Charge state determination of single ion detection events

Abstract

[0006] An exemplary method for classifying the charge state of detected ions may include generating a pulse for each ion in a plurality of ions detected by a detector, each pulse having a pulse characteristic (e.g., pulse height, width, or area), generating a pulse characteristic distribution of the generated pulses, and generating an identification of the charge state of one or more ions in the plurality of ions based on the pulse characteristic distribution.

Description

[Technical Field]

[0001] (CROSS-REFERENCE TO RELATED APPLICATIONS) This application claims the benefit of priority to U.S. Patent Application No. 63 / 024,987, filed May 14, 2020, as a PCT International Patent Application, filed May 14, 2021, the entire disclosure of which is incorporated herein by reference in its entirety. [Background technology]

[0002] Generally, mass spectrometry (MS) is an analytical technique for the detection and quantification of chemical compounds based on analysis of the mass-to-charge (m / z) values ​​of ions formed from those compounds. MS involves ionizing one or more compounds of interest from a sample, generating precursor ions, and mass analyzing the precursor ions. Tandem mass spectrometry or mass spectrometry / mass spectrometry (MS / MS) involves ionizing one or more compounds of interest from a sample, selecting one or more precursor ions of the one or more compounds, fragmenting the one or more precursor ions into product ions, and mass analyzing the product ions.

[0003] Both MS and MS / MS can provide qualitative as well as quantitative information. The measured precursor or product ion spectra can be used to identify molecules of interest. The intensities of the precursor and product ions can also be used to quantify the amount of compounds present in a sample.

[0004] Mass spectrometry techniques often generate mass spectral data that utilize the mass-to-charge ratio (m / z) of detected ions. However, knowledge of the actual charge or mass of detected ions is often not directly measurable. As a result, some degree of overlap of detected ions may occur in certain scenarios. For example, a singly charged ion with a certain mass may appear in a mass spectrum as a doubly charged ion with double the mass and having the same mass-to-charge ratio. This problem may generally be referred to as the peak overlap problem.

[0005] In top-down mass spectrometry (MS) protein analysis, for example, overlapping mass or mass / charge (m / z) peaks in mass spectra is a significant problem. In this type of analysis, a very wide range of different fragment or product ions is generated, including product ions with lengths ranging from 1 to 200 amino acids and 1 to 50 different charge states. Product ion peaks significantly overlap with each other in a single spectrum. In addition, the overlap can be so extensive that even mass spectrometers with the highest mass resolution (Fourier transform ion cyclotron resonance (FT-ICR) or Orbitrap) are unable to deconvolute such overlapping peaks. As a result, large product ions are often lost in top-down protein analysis, limiting sequence coverage of large proteins. International Publication No. WO2020 / 157720, published August 6, 2020, and International Publication No. WO2019 / 197983, published October 17, 2019, both provide additional discussion of top-down MS protein analysis and associated challenges. Summary of the Invention [Means for solving the problem]

[0006] In one aspect, the present technology relates to a method for classifying charge states of detected ions, the method including generating a pulse for each ion in a plurality of ions detected by a detector, each pulse having pulse characteristics; generating a pulse characteristic distribution of the generated pulses; and generating an identification of the charge state of one or more ions in the plurality of ions based on the pulse characteristic distribution.

[0007] In an embodiment, the pulse characteristic distribution is a plot of probability versus pulse characteristic. In another embodiment, the pulse characteristic is at least one of pulse height, pulse width, or pulse area. In a further embodiment, the pulse characteristic is pulse height, and the pulse height is a maximum voltage of the pulse. In yet another embodiment, the detector is an electron multiplying detector, and the detector is configured to primarily detect single ion events. In yet another embodiment, generating an identification of the charge state includes comparing the pulse characteristic distribution to a reference pulse characteristic distribution. In yet another embodiment, the ions detected by the detector are generated from ionization of a sample, and the reference pulse characteristic distribution is identified based on known characteristics of the sample.

[0008] In another embodiment, the generated identification comprises a probability of the charge state. In a further embodiment, the method further includes generating a deconvoluted mass spectrum for the detected ions based on the charge state identification, wherein one axis of the mass spectrum is mass rather than mass / charge (m / z). In yet another embodiment, the plurality of ions are grouped into different groups using the m / z domain, and charge state identification based on the pulse characteristic distribution is performed for each group. In yet another embodiment, the grouping step includes generating a mass spectrum based on the plurality of detected ions, identifying a first peak in the mass spectrum, the first peak having a mass / charge (m / z) value, and grouping ions within a mass / charge (m / z) range based on the m / z value of the first peak. In still yet another embodiment, the grouping step includes selecting a first subset of the plurality of ions in a first intensity band and a second subset of the plurality of ions in a second intensity band; generating a first mass spectrum for the first intensity band; generating a second mass spectrum for the second intensity band; identifying a first peak in at least one of the mass spectra, the first peak having a mass-to-charge (m / z) value; and grouping the ions within the mass-to-charge (m / z) range based on the m / z value of the first peak.

[0009] In another embodiment, the method further includes generating a second pulse characteristic distribution for ions within the m / z range, and generating an identification includes determining that ions forming the first pulse characteristic distribution have a first charge state and determining that ions forming the second pulse characteristic distribution have a second charge state. In a further embodiment, the method further includes determining that one or more isotopes correspond to one or more ions forming the first peak based on the charge state identification. In yet another embodiment, generating an identification of the charge state includes comparing a reference pulse characteristic distribution to the first pulse characteristic distribution. In yet another embodiment, the ions detected by the detector are generated from ionization of a sample, and the reference pulse characteristic distribution is identified based on known characteristics of the sample. In yet another embodiment, the generated identification comprises a charge state probability. In another embodiment, the method is performed as part of a top-down protein analysis.

[0010] In another embodiment, the method further includes identifying a second peak, determining a consensus m / z distance based on at least the first peak and the second peak, and identifying the first peak and the second peak as forming a feature, wherein generating a charge state identification is based on the consensus distance. In a further embodiment, identifying a peak as forming a feature includes comparing pulse characteristic distributions of the peaks and selecting peaks with substantially identical pulse characteristic distributions. In yet another embodiment, comparing the pulse characteristic distributions is performed by calculating a Euclidean distance between the pulse characteristic distributions and comparing it to a predetermined threshold. In yet another embodiment, the method further includes identifying a missing peak corresponding to the feature based on the consensus distance. In yet another embodiment, the method further includes generating a deconvoluted mass spectrum for the detected ions based on the identification of the ions' charge states, wherein one axis of the mass spectrum is mass rather than m / z. In another embodiment, the pulse characteristic is a maximum voltage of the pulse.

