independent way. As established in the art, precursor ions of interest are either chosen in data-dependent manner, e.g. ions are ranked by their relative abundance (topN method), or can also be specified in a 'target list'.

Accordingly, in a further preferred embodiment of the method in accordance with the first aspect, said method further comprises, after step (f), (g) or (h), performing one or more fragment ion scans (MS2 scans) across an m/z range of interest, preferably using one or more data-independent m/z selection window(s).

It is understood that "data-independent acquisition" refers to a method that performs a plurality of fragment ion scans one or more times in a single experiment across a mass range using a plurality of pre-determined, i.e., data-independent mass selection windows. These windows are typically broader than those used for data-dependent methods, noting that data- dependent methods typically focus on peaks of interest.

The present invention can advantageously be combined with data-independent acquisition to extend the library strategy disclosed herein below. In particular, while the state-of-the-art is limited by the low dynamic range of regular full scans, this is addressed by the present invention.

Furthermore, the present invention can advantageously be combined with data-independent acquisition to infer a fragment ion spectrum in silico by correlating the chromatographic elution peak of a precursor ion with the chromatographic elution peak of its corresponding fragment ions. The present invention improves the signal of the precursor ions, which is limiting using state-of-the-art full scans. In a particularly preferred embodiment, said mass selection window(s) is/are identical with the regions defined in step (a) or are a subset of the regions defined in step (a).

In a further preferred embodiment, fragmentation is effected by one, more or all of higher energy collisional dissociation (HCD); collision-induced dissociation (CID); electron transfer dissociation (ETD); electron capture dissociation (ECD); ultraviolet photon dissociation (UVPD); and infrared multiphoton dissociation (IRMPD). In a further preferred embodiment, the method of the first aspect further comprises (j) determining one or more physico-chemical property(ies) other than m/z for one or more molecules of interest; (k) determining the m/z for ions obtained in step (c) from said one or more molecules of interest; and (I) identifying said one or more molecules of interest by comparing the information obtained in steps (j) and (k) with a reference data set; thereby identifying and quantifying said one or more molecules of interest.

Mass spectrometry provides (i) identification of species, for instance by fragmenting them, and (ii) quantification of species by their relative ion abundance. Here, the idea is to decouple both parts. Such decoupling becomes possible by relying on a reference dataset. As such, "identifying" in accordance with said preferred embodiment is effected by relying on a reference dataset, and "quantifying" in accordance with said preferred embodiment is effected by the method in accordance with the present invention. Physico-chemical properties in accordance with the present invention include the chromatographic retention time; the relative abundance of isotopes (isotopic distribution); the ion abundance of the identified peptides relative to the ion abundance of other peptides for the same protein (peptide abundance rank); the chromatographic retention time in any preceding dimension of separation (to the extent more than one separation dimension has been used); the collisional cross-section (in those cases where ion mobility spectrometry is coupled to the mass spectrometer); and the fragmentation pattern as seen, e.g., in an MS2 scan.

Comparing in accordance with step (I) can be done either in real-time or in the course of post-processing of the obtained data.

This preferred embodiment distinguishes between a reference dataset and a dataset currently under consideration. As noted above, a separate identification of analytes in the dataset under consideration is not necessary. This avoids time-consuming fragmentation scans in the mass spectrometer (assuming identification were to be performed by means of MS). On the other hand, for the purpose of generating the reference dataset, a fragment scan may be performed. This is the subject of a further preferred embodiment disclosed below. An exemplary implementation is shown in Figure 8. To explain further, and to increase the proteome coverage, i.e. the depth of protein identification, usually more acquisition time and more sample material is required. A common approach is to pre-fractionate complex cell lysates, either on the protein or on the peptide level.

The left panel in Figure 8 (library generation, i.e. generation of a reference dataset) illustrates the pre-fractionation of a tryptic digest into 24 fractions by high-pH reverse-phase chromatography. Each fraction is separately subjected to LC-MS analysis with a standard acquisition method (total acquisition time ~1 day). This experiment yields numerous peptide identifications along with their retention time and exact mass, retention time being a 'physico- chemical' parameter in the sense of the above preferred embodiment.

The goal is always to transfer identifications from an experiment (or database) with maximum depth (here termed library) into another (quantification) experiment, while the latter one is acquired with the boxcar method. For the quantification experiment, this requires only a fraction of the measurement time and sample material, while the same depth of the proteome is achieved.

The 'library generation' could be done as described above, or assembling boxcar LC-MS experiments from a multitude of samples (hundreds or thousands). The quantification experiment could be a 'single shot' (LC-MS of the whole digest without fractionation, illustrated in the figure) or an experiment that includes pre-fractionation itself. In the latter case, the 'library' could have been built from combinations of orthogonal separation methods, with the final one being the same as in the quantification experiment. In a preferred embodiment, said reference data set is (i) contained in a database; and/or (ii) obtained by performing steps (j) and (k) for one or more given molecules.

In a further preferred embodiment, said given molecule has been identified by MS, for example by a fragment ion scan.

In a second aspect, the present invention relates to a mass spectrometer comprising a mass filter, an ion store and a detector, said mass spectrometer being adapted to execute the steps of the method of the first aspect of the invention. A preferred mass filter is a quadropole mass filter.

Preferred implementations of mass filter, ion store and detector are as disclosed herein above. In a preferred embodiment, said mass spectrometer further comprises a record of all acquired mass spectra. The term "adapted to execute" means that the mass spectrometer has a computer program loaded, which computer program is subject of the third aspect of the present invention.

In a third aspect, the present invention relates to a computer program comprising instructions to cause the mass spectrometer of the second aspect to execute the steps of the method in accordance with the first aspect.

Said computer program may be written in any program language. It presents the steps of the method in accordance with the first aspect in a computer readable and/or computer compilable version.

In a fourth aspect, the present invention relates to a computer-readable medium (a) comprising instructions which, when executed on a mass spectrometer, cause said mass spectrometer to execute the steps of the method of the first aspect; and/or (b) having stored thereon the computer program of the third aspect of the present invention. A preferred mass spectrometer in this respect is the mass spectrometer of the second aspect of the invention.

In a fifth aspect, the present invention provides a computational method to quantitatively analyze data generated with the method of MS of the first aspect to the extent said method of MS comprises step (h), said method comprising: (a) aligning on a common m/z grid mass spectra acquired by performing said method of MS; (b) comparing ion abundances observed in steps (e), (f) and/or (g) of said method of MS with those observed in step (h) of said method of MS; (c) determining, from said comparing, relative transmission factors; and (d) (i) multiplying the intensities in the mass spectra obtained in step (e), (f) and/or (g) of said method of MS with the inverse of said relative transmission factors determined in step (c) of said computational method, thereby quantifying ion abundances; or (ii) quantifying ion abundances as the weighted average of all abundances observed for a given ion species in steps (e), (f) and (h) of said method of MS, the weighting factors being the relative transmission factors determined in step (c) of said computational method. As mentioned above, partitioning and defining at least two sets of regions in accordance with steps (a) and (b) of the first aspect of the invention and furthermore implementing the boxcar function in accordance with a preferred embodiment of the first aspect may, when effected in a real-world-setting, suffer from certain imperfections. The purpose of the computational method of the fifth aspect is to account for such imperfections. Imperfections lead to ion transmissions offered by the instrument which may be lower than 100%, for example at the edges of the boxcar function (also referred to as "edge effect").

Said aligning in accordance with step (a) allows for comparability. Aligning refers to ensuring that peaks corresponding to the same ion species in different spectra coincide.

Assuming that transmission in a standard MS scan is 100% throughout, any reduced transmission, for example resulting from the mentioned edge effects, can be determined by comparing in accordance with step (b).