[0011] In another aspect, the present technology relates to a mass spectrometry system. The mass spectrometry system includes a detector configured to generate a pulse for each ion detected by the detector, a processor, and a memory storing instructions configured, when executed by the processor, to cause the system to perform a set of operations. The operations include generating a pulse for each ion in a plurality of ions impinging on an electron multiplying detector, each pulse having a pulse characteristic; generating a pulse characteristic distribution of the generated pulses; and generating an identification of the charge state of one or more ions in the plurality of ions based on the pulse characteristic distribution. In an embodiment, the detector is an electron multiplying detector. In another embodiment, the mass spectrometry system further includes an ion source device, a dissociation device, and a mass analyzer.

[0012] In another aspect, the present technology relates to a method for classifying charge states of detected ions, the method including the steps of: using a processor to detect a transient time domain signal induced on an image charge detector of a mass analyzer by oscillation of a plurality of ions in the mass analyzer; converting the transient time domain signal into a plurality of frequency domain (FD) peaks corresponding to ions in the plurality of ions; generating an FD peak characteristic distribution of the generated pulses; and generating an identification of the charge state of one or more ions in the plurality of ions based on the FD peak characteristic distribution.

[0013] In an embodiment, the FD peak characteristic distribution is a plot of probability versus FD peak characteristic. In another embodiment, the FD peak characteristic is peak intensity. In yet another embodiment, generating an identification of the charge state comprises comparing the FD peak characteristic distribution to a reference FD peak characteristic distribution. In a further embodiment, the ions detected by the detector are generated from ionization of a sample, and the reference FD peak characteristic distribution is identified based on known characteristics of the sample. In yet another embodiment, the generated identification comprises a probability of the charge state. In yet another embodiment, the method further includes generating a deconvoluted mass spectrum for the detected ions based on the identification of the charge state, wherein one axis of the mass spectrum is mass rather than mass / charge (m / z).

[0014] In another aspect, the present technology relates to a method for classifying charge states of detected ions, the method including generating a pulse for each ion in a plurality of ions detected by a detector, each pulse having pulse characteristics, generating a pulse characteristic distribution for the generated pulses, identifying a crude charge state based on the pulse characteristic distribution, identifying peak pairs of a first ion peak and a second ion peak such that the peaks have adjacent charge states, and determining a refined charge state of the second ion peak based on the m / z value for the first ion peak, the m / z value for the second ion peak, and the mass of a charge carrier.

[0015] In an embodiment, the crude charge state identification is precise to a range of possible charge states for at least one peak forming the pair, with at least one charge state from the range adjacent to the charge state identified for the second peak. In another embodiment, the method further includes accepting the refined charge state identification if the refined identified charge state is an integer within a threshold. In yet another embodiment, a third peak with an adjacent charge state is identified, the third peak forms a pair with at least one of the peaks, and the charge state identification for the common peak matches in both pairs. In yet another embodiment, the method further includes determining the charge state of the first ion peak based on the determined charge state of the second ion peak. In yet another embodiment, the method further includes obtaining the mass of the charge carrier based on known properties of the sample ionized to produce the plurality of ions.

[0016] This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. Additional aspects, features, and / or advantages of embodiments will be set forth in part in the description that follows, or will be obvious from the description, or may be learned by practice of the disclosure. The present invention provides, for example, the following. (Item 1) 1. A method for classifying the charge state of detected ions, said method comprising: generating a pulse for each ion in the plurality of ions detected by the detector, each pulse having pulse characteristics; generating a pulse characteristic distribution of the generated pulses; generating an identification of the charge state of one or more ions in the plurality of ions based on the pulse characteristic distribution; A method comprising: (Item 2) Item 10. The method of item 1, wherein the pulse characteristic distribution is a plot of probability versus pulse characteristic. (Item 3) 3. The method of any one of items 1-2, wherein the pulse characteristic is at least one of a pulse height, a pulse width, or a pulse area. (Item 4) 4. The method of claim 3, wherein the pulse characteristic is a pulse height, and the pulse height is a maximum voltage of the pulse. (Item 5) 5. The method of any one of items 1-4, wherein the detector is an electron multiplying detector, and the detector is configured to primarily detect single ion events. (Item 6) 6. The method of any one of claims 1-5, wherein generating the charge state identification comprises comparing the pulse characteristic distribution to a reference pulse characteristic distribution. (Item 7) 7. The method of claim 6, wherein the ions detected by the detector are generated from ionization of a sample, and the reference pulse characteristic distribution is identified based on known characteristics of the sample. (Item 8) 8. The method of any one of items 1-7, wherein the generated identification comprises a probability of the charge state. (Item 9) 9. The method of any one of claims 1-8, further comprising generating a deconvoluted mass spectrum for the detected ions based on the identification of the charge states, wherein one axis of the mass spectrum is mass rather than mass / charge (m / z). (Item 10) 10. The method of any one of items 1-9, wherein the plurality of ions are grouped into different groups using m / z domains, and the identification of charge states based on the pulse characteristic distribution is performed on a group-by-group basis. (Item 11) The grouping step includes: generating a mass spectrum based on the plurality of detected ions; identifying a first peak in the mass spectrum, the first peak having a mass-to-charge (m / z) value; grouping ions within a mass-to-charge (m / z) range based on the m / z value of the first peak; Item 11. The method according to item 10, comprising: (Item 12) The grouping step is selecting a first subset of the plurality of ions within a first intensity band and a second subset of the plurality of ions within a second intensity band; generating a first mass spectrum for the first intensity band; generating a second mass spectrum for the second intensity band; identifying a first peak in at least one of the mass spectra, the first peak having a mass-to-charge (m / z) value; grouping ions within a mass-to-charge (m / z) range based on the m / z value of the first peak; Item 11. The method according to item 10, comprising: (Item 13) generating a second pulse characteristic distribution for the ions within the m / z range, wherein generating the identification comprises: determining that ions forming the first pulse characteristic distribution have a first charge state; determining that ions forming the second pulse characteristic distribution have a second charge state; Item 11. The method according to item 10, comprising: (Item 14) 11. The method of claim 10, further comprising determining, based on the identification of the charge state, that one or more isotopes correspond to the one or more ions forming the first peak. (Item 15) 15. The method of any one of claims 10-14, wherein generating the charge state identification comprises comparing a reference pulse characteristic distribution to the first pulse characteristic distribution. (Item 16) 16. The method of claim 15, wherein the ions detected by the detector are generated from ionization of a sample, and the reference pulse characteristic distribution is identified based on known characteristics of the sample. (Item 17) 17. The method of any one of items 10-16, wherein the generated identification comprises a probability of the charge state. (Item 18) 18. The method of any one of items 1-17, wherein the method is performed as part of a top-down protein analysis. (Item 19) identifying a second peak; determining a consensus m / z distance based on at least the first peak and the second peak; identifying the first peak and the second peak as forming a feature; further comprising generating the identification of the charge state based on the consensus distance; 19. The method according to any one of items 11-18. (Item 20) 20. The method of claim 19, wherein identifying the peaks that form the feature comprises comparing pulse characteristic distributions of the peaks and selecting peaks with substantially identical pulse characteristic distributions. (Item 21) 21. The method of claim 20, wherein the comparison of pulse characteristic distributions is performed by calculating a Euclidean distance between the pulse characteristic distributions and comparing it with a predetermined threshold. (Item 22) 22. The method of any one of items 19-21, further comprising identifying missing peaks corresponding to the features based on the consensus distance. (Item 23) 22. The method of any one of items 19-21, further comprising generating a deconvoluted mass spectrum for the detected ions based on the identification of the charge states of the ions, wherein one axis of the mass spectrum is mass rather than m / z. (Item 24) 24. The method of any one of items 1-23, wherein the pulse characteristic is a maximum voltage of the pulse. (Item 25) 1. A mass spectrometry system comprising: a detector configured to generate a pulse for each ion detected by the detector; a processor; a memory storing instructions that, when executed by the processor, cause the system to: generating a pulse for each ion in a plurality of ions impinging on an electron multiplying detector, each pulse having a pulse characteristic; generating a pulse characteristic distribution of the generated pulses; generating an identification of a charge state of one or more ions in the plurality of ions based on the pulse characteristic distribution; a memory configured to cause the device to perform a set of operations including: A mass spectrometry system comprising: (Item 26) 26. The mass spectrometry system of claim 25, wherein the detector is an electron multiplying detector. (Item 27) 27. The mass spectrometry system of any one of items 25-26, further comprising an ion source device, a dissociation device, and a mass analyzer. (Item 28) 1. A method for classifying the charge state of detected ions, said method comprising: using a processor to detect transient time domain signals induced on an image charge detector of the mass analyzer by oscillation of a plurality of ions within the mass analyzer; transforming the transient time-domain signal into a plurality of frequency-domain (FD) peaks corresponding to ions in the plurality of ions; generating an FD peak characteristic profile of the generated pulse; generating an identification of the charge state of one or more ions in the plurality of ions based on the FD peak characteristic distribution; A method comprising: (Item 29) 29. The method of claim 28, wherein the FD peak characteristic distribution is a plot of probability versus FD peak characteristic. (Item 30) 30. The method of any one of items 28-29, wherein the FD peak characteristic is a peak intensity. (Item 31) 31. The method of any one of items 28-30, wherein generating the charge state identification comprises comparing the FD peak characteristic distribution with a reference FD peak characteristic distribution. (Item 32) 32. The method of any one of items 28-31, wherein the ions detected by the detector are generated from ionization of a sample, and the reference FD peak characteristic distribution is identified based on known characteristics of the sample. (Item 33) 33. The method of any one of items 28-32, wherein the generated identification comprises a probability of the charge state. (Item 34) 34. The method of any one of items 28-33, further comprising generating a deconvoluted mass spectrum for the detected ions based on the identification of the charge states, wherein one axis of the mass spectrum is mass rather than mass / charge (m / z). (Item 35) 1. A method for classifying the charge state of detected ions, said method comprising: generating a pulse for each ion in the plurality of ions detected by the detector, each pulse having pulse characteristics; generating a pulse characteristic distribution for the generated pulses; identifying a coarse charge state based on the pulse characteristic distribution; identifying peak pairs of a first ion peak and a second ion peak such that the peaks have adjacent charge states; determining a refined charge state of the second ion peak based on the m / z value for the first ion peak, the m / z value for the second ion peak, and the mass of a charge carrier; A method comprising: (Item 36) 36. The method of claim 35, wherein the crude charge state identification is precise to a range of possible charge states for at least one peak of the pair, and at least one charge state from said range is adjacent to the charge state identified for the second peak. (Item 37) 37. The method of any one of items 35-36, further comprising accepting the refined charge state identification if the refined identified charge state is an integer within a threshold. (Item 38) 38. The method of any one of items 35-37, wherein a third peak with an adjacent charge state is identified, said third peak being paired with at least one of said peaks, and the charge state identification for the common peak being consistent in both pairs. (Item 39) 39. The method of any one of items 35-38, further comprising determining the charge state of the first ion peak based on the determined charge state of the second ion peak. (Item 40) 40. The method of any one of items 35-39, further comprising obtaining the masses of the charge carriers based on known properties of a sample ionized to produce the plurality of ions. [Brief explanation of the drawings]

[0017] Non-limiting and non-exhaustive examples are described with reference to the following figures.

[0018] [Figure 1A] FIG. 1A depicts an exemplary system for performing mass spectrometry.

[0019] [Figure 1B] FIG. 1B depicts an exemplary plot of an ion pulse.

[0020] [Figure 1C] FIG. 1C depicts an exemplary mass spectrum separated into different bands or channels based on ion pulse intensity.

[0021] [Figure 2] FIG. 2 depicts a plot of an exemplary pulse height distribution.

[0022] [Figure 3A] FIG. 3A depicts an exemplary method for charge state assignment.

[0023] [Figure 3B] FIG. 3B depicts another exemplary method for charge state assignment.

[0024] [Figure 3C] FIG. 3C depicts another exemplary method for charge state assignment.

[0025] [Figure 3D] FIG. 3D depicts another exemplary method for charge state assignment.

[0026] [Figure 3E] FIG. 3E depicts another exemplary method for charge state assignment.

[0027] [Figure 3F] FIG. 3F depicts another exemplary method for charge state assignment.