Step (d) finally provides for the desired correction in that data obtained with less than 100% transmissions are scaled up such that they are eventually presented in a manner as if they would have been obtained with 100% transmission. Items (i) and (ii) define two distinct algorithms for correction.

An exemplary implementation is described in Example 4. This principle is illustrated in Figure 9. Figure 9 illustrates this procedure for a method comprising one full scan (depicted in black) and two boxcar scans (depicted in grey) in each scan cycle. The total LC-MS time was 45 min, the method cycle time about 1.2 s. In order to sum up all signals over the retention time (z axis), all scans need to be aligned on a common m/z axis first (step (a)). Next, all scans are summed up. In the shown experiment, this yields a summed mass spectrum for the full scan (black), a summed mass spectrum for the first set of regions and a summed mass spectrum for the second set of regions (grey). A point-by- point comparison of the summed intensities for each set of regions with the full scan yields relative transmission factors for each m/z value, assuming that the transmission for the full scan is 100% at any point. The transmission factors, as calculated from the summed spectra, are then applied to correct the ion abundances in every single scan from every set of regions. Having that, one can also combine all scans from each scan cycle (here: 1 full scan and 2 boxcar scans). In a preferred implementation, the weighted average of all non-zero abundances for each m/z value is calculated. The weighing factors equal the transmission factors calculated above. The resulting hybrid spectra can be subjected to the established post-processing procedures without further adaption (Feature detection and quantification). In a further preferred embodiment, said mass spectra are acquired with the method of MS of the first aspect, to the extent said method employs a chromatography device, and wherein said mass spectra of (e), (f) and/or (g) as well as of (h), respectively, of said method of MS are summed up across retention times as observed during chromatography in said chromatography device.

To explain further, further reliability of the computational method may be attained when data obtained over time (such as from an online coupled chromatography device) are summed up. This preferred implementation is depicted in Figure 9.

In conjunction with the preferred embodiment relating to overlapping edges as disclosed above (adjacent regions overlap, preferably by about 0.1 to about 10 Th, more preferably about 1 Th), ions inside the overlapping regions will be detected in two scans in accordance with step (e) of the method of MS of the first aspect. In this case, the computational method of the fifth aspect may comprise at any point before step (d), preferably before step (a), a step of omitting ions originating from one scan in those regions where scans overlap. This can be done by omitting ions from the scans generating the overlap. Preferably, ions at the high m/z edge from each region in one scan are omitted. Alternatively, in another preferred embodiment, a smaller region at both, the high and low m/z edge of each region in both scans is omitted. Furthermore, but less preferred, edges may also be omitted partially such that there still remain overlapping regions where ions of a given species will be detected and used for quantifying twice, i.e., once in the course of a first scan in accordance with step (e) of the method of MS of the first aspect, and a second time in the course of a second scan (such as a complementary boxcar scan).

It is understood that the remaining ions (those which are not omitted), when taken together, cover the full mass range of interest as defined in step (a) of the method of MS of the first aspect. It is noted that in another aspect of the invention, the use of overlapping edges allows for dispensing altogether with the correction method implemented by the fifth aspect. Also step (h) of the method of MS the first aspect (which step (h) is a prerequisite of the computational method of the fifth aspect owing to the design of said computational method) is then dispensable, but may still be used in an optional manner.

Accordingly, in a sixth aspect, a further computational method is provided which method comprises (a) aligning on a common m/z grid mass spectra acquired by performing said method of MS; (b) omitting ions originating from one scan in those regions where scans overlap; and (c) quantifying ion abundances.

Said omitting is preferably effected as defined above.

Said quantifying is preferably effected by using the ions resulting from step (b) of the computational method of the sixth aspect. Alternatively, said quantifying is effected by determining averages of the abundances of a given ion as observed in step (e) and observed in one or both of steps (f) and (h), said steps (e), (f) and (h) being steps of the method of MS of the first aspect.