[0028] [Figure 4] FIG. 4 depicts another exemplary method for charge state assignment.

[0029] [Figure 5] FIG. 5 depicts an exemplary expanded mass spectrum.

[0030] [Figure 6] FIG. 6 depicts an exemplary peak detection operation for a fringe mass spectrum.

[0031] [Figure 7] FIG. 7 depicts an exemplary pulse height distribution for a selected peak from FIG.

[0032] [Figure 8] FIG. 8 depicts an exemplary pairwise similarity matrix.

[0033] [Figure 9] FIG. 9 depicts an exemplary calculation of the most likely charge states.

[0034] [Figure 10] FIG. 10 depicts an example plot for performing calculations on feature attributes.

[0035] [Figure 11] FIG. 11 depicts an example plot for adjacent charge state peaks.

[0036] [Figure 12] FIG. 12 depicts an example plot of a transient time-domain signal measured by an image charge detector.

[0037] [Figure 13] FIG. 13 depicts an exemplary system including an image charge detector. DETAILED DESCRIPTION OF THE INVENTION

[0038] Detailed Description As briefly discussed above, peak overlap of detected ions is problematic for analyzing MS results. To solve this peak overlap problem, one solution is to determine or infer the charge state of the ions forming the peaks. By determining the charge state, the mass of the ions can then be resolved, and ions from different species can be distinguished from one another. In addition, multiple peaks in a mass spectrum may represent isotopic clusters or features. However, in some cases, it may be unclear which peaks belong to which clusters. Charge state identification techniques can further resolve the identification of the appropriate peaks for such clusters.

[0039] Analog-to-digital conversion (ADC) banding methods have previously been proposed to separate ions based on their intensity. One such example of a banding method is disclosed in International Publication No. WO 2020 / 157720 ('720 Publication), published August 6, 2020, which is incorporated herein by reference in its entirety. Such separation based on ion intensity promotes charge state separation and therefore improves peak capacity. However, the described method does not teach how ions are separated based on their charge state. This leads to two problems. First, signals from the same species are diluted among multiple data channels. Second, there is no proposed method for constructing a deconvoluted mass spectrum that is convenient for subsequent data interpretation. Therefore, an improved method that can assign charge states to individual ion detection events is desirable.

[0040] One such method has recently been published in the following document: "Multiplexed mass spectrometry of individual ions improves measurement of proteoforms and their complexes" by Kafader et al., Nature Methods, Nature Methods volume 17, pages 391-394 (2020). However, the method described in that document is limited to mass spectrometers with detection systems in which the detected signal has a deterministic relationship to the measured charge (e.g., image charge induction detectors). Therefore, among other things, the method described in that document does not teach how to set the charge assignment for each individual ion measurement event for a mass spectrometer based on a detection system in which the measured signal has a probabilistic relationship to the measured charge (e.g., electron multiplication-based detection systems).

[0041] In some of these new classes of acquisition strategies, direct identification of charge states is attempted prior to co-adding the corresponding signals to the mass spectrum (see Kafader et al.). However, such strategies are not applicable to charge state assignment in electron multiplying detection systems. In many cases involving systems based on electron multiplying detection systems, each charge state does not have a distinct detector response, but instead has a distinct pulse height distribution (or more generally, intensity distribution) specific to each charge state and m / z value.

[0042] This technique allows the determination or inference of the charge state of an ion according to the characteristics of a pulse generated by a detector in response to detecting the ion. To do so, the technique generates a distribution of pulse characteristics for a plurality of detected ions. The pulse characteristics may include pulse height, pulse width, or pulse area, among other possible characteristics. The distribution of pulse characteristics forms distinct profiles depending on the charge state of the detected ions. Thus, the charge state can be determined from the pulse characteristic distribution. Once the charge state of an ion is determined, the mass of the ion can be determined based on the m / z of the ion, and the ion can be distinguished from other ions. Ultimately, the compound being analyzed by the MS technique can be identified based on the determined charge state of the detected ion. Thus, by identifying and / or assigning the charge state of an ion, the measurement capabilities of the mass spectrometry instrument are improved. The accuracy of the mass spectrometry instrument can also be improved.

[0043] FIG. 1A depicts an exemplary mass spectrometry system 100 for performing mass spectrometry techniques. In some embodiments, the system 100 may be a mass spectrometer. The exemplary system 100 includes an ion source device 101, a dissociation device 102, a mass analyzer 103, a detector 104, and computing elements such as a processor 105 and a memory 106. The ion source device 101 may be, for example, an electrospray ion source (ESI) device. The ion source device 101 may be shown as part of a mass spectrometer or may be a separate device. The dissociation device 102 may be, for example, an electron-based dissociation (ExD) device or a collision-induced dissociation (CID) device. Electron-based dissociation (ExD), ultraviolet photodissociation (UVPD), infrared photodissociation (IRMPD), and collision-induced dissociation (CID) are often used as fragmentation techniques for tandem mass spectrometry (MS / MS). ExD can include, but is not limited to, electron capture dissociation (ECD) or electron transfer dissociation (ETD). CID is the most conventional technique for dissociation in tandem mass spectrometers. As explained above, in top-down and middle-down proteomics, intact or digested proteins are ionized and subjected to tandem mass spectrometry. For example, ECD is a dissociation technique that preferentially dissociates peptides and protein backbones. As a result, this technique is an ideal tool for analyzing peptide or protein sequences using top-down and middle-down proteomics approaches.

[0044] The mass analyzer 103 can be any type of mass analyzer used for the desired technique, such as a time-of-flight (TOF), ion trap, or quadrupole mass analyzer. The detector 104 may be any suitable detector for detecting ions and generating signals, as discussed herein. For example, the detector 104 may include an electron multiplying detector, which may include analog-to-digital conversion (ADC) circuitry. The detector 104 may also be an image charge-induced detector. The detector 104 generates a detection pulse for detected ions.

[0045] The computing elements of system 100, such as processor 105 and memory 106, may be included in the mass spectrometer itself, located adjacent to the mass spectrometer, or located remotely from the mass spectrometer. Generally, the computing elements of the system may be in electronic communication with detector 104 such that the computing elements are able to receive signals generated from detector 104. Processor 105 may include multiple processors and may include any type of suitable processing components for processing signals and generating the results discussed herein. Depending on the exact configuration, memory 106 (which stores, among other things, mass spectrometry programs and instructions to perform the operations disclosed herein) can be volatile (such as RAM), non-volatile (such as ROM, flash memory), or some combination of the two. Other computing elements may also be included in system 100. For example, system 100 may include storage devices (removable and / or non-removable), including, but not limited to, solid-state devices, magnetic or optical disks, or tape. System 100 may also have input devices such as a touch screen, keyboard, mouse, pen, voice input, etc., and / or output devices such as a display, speakers, printer, etc. One or more communication connections such as a local area network (LAN), a wide area network (WAN), point-to-point, Bluetooth, RF, etc. may also be incorporated into system 100.

[0046] 1B depicts an example plot 110 of an ion pulse generated from a detector, such as an electron multiplying detector. The y-axis represents intensity and the x-axis represents time. Intensity may be in units of voltage. For example, for an electron multiplying detector, the detector output may be a voltage (often expressed in millivolts (mV)) based on the detected electrons.

[0047] In FIG. 1B, three pulses are depicted: a first pulse 111, a second pulse 112, and a third pulse 113. Pulses 111-113 each represent a different single ion arriving at the detector. Pulses 111-113 can be digitized, and peaks can be found from each digitized pulse. Intensity (or peak height) and arrival time pairs can be calculated and stored for each pulse. Rectangles 131, 132, and 133 represent the intensities or pulse heights of the individual pulses.

[0048] Each pulse may be characterized by pulse characteristics. The pulse characteristics may include characteristics such as pulse height, pulse width, and / or area under the curve of the pulse. The pulse height of each pulse is indicated by rectangles 131, 132, and 132. The pulse height may be the maximum pulse height for the individual pulse, and the pulse height may have units of voltage. The pulse width may be at any point in the pulse, but one measurement of the pulse width may be full width at half maximum (FWHM). The pulse width may have units of time. The area under the pulse curve may be generated by integrating the area under the individual pulse signals for each pulse.

[0049] Pulse characteristics may be used to separate detected ions into different bands. FIG. 1C depicts an exemplary mass spectrum 150 separated into different bands or channels based on ion pulse characteristics, specifically, maximum pulse height. A first mass spectrum 160 is generated for detected ions having a maximum pulse height of 10-20 mV. A second mass spectrum 170 is generated for detected ions having a maximum pulse height of 20-30 mV. A third mass spectrum 180 is generated for detected ions having a maximum pulse height of 20-30 mV. Additional details regarding such separation and banding are further discussed in the '720 publication. As discussed above, separating ions into different bands has benefits, but separation does not allow for identification of the charge state of specific ions. As discussed further herein, pulse characteristics may be utilized to generate distribution profiles that enable charge state classification of ions.

[0050] FIG. 2 depicts an exemplary pulse characteristic distribution plot 200. The pulse characteristic distribution in plot 200 is based on pulse height characteristics. Therefore, the pulse characteristic distribution may be referred to as a pulse height distribution or an intensity distribution. In the plot, the x-axis represents pulse height, and the y-axis indicates the probability or frequency of detection. For example, a higher probability value indicates that ions with the corresponding pulse height were detected more frequently.

[0051] A first pulse height distribution 202 and a second pulse height distribution 204 are depicted in plot 200. As can be seen from plot 200, the pulse height distributions overlap, but first pulse height distribution 202 has a profile that is distinctly different from the profile of second pulse height distribution 204. The difference in profile shape is primarily due to differences in the charge states of the detected ions that form the individual pulse height distributions. For example, the detected ions that form first pulse height distribution 202 correspond to 3+ charged ions, while the detected ions that form second pulse height distribution 204 correspond to 7+ charged ions. Thus, once various pulse height distributions have been established or generated, it may be possible to determine the charge state of any single detected ion by determining the pulse height distribution profile to which the corresponding ion pulse fits.

[0052] As some additional detail, pulse height distributions 202, 204 were generated for product ions with very similar m / z values ​​at approximately 517. The product ions were generated from a top-down ECD analysis of carbonic anhydrase 2 (CA2). As discussed above, as can be seen from plot 200, pulse height distributions resulting from different charge states can overlap significantly. In the case of such overlap, a single intensity data point is insufficient, and any charge state determination based on a single intensity data point will have a significant chance of being incorrect.

[0053] However, it may sometimes be possible to infer charge states for an ensemble of single-ion detection events arising from the same sample. This can be achieved either by comparing the pulse height distribution of such an ensemble to a set of pulse height distributions of known compounds with similar m / z and charge states, or by using redundant information that is typically available based on the properties of the analyzed sample. An example of such redundant information could be distinct isotope patterns, when resolved by a mass spectrometer, that encode charge state information in their relative positions in m / z space. Alternatively, charge state distributions can be used for this purpose, which also encode charge state information in their relative positions in m / z space. This disclosure describes a class of methods for charge state identification of single detection events based on grouping similar detection events in ensembles, the assignment of charge states to ensembles, and the subsequent assignment of charge state information for individual events.

[0054] In an embodiment, a data set containing pulse height distributions for different charge states and m / z is collected. The pulse height distributions for a group of ions may be used to assign charge states to individual detection events. For detection events corresponding to unknown species, an ensemble of detection events may be selected based on their relative proximity in m / z space. A pulse height distribution may be calculated for such an ensemble. This pulse height distribution may then be matched to known pulse height distributions from similar m / z, and the "best" match may be selected. As will be understood, the term "best" may be used to identify a relatively determined optimal state given the available data and the determination effort applied. The best-matching charge state is then inferred for each detection event and / or corresponding ion from the ensemble. Alternatively, or in addition, a model may be constructed based on well-characterized data of intensity distributions for different m / z, and charge states can be predicted based on this model. This may be achieved, for example, using machine learning techniques, where the training dataset may contain annotated data that is collected a priori.

[0055] In another example, in a first step, collected data for individual detection events may be summed to form a conventional mass spectrum, or alternatively, multiple mass spectra based on their intensities. In a second step, an algorithm is employed to perform feature extraction and feature charge state assignment, where features are isotope clusters or charge state clusters corresponding to the same molecule. In a third step, determination of the ensemble of ion detection events corresponding to those features and subsequent inference of charge states is performed.

[0056] In another example, a data acquisition and processing strategy is described that combines both information on the pulse height distributions of individual groups of ions with redundant information known about the properties of the sample, such as isotope patterns or charge state distributions, to identify ensembles of detection events arising from similar ions and subsequently assign charge states to the individual ions.