The above disclosed data analysis strategies are further illustrated in Figure 14.

As regards the embodiments characterized in this specification, in particular in the claims, it is intended that each embodiment mentioned in a dependent claim is combined with each embodiment of each claim (independent or dependent) said dependent claim depends from. For example, in case of an independent claim 1 reciting 3 alternatives A, B and C, a dependent claim 2 reciting 3 alternatives D, E and F and a claim 3 depending from claims 1 and 2 and reciting 3 alternatives G, H and I, it is to be understood that the specification unambiguously discloses embodiments corresponding to combinations A, D, G; A, D, H; A, D, I; A, E, G; A, E, H; A, E, I; A, F, G; A, F, H; A, F, I; B, D, G; B, D, H; B, D, I; B, E, G; B, E, H; B, E, I; B, F, G; B, F, H; B, F, I; C, D, G; C, D, H; C, D, I; C, E, G; C, E, H; C, E, I; C, F, G; C, F, H; C, F, I, unless specifically mentioned otherwise. Similarly, and also in those cases where independent and/or dependent claims do not recite alternatives, it is understood that if dependent claims refer back to a plurality of preceding claims, any combination of subject-matter covered thereby is considered to be explicitly disclosed. For example, in case of an independent claim 1 , a dependent claim 2 referring back to claim 1 , and a dependent claim 3 referring back to both claims 2 and 1 , it follows that the combination of the subject-matter of claims 3 and 1 is clearly and unambiguously disclosed as is the combination of the subject-matter of claims 3, 2 and 1. In case a further dependent claim 4 is present which refers to any one of claims 1 to 3, it follows that the combination of the subject-matter of claims 4 and 1 , of claims 4, 2 and 1 , of claims 4, 3 and 1 , as well as of claims 4, 3, 2 and 1 is clearly and unambiguously disclosed. The figures show:

Figure 1 : Schematic representation of three complementing, rectangular transmission functions as applied in the boxcar scan method. The overlay of all boxcar scans yields a full spectrum over the entire m/z range as indicated in the uppermost row.

Figure 2: (a) Representative example of a full mass spectrum from a complex tryptic digest, (b) The same spectrum as in (a), but acquired with the method of the present invention. Signal-to-noise ratios (S/N) are annotated for selected peaks and total ion injection time for each scan is depicted on top of the respective spectrum. Ion injection times and thus signal- to-noise ratios were several-fold increased in the boxcar scan. The inserts highlight a single m/z region from the acquired set of regions.

Figure 3: Representative mass spectra from a tryptic digest of a human cancer cell line, (a) Full scan in the center of the LC gradient, (b) Corresponding boxcar scans from the same scan cycle. Boxcar scans were acquired with a box width of 45.4 Th, including a 1 Th overlap. The respective ion injection times in milliseconds are depicted above the corresponding mass spectra. The total AGC target value was set to 3 x 10⁶ for the full scan and 1 x 10⁶ for each of the boxcar scans.

Figure 4: Detection of isotope patterns from human plasma samples in the m/z-retention time plane with a standard full scan method (Panel a) and the method invented herein (Panel b). In the inserts, the median ion injection times per scan are illustrated as a function of m/z (white line). Panel c is a histogram of the number of potential peptide features detected with the standard method (light grey) and the boxcar method (dark grey). Panels d and e illustrate the dynamic range of all detected features as a function of the retention time (d) and m/z (e).

Figure 5: Schematic illustration of a mass spectrometer according to the invention. Arrows indicate the direction of the ion beam.

Figure 6: Application of the "library approach" to single shot proteomics of mouse cerebellum, (a) Number of identified protein groups per replicate and the total number of protein groups in the library, (b) The method provides consistent sampling of over 9,000 protein groups in each replicate, (c) Coverage of different cellular compartments in boxcar single shots as compared with all protein groups present in the library, (d) Ranked abundance of protein groups that were quantified in at least two replicate boxcar single shots. Highlighted protein groups have been previously reported to be specifically enriched in the cerebellum over other brain regions. The density distribution on the right illustrates the relative distribution of proteins associated with key cellular and neuronal functions.