[0057] FIG. 3A depicts an exemplary method 300 for charge state assignment. In operation 301, a pulse is generated for each of a plurality of ions detected by an ion detector. For example, a pulse is generated when an ion strikes an electron-multiplying detector. The generated pulses may be similar to the pulse depicted in FIG. 1B and discussed above. Each pulse may have or be described by its individual pulse characteristics. The detected ions may be ions ionized from a sample being investigated or analyzed by a mass spectrometry technique. In operation 302, one or more pulse characteristic distributions are generated based on the characteristics of the pulses. For example, the generated pulse characteristic distribution may be generated with respect to a particular pulse characteristic, such as pulse height. Thus, the pulse characteristic distribution generated in operation 302 may be a pulse height distribution similar to the pulse height distribution depicted in FIG. 2 and discussed above.

[0058] Based on the pulse-characteristic distribution generated in operation 302, an identification of the charge state of one or more ions in the plurality of ions may be generated in operation 303. For example, the generated pulse-characteristic distribution may be compared to a set of reference pulse-characteristic distributions with known charge states to determine a closest match to the generated pulse-characteristic distribution. The charge states of the ions forming the generated pulse-characteristic distribution may then be assigned to charge states associated with the reference pulse-characteristic distributions. The number of reference pulse-characteristic distributions to which the generated pulse-characteristic distribution is compared may be limited or reduced based on external information or known characteristics about the sample being analyzed and / or the m / z values ​​of the ions forming the generated pulse-characteristic distribution. For example, a subset of reference pulse-characteristic distributions may exist for particular m / z values ​​or ranges, and / or the subset of pulse-characteristic distributions may correspond to particular isotopes, compounds, and / or samples.

[0059] In other examples, a machine learning model (e.g., a neural network) may be trained based on a set of reference pulse characteristic distributions with known charge states. The generated pulse characteristic distributions may be provided as inputs into the trained machine learning model. The trained machine learning model processes the input generated pulse characteristic distributions, and the output of the trained machine learning model indicates the charge state or likely charge states corresponding to the generated pulse characteristic distribution. For example, the output of the machine learning model may be a charge state indication and / or an indication of the reference pulse characteristic distribution that most closely matches the generated pulse characteristic distribution. In some examples, the input of the trained machine learning model may also include m / z values ​​or m / z ranges for the ions that form the generated pulse characteristic distribution. Additionally or alternatively, the input may also include external data about the sample, such as expected compound or isotope types. In other examples, different machine learning models may be trained on different types of samples, and the machine learning model for the sample being analyzed may be selected for use in analyzing the generated pulse characteristic distribution.

[0060] Charge state identification or assignment may also, or alternatively, be based on grouping peaks together as features and then analyzing the relative distances between the grouped peaks. Additional details regarding such charge assignment processes are discussed in more detail below with respect to method 3000 in FIG. 3D.

[0061] 3A , in operation 304, a mass spectrum is generated for the detected ions. The mass spectrum may be generated based on the charge states identified in operation 303. For example, overlapping peaks with known charge states may be resolved or otherwise represented in the mass spectrum. For example, because the charge states for the ions forming the mass spectrum are known or identified in operation 303, a deconvoluted mass spectrum may be generated for the detected ions. For a deconvoluted mass spectrum, one axis of the mass spectrum may be mass rather than mass / charge (m / z). In operation 305, compounds or amounts of compounds in the sample corresponding to the detected ions may be identified. The compounds or amounts of compounds may be identified from the mass spectrum generated in operation 304 and / or from the charge states identified in operation 303.

[0062] FIG. 3B depicts another exemplary method 310 for charge state assignment. In operation 311, a pulse is generated for each ion in a plurality of ions detected by a detector. Operation 311 may be substantially the same as or similar to operation 301 discussed above. The generated pulses may be stored in memory for subsequent analysis. In operation 312, a mass spectrum is generated based on the m / z values ​​of the detected ions. In operation 313, peaks in the mass spectrum may be selected and / or identified. For example, peaks may be automatically identified through a peak-finding algorithm. In other examples, peaks may be identified and / or selected based on manual input received from a user. Identified peaks have associated m / z positions or values. The m / z positions may be the center locations of the peaks and / or the centroid or weighted average m / z positions over the peaks.

[0063] In operation 314, one or more pulse characteristic distributions are generated for the ions forming the peak identified in operation 313. The ions forming the peak may be ions identified via a peak-finding algorithm. The ions may also be ions that fall within a particular m / z range of m / z values ​​of the identified peak. For example, the m / z range may be selected based on the characteristics of the peak and / or may be a preset range (e.g., fixed m / z values). As an example, the m / z range may be from the beginning of the m / z value of the peak (i.e., where the peak begins) to the end of the m / z value of the peak (i.e., where the peak ends).

[0064] For each ion forming a peak and / or within an m / z range, the corresponding pulse for those ions may be accessed. The pulses may be plotted or stored in a manner showing probability or frequency versus pulse characteristic (e.g., pulse height). For example, the plot may be similar to the plot in FIG. 2, or an array of pulse characteristic and probability / frequency pairs may be generated and / or stored. One or more pulse characteristic distributions may then be identified or generated for the pulse. For example, if peaks are formed from ions with different charge states, multiple pulse characteristic distributions may be generated.

[0065] In operation 315, the charge states of one or more ions forming the identified peaks are identified and / or assigned based on the pulse characteristic distribution generated in operation 314. Operation 315 may be similar to operation 303 discussed above in FIG. 3A , and the step of identifying the charge states may be similarly identified as discussed above. Based on the identified and / or assigned charge states, a mass spectrum may be generated, and / or compounds or compound amounts may be identified, as discussed above with reference to operations 304 and 305.

[0066] 3C depicts another exemplary method 320 for charge state assignment. In operation 321, a pulse is generated for each ion in a plurality of ions detected by a detector. Operation 321 may be the same as or similar to operation 301 discussed above. In operation 322, the detected ions are grouped according to their individual pulse characteristics. As an example, a pulse characteristic of pulse height may be used, and ions may be grouped into intensity bands based on their individual pulse height. For example, a first subset of the plurality of ions may be grouped into a first intensity band and a second subset of the plurality of ions may be grouped into a second intensity band based on the pulse characteristics for each ion.