Figure 7: Validation of the quantitative reproducibility of the MS method disclosed in here by replicate injections of a whole protein digest from a human cancer cell line. The method is cross validated with an art-established method. Very high correlations of the protein quantification indicate an excellent quantitative accuracy and reproducibility of the acquisition method as well as the post-processing workflow, including the correction for ion transmission.

Figure 8: Disentangling identification and quantification. Quantification is done by the method of the present invention.

Figure 9: Illustration of the computational method of the invention.

Figure 10: Quantification of (a) peptide (Λ/=26,246) and (b) protein (Λ/=1 ,647) ratios from a human cancer cell line in a two-channel SILAC experiment, acquired in triplicate single runs with the BoxCar method and applying the intensity correction as described in the main text. The heavy and light channels were mixed in a 1 :3 ratio, which is accurately reflected in the density plots.

Figure 11 : Label-free quantification of E.coli lysate mixed with a human cancer cell line (HeLa) lysate in 1 :2 and 1 : 12 ratios (E.coli : HeLa). The scatter plot indicates MaxLFQ ratios of human (red) and E.coli (blue) proteins that were fully quantified in triplicate single runs of each sample with the (a) shotgun (Λ/=5,214) and (b) BoxCar (N=5,699) acquisition method. One-sided student's t-test returns 35% more (in total 962) significantly changing E.coli proteins at a permutation-based FDR below 0.05 for BoxCar as compared with the standard method. Figure 12: Assessment of missing value rates in ten replicate 45 min analyses of HeLa digest, (a) Number of unique peptide sequences quantified in specific numbers of replicates (N) with our standard shotgun method, with 'matching between runs' (MBR) and with BoxCar in conjunction with a matching library, (b) Same for quantified protein groups, (c) Completeness of the peptide quantification matrix as a function of descending peptide abundance in all three experiments, (d) Same for quantified protein groups. Figure 13: Comparison of the number and dynamic range of identified features by matching from a peptide library into single shotgun with standard full scans (dark grey) and BoxCar (light red) runs, (a-c) Analysis of a human cancer cell line digest in a 45 min gradient, (d-f) Analysis of a mouse cerebellum digest in a 100 min single run.

Figure 14: Data analysis strategies with different handling of the edges of regions. In each case, only the grey edge regions are considered for data analysis. Option 1 includes data from both boxcar scans (also in the region of the overlap) and furthermore from the full scan. Option 2 omits data from one boxcar scan in the region of the overlap. Option 3 also dispenses with the full scan.

The examples illustrate the invention.

Example 1

Comparison of standard full scans with the method of the invention

The method of the present invention was implemented on a hybrid quadrupole-Orbitrap mass spectrometer. In the example, the m/z range of interest (m/z = 400-1200) was evenly partitioned into three sets of regions, each set of regions comprising six interleaved regions. The resulting transmission function is referred to as a boxcar function. The acquired boxcar MS spectra resemble closely the standard full scan, but the average ion injection times for each boxcar scan were more than 10-fold higher than the injection time for the corresponding full scan (Fig. 3). Closer inspection revealed that the ion injection time remained low (<1 ms) for boxes (i.e. regions) with high abundant species, but were many-fold increased in m/z regimes with low ion current. As a consequence, signal-to-noise ratios (S/N) for low abundant species were more than 10 to 40-fold increased in boxcar scans as compared with the standard full scan; whereas the signal-to-noise ratios for high abundant species remained nearly constant, as expected from the relative ion injection times.