[0067] In operation 323, a mass spectrum for each intensity band may be generated. For example, if two intensity bands are utilized, a first mass spectrum for the first intensity band may be generated, and a second mass spectrum from the second intensity band may be generated. The mass spectra may be similar to the mass spectrum depicted in FIG. 1C and discussed above.

[0068] In operation 324, one or more peaks are identified from the mass spectrum generated in operation 323. Peak identification and / or selection may be performed in the same or similar manner as operation 313 discussed above. By generating a mass spectrum prior to identifying peaks, background noise may be reduced or removed. For example, one or more of the intensity bands may represent clearer signals and / or more clearly defined peaks than a single aggregate mass spectrum for all ions. Thus, peaks may be more accurately identified.

[0069] In operation 325, one or more pulse characteristic distributions may be generated for ions identified in operation 323 and / or forming peaks within an m / z range of m / z values ​​of the identified peak. For example, a first pulse characteristic distribution corresponding to ions having a first charge state and a second pulse characteristic distribution corresponding to ions having a second charge state may be generated in operation 314. The pulse characteristic distributions may be generated as discussed above, e.g., with reference to operations 315 and 303.

[0070] In operation 326, an identification of the charge state of the ion forming the identified peak is generated based on at least one of the pulse characteristic distributions generated in operation 325. Identifying the charge state from the pulse characteristic distribution may be performed as discussed above. For example, operation 326 may be similar to operation 303 shown in FIG. 3A and discussed above, and identifying the charge state may similarly be identified as discussed above. Based on the identified and / or assigned charge states, a mass spectrum may be generated and / or compounds may be identified as discussed above with reference to operations 304 and 305.

[0071] FIG. 3D depicts another exemplary method 3000 for charge state assignment. FIG. 3D and the following description of method 3000 are illustrative of exemplary steps, with step sets of actions that may be employed for each individual step's task and outcome for an exemplary selected data set. In this example, data from a top-down ECD experiment of carbonic anhydrase 2 (CA2) is used. FIG. 5 depicts an exemplary enlargement or data snippet of a mass spectrum 500 of data from the experiment. More specifically, FIG. 5 depicts an enlarged mass spectrum 500 in the m / z range 517-520 for a top-down ECD experiment of CA2.

[0072] 3D, in step 3010, data from the detector is recorded in a manner such that for each pair of intensity detection events, a corresponding m / z value is recorded and stored. For example, for each pulse generated by the detector (resulting from a detected ion), at least one pulse characteristic (e.g., pulse height, pulse width, pulse area) and the corresponding m / z value for the corresponding detected ion may be stored as a pair. The data may be generated or recorded as discussed above and / or according to the methods described in the '720 publication.

[0073] In step 3020, the recorded data is summed into a single spectrum or multiple mass spectra based on peak intensities, as described above and / or in the '720 publication. In this example, the data is summed to form multiple mass spectra corresponding to different intensity bands. FIG. 6 depicts an example of such a banded mass spectrum. For example, FIG. 6 depicts multiple banded mass spectra 600, including (1) a first mass spectrum 602 for ions having corresponding pulse heights between 20 and 30 mV, (2) a second mass spectrum 604 for ions having corresponding pulse heights between 30 and 40 mV, (3) a third mass spectrum 606 for ions having corresponding pulse heights between 40 and 50 mV, and (4) a fourth mass spectrum 608 for ions having corresponding pulse heights between 50 and 60 mV.

[0074] Turning again to FIG. 3D , in step 3030, a peak detection operation is performed. Various algorithms can be employed for this step, as would be understood by one skilled in the art. For example, a continuous wavelet transform (CWT) algorithm can be used. The highlighting / shading in FIG. 6 indicates the results of a peak detection procedure based on a continuous wavelet transform algorithm. For example, each peak detected by the peak detection algorithm is highlighted / shaded. The lines through each of the peaks depicted in FIG. 6 indicate the m / z value for the individual peak. The outer boundaries of the shading / highlighting may indicate the m / z range for the peak as discussed above.

[0075] In step 3040, a pulse characteristic distribution is calculated for each peak using detected events filtered by their proximity to the peak apex. For example, all pulses corresponding to ions within the highlighted area for a particular peak in FIG. 6 may be used to generate the pulse characteristic distribution. While multiple fringe mass spectra may be generated and used for peak identification (to improve the accuracy of the peak detection process), the pulses used to generate the pulse characteristic distribution are taken from all fringe mass spectra. For example, if the peak selected to generate the pulse characteristic distribution is the peak located at approximately 518.0 in the 30-40 mV mass spectrum 604 in FIG. 6, the pulses used to generate the pulse characteristic distribution include pulses corresponding to ions from all fringe mass spectra 602-608 having m / z values ​​near the m / z value of the peak.

[0076] The generated pulse characteristic distribution may be a pulse height distribution. An example of a calculated pulse height distribution for the multiple peaks of FIG. 6 is shown in FIG. 7. As can be seen from FIG. 7, two clusters of pulse height distributions are formed: a first cluster 702 and a second cluster 704. The first cluster 702 corresponds to ions having a first charge state, and the second cluster 704 corresponds to ions having a second charge state. While the pulse height distributions in each cluster are not the same, two distinct groupings of profiles or distributions are evident from the plot in FIG. 7.

[0077] Returning to FIG. 3D , in step 3050, the pulse characteristic distributions generated in operation 3040 may be compared with each other. The comparison may be performed to group or cluster the pulse characteristic distributions with each other. Based on the grouping or clustering, ions and / or peaks belonging to different isotopes may be grouped together. One way to perform such a comparison is to calculate the relative distance between the pulse characteristic distributions. For this, appropriate binning may be employed, and the pulse characteristic distributions may be represented as vectors containing probabilities for each intensity range. The Euclidean distance or any other appropriate norm may be calculated for each pair of vectors representing the intensity distributions. An example of the calculated pairwise Euclidean distance is shown in table 800 depicted in FIG. 8 . Peaks having corresponding pulse characteristic distributions with a relative distance less than a predefined threshold may be grouped together to form a feature. In an exemplary analysis, peaks are grouped into features if the relative distance between each pulse characteristic distribution within such a group is less than 0.1, thus forming two separate groups of peaks corresponding to two distinct features. For example, two peaks identified in a mass spectrum may have corresponding pulse characteristic distributions that are very similar to each other (i.e., in the same cluster). The pulse characteristic distribution similarity indicates that the peaks are likely formed from ions with the same charge state. Thus, the peaks may be grouped together as features and considered to be part of an isotope cluster.