To demonstrate the improvements characterizing the method of the invention, we applied the method to mass-spectrometry based proteomics of clinical samples, more precisely human plasma samples. Figure 4 shows a two-dimensional representation of detected isotope patterns, i.e. ionized peptides, as they are eluting from the chromatographic column and detected by the mass spectrometer using the art-established full scan (panel a) and the method of the present invention (panel b). Note that, in the standard full scan, the very same ion injection time is applied across the m/z range of interest, which, compared to the method of the invention, compresses the dynamic range of the analysis and discards a large proportion of the generated ion beam. In contrast, the boxcar method makes more efficient use of the incoming ion beam and dedicates high ion injection times to spectral regimes with low ion current. This multiplied the number of detected peptide throughout the LC- S experiment and increased the dynamic range by more than one order of magnitude.

We further validated the quantitative accuracy of the disclosed method by replicate injections of tryptic protein digests from a human cancer cell line (HeLa S3). Pearson correlations close to 1 indicate a very high reproducibility and the cross-validation with the quantification based on art-established full scans did not indicate any systematic biases (Fig. 7).

Example 2

Further validation

To test the quantitative accuracy of the resulting hybrid spectra, we first analyzed a mixture of heavy and light SILAC-labeled HeLa lysate, in which BoxCar accurately quantified the pipetting ratio of 1 :3 throughout the entire m/z range (Fig. 10). In a two-proteome experiment (HeLa and E.coli lysate mixed in 1 :2 and 1 :12 ratios), BoxCar performed particularly well for low abundant E.coli proteins that were quantified with either low accuracy or missing values with the standard method, indicating an approximately 10- fold increased dynamic range (Fig. 1 1 ). Example 3

Application of the library approach

Given the greatly improved dynamic range and number of detected features of the BoxCar method, we asked if comprehensive proteome characterization could be reached by combining single BoxCar runs with a peptide library. In such an approach, peptide identifications would be transferred from the library, for example using the 'match between runs' feature of MaxQuant, while the quantitative information is provided by BoxCar single runs (Fig. 8). Unlike a standard shotgun approach, the library strategy is not limited by the MS/MS scan speed of the mass spectrometer and, with BoxCar, a much larger proportion of the incoming ion current is used for quantification. Figure 6 illustrates the application of the library approach (Fig. 8) for in-depth proteome analysis of mouse cerebellum tissue. Strikingly, nearly the same number of protein groups (>10,000) was identified in 100 min single shot analyses as compared with extensively fractioned samples of the same origin (>1 day of acquisition time). The high proportion of proteins that were quantified in all replicates indicates that the method provides very consistent sampling of the proteome, and thus mitigates a major bottleneck in proteomics often referred to as 'missing value problem'. Further analysis revealed a very high proteome coverage of important cellular compartments and biological functions, which underscores the potential benefits of the method for cellular biology and biomedical research. Figure 13 demonstrates that likewise transferring identifications to standard shotgun runs with conventional full scans is limited by the dynamic range of the full scan, which is expanded up to one order of magnitude with boxcar throughout the LC gradient and mass range. The potential of BoxCar is best exploited in high dynamic range samples such as bodily fluids and tissue, yielding over 100,000 identified features per single run in the cerebellum sample, 60% more than the analogue shotgun/library approach and spanning more than one additional order of magnitude.

Example 4 Alleviating the 'missing value' problem

The semi-stochastic nature of precursor selection in shotgun proteomics results in not every peptide being quantified in every run - often referred to as the 'missing value' problem (Bruderer, R. et al. Extending the Limits of Quantitative Proteome Profiling with Data- Independent Acquisition and Application to Acetaminophen-Treated Three-Dimensional Liver Microtissues. Mol. Cell. Proteomics 14, 1400-1410 (2015); Rost, H. L, Malmstrom, L. & Aebersold, R. Reproducible quantitative proteotype data matrices for systems biology. Mol. Biol. Cell 26, 3926-3931 (2015).). In 45 min single run analyses of 1 ug tryptic peptides from a HeLa cell line each, the complete quantification matrix of ten replicate injections analyzed with a standard shotgun method comprised 37,869 unique peptide sequences and 5,050 protein groups of which only 11 ,626 peptides and 2,789 proteins were quantified in each of the 45 min runs (Fig. 12 a,b). 'Matching between runs' in MaxQuant greatly mitigated this problem by transferring peptide identifications between experiments and lead to complete quantification of 25,653 peptides and 4,180 protein groups in ten replicates. BoxCar in conjunction with the peptide library workflow disclosed herein (Fig. 8) achieved even more consistent profiles across all replicates, providing quantification of approximately 70% of all peptides and >95% of all proteins from the entire shotgun experiment in every replicate. In total, BoxCar quantified 7,222 proteins per 45 min run with a median CV of 10% and more than doubled the proportion peptides and proteins quantified in ten of ten runs to 36,377 and 6,216, respectively. As in any proteomics method, missing values are concentrated in low abundance proteins with peptides close to the level of detection. We therefore calculated the completeness of the data matrices as a function of the peptide and protein abundance rank (Fig. 12 c,d). For the standard shotgun method, both curves decline fast and reach 60% and 75% at the full depth. With matching between runs the curve is much shallower and only 11 % of the peptide and 6% of the protein data points were missing in any of the ten replicates. The BoxCar workflow outperformed both methods in terms of proteome coverage and reproducibility. From the 54,000 most abundant peptides, only 10% of the values are missing and the protein data completeness does not drop below 95% until a depth of 7,450 proteins. Remarkably, even at the full depth of 8,253 protein groups, the fraction of missing data points was as low as 12.5%. Our results compare favorably to reported sparsity levels in data-independent acquisition methods, which are furthermore reached at much lower coverage of the proteome.

Example 5

Raw data processing for BoxCar scans

We developed a new data analysis mode for processing BoxCar scans, which assembles the separate BoxCar scans and the preceding full scan into a single high dynamic range scan. For this purpose the BoxCar scans as well as the full scans are first transformed to a common m/z grid. This is facilitated by having the same mass resolution in each of these scans. The m/z-dependent bin size of the common m/z grid is chosen to be the local mean of the bin sizes of the full scan and the BoxCar scans in the raw data. Note that the typical spacing of the raw data points in 'profile mode' is about ten times more finely grained than the mass resolution. The intensities on the new m/z grid with common binning are calculated from the original scans as linear interpolations from the closest m/z values above and below the m/z value on the new grid. Next, BoxCar scans for each range combination as well as the full scans are summed up over the whole retention time range. The summed full scan as well as the summed BoxCar scans allow us to calculate a transmission function for each BoxCar range as follows. Since the full scan summed over the whole LC-MS run, as well as the summed BoxCar scans have the same common m/z grid, we can divide intensity values point wise. The transmission function for BoxCar scan j is the point-wise ratio between the intensities of the full scan and the BoxCar scan: transmission_j(m/z) = boxcarjntensity m/z) / full_scan_intensity(m/z)

These transmission functions are used to calculate a single high dynamic range scan, by using the transmission function as a weight for a weighted average of the full scan and all BoxCar scans from one acquisition cycle. Importantly, the algorithm does not require any user input but rather adapts to the experimental design, preserving full flexibility in choosing the scan range for full scans and BoxCar scans, the number of BoxCar scans as well as the number of boxes per scan. All subsequent parts of the computational workflow take these scans as input as a replacement for the commonly used MS1 survey scans.

Claims

Claims

1. A method of mass spectrometry (MS) comprising:

(a) partitioning an m/z range of interest in two or more regions;

(b) defining at least two sets of regions, each set of regions consisting of at least one region as defined in step (a), provided that all sets of regions, when taken together, cover said m/z range in its entirety;

(c) ionizing a sample of interest and storing (a) defined number(s) of ions for each region of a given set of regions, thereby obtaining one or more ion populations;

(d) combining, if applicable, the more than one ion populations of (c);

(e) analyzing the one ion population obtained in step (c) or the combined ion populations obtained in step (d), respectively, in a detector, thereby obtaining a partial mass spectrum; and

(f) repeating steps (c) to (e) for each set of regions.

2. The method of claim 1 , wherein the number of regions per set of regions is 2 or more.

3. The method of claim 1 or 2, wherein said partitioning is effected in a quadrupole mass filter.

4. The method of any one of claims 1 to 3, further comprising

(g) combining the mass spectra obtained in steps (e) and (f), thereby obtaining a combined mass spectrum of said m/z range of interest.

5. The method of any one of claims 1 to 4, wherein the sum of the defined numbers of ions, said sum being over all regions in a given set of regions,

(i) does not exceed the total ion capacity of said ion store and/or said detector; and/or

(ii) is less or equal about 5 x 10⁵, preferably less or equal about 10⁶, more preferably less or equal about 3 x 10⁶.

6. The method of any one of the preceding claims, wherein said sets of regions are interleaved on the m/z axis.

7. The method of any one of the preceding claims, wherein said regions in each set of regions are interspaced on the m/z axis.

8. The method of any one of the preceding claims, wherein filling of the ion store(s) used for storing ions is controlled by an automatic gain control (AGC) algorithm.

9. The method of any one of the preceding claims, wherein each region is defined by an ion transmission function which is a boxcar function of m/z.

10. The method of any one of the preceding claims, furthermore comprising:

(h) analyzing, in the absence of prior partitioning, all ions in said m/z range of interest.

11. The method of any one of the preceding claims, wherein said sample of interest is the eluate of a chromatography device, said chromatography device preferably being coupled online to said mass spectrometer.

12. The method of any one of the preceding claims, further comprising

G) determining one or more physico-chemical properties other than m/z for one or more molecules of interest;

(k) determining m/z for ions obtained in step (c) from said one or more molecules of interest; and

(I) identifying said one or more molecules of interest by comparing the information obtained in steps (j) and (k) with a reference data set;

thereby identifying and quantifying said one or more molecules of interest.

13. The method of claim 12, wherein said reference data set is

(i) contained in a database; and/or

(ii) obtained by performing steps (j) and (k) for a given molecule.

14. A mass spectrometer comprising a mass filter, an ion store and a detector, said mass spectrometer being adapted to execute the steps of the method of any one of the preceding claims.

15. A computer program comprising instructions to cause the mass spectrometer of claim 14 to execute the steps of the method of any one of claims 1 to 3. A computer-readable medium

(a) comprising instructions which, when executed on a mass spectrometer as defined in claim 14, cause said mass spectrometer to execute the steps of the method of any one of claims 1 to 13; and/or

(b) having stored thereon the computer program of claim 15.

A computational method to quantitatively analyze data generated with the method of MS of any one of claims 10 or 11 to 13, to the extent claims 11 to 13 refer back to claim 10, said method comprising:

(a) aligning on a common m/z grid mass spectra acquired by performing said method of MS;

(b) comparing ion abundances observed in steps (e), (f) and/or (g) of said method of MS with those observed in step (h) of said method of MS;

(c) determining, from said comparing, relative transmission factors; and

(d) (i) multiplying the intensities in the mass spectra obtained in step (e), (f) and/or (g) of said method of MS with the inverse of said relative transmission factors determined in step (c) of said computational method, thereby quantifying ion abundances; or

(ii) quantifying ion abundances as the weighted average of all abundances observed for a given ion species in steps (e), (f) and (h) of said method of MS, the weighting factors being the relative transmission factors determined in step (c) of said computational method.

The computational method of claim 17, wherein said mass spectra are acquired with the method of MS of any one of claims 11 , 12 or 13, to the extent claims 12 or 13 refer back to claim 11 , and wherein said mass spectra of (e), (f) and/or (g) as well as of (h), respectively, of said method of MS are summed up across retention times as observed during chromatography in said chromatography device as defined in claim 1 1.