[0078] Returning to FIG. 3D , in step 3060, the charge states of the features are identified, assuming that the peaks form isotope clusters. To accomplish this, the peaks corresponding to or grouped with the features in operation 3050 may first be sorted in ascending or descending order. The distance between adjacent peaks corresponding to the features can be calculated in m / z space (e.g., the m / z distance between the first and second peaks). If the consensus distance can then adequately account for other observed distances (e.g., other distances that are multiples of the consensus distance), a distance can be selected based on either the most abundant (and therefore most likely) distance for a particular feature or the smallest distance. The distance between adjacent isotopes in an isotope cluster is inversely proportional to the feature charge, and therefore the feature charge can be inferred from those distances.

[0079] An example of such a calculation is shown in FIG. 9. FIG. 9 depicts two features in which peaks are grouped based on the similarity of their corresponding pulse characteristic distributions. For example, for the first feature, four peaks are grouped together. For the second feature, three peaks are grouped together. The peak locations and m / z distances between the peaks are shown for each individual peak. Based on the distances, predicted charge states can be generated based on the reciprocal of the distances. For example, the reciprocal of 0.142 is 7.04 (i.e., 1 / 0.142 = 7.04), and the reciprocal of 0.334 is 2.99 (i.e., 1 / 0.334 = 2.99). The most frequently predicted charge or distance may then be used to identify or assign a charge state. For example, for the first feature, the most common predicted charge state due to the m / z distance is approximately 7. Because charge states must be integers, the consensus charge state is then determined to be 7. A similar calculation is performed for the second feature to determine that the consensus charge state is 3. Notably, the distance between peak 2 and peak 3 of the first feature differs from the consensus distance; addressing that difference is discussed further below.

[0080] 3D , method 3000 may continue to step 3070, where missing peaks may be calculated. Multiple strategies may be used to perform this step. For example, to find missing interior peaks, an algorithm first finds a distance between adjacent peaks that is greater than the consensus distance and is substantially a multiple of the consensus distance. A multiplication factor N is then calculated as N = (measured distance) / (consensus distance). An extra N-1 peaks can be inserted between the peaks at their respective positions to form a complete isotopic cluster.

[0081] An example of the steps of using such an algorithm can be demonstrated with reference to FIG. 9. In FIG. 9, the distance between peak 2 and peak 3 of the first feature is 0.286, which is greater than the distance between the other peaks in the first feature. The consensus distance for the first feature is approximately 0.143. Using the above calculation, N = 0.286 / 0.143 = 2. Therefore, one (i.e., N-1) peak may be inserted between peak 2 and peak 3. Subsequent steps provide an extra peak that describes the isotope cluster. The peak is located at m / z position 517.994 m / z (calculated by averaging the m / z of labeled peaks 2 and 3).

[0082] Another algorithm may also be employed to search for missing peaks, alternatively or additionally, and is not limited to finding only inner peaks. The algorithm calculates the locations of possible neighboring peaks and then extracts the recorded signal corresponding to such locations. This signal is further processed to form a pulse characteristic distribution, which can then be compared to one (or alternatively, the average) of the pulse characteristic distributions calculated for the peaks in the group using similar methods, as described in steps 3050-3060. The identified peaks are then applied to the feature peak list.

[0083] Step 3080 may include performing an analysis to find overlapping peaks in multiple features. This step can be accomplished by comparing peak positions in each feature and finding peaks at substantially the same position. For example, from the previous step 3070, a newly found peak with an m / z of 517.994 is at substantially the same position as Peak 0 from Feature 2 (see FIG. 9), which has a mass of 518.003. Overlapping peaks may be identified by finding the m / z difference or distance between the peaks and comparing it to an overlap threshold. If the m / z distance is below the threshold, the peaks may be considered overlapping.

[0084] Once the overlapping peaks are identified, a further step of finding the contribution of each feature to the overlapping peaks can be performed. For this, a system of linear equations can be written and approximately solved, with or without constraints. Such constraints can include the requirement that each contribution be non-negative. This step can be performed, for example, using a non-negative least-squares approximation algorithm.

[0085] In step 3090, a detection event (e.g., an ion corresponding to a pulse) may be assigned to a feature. Step 3090 may be performed, for example, using the following algorithm: First, using the consensus distance from step 3070 and one or more additional instrument parameters, such as instrument resolution, an m / z distribution can be modeled for each peak for the feature. Second, using the consensus pulse characteristic distribution for the feature (which can be calculated as the average of the pulse characteristic distributions of all non-overlapping peaks) and the m / z distribution from the first part of this step, two values ​​can be calculated that reflect the probability of the feature for having such a detection event. These values ​​are the intersection of the m / z position and the calculated m / z distribution, and the similarity intersection of the detection event intensity and the consensus intensity distribution attributed to the feature. The production of these values ​​is a score that can be used to attribute a detection event to a feature using a threshold.

[0086] An example of calculating such a score may be provided with reference to FIG. 10, which depicts a modeled peak 1000 and a consensus pulse profile distribution 1050. A detected ion to be assigned had an m / z value of 517.994 and a pulse height of 34. These values ​​are indicated by vertical lines in the plot. The m / z value of 517.994 intersects the modeled peak at an intensity of 0.31 (note that intensity has been normalized to 1 for the modeled peak). The pulse height of 34 intersects the consensus pulse profile distribution at 0.25. Therefore, the score may be calculated to be 0.0775 (i.e., 0.31 × 0.25 = 0.0775).

[0087] In the case of multiple overlapping features, the feature that results in the highest score is selected. Additional constraints that balance the total contribution of the features to the overlapped peak may also be set and implemented. Additional or alternative algorithms may also be implemented in connection with this step to determine the most appropriate feature for assigning a detection event corresponding to a detected ion. Such an algorithm may estimate the probability of a detection event belonging to a certain feature by using a Bayesian structure, or the like. In step 3010, a feature charge state is assigned to the detection event.

[0088] FIG. 3E depicts another exemplary method 3200 for charge state assignment. In operation 3202, a pulse is generated for each ion in a plurality of ions detected by a detector. For example, a pulse is generated when an ion strikes an electron-multiplying detector. The generated pulses may be similar to the pulse depicted in FIG. 1B and discussed above. Each pulse may have or be described by its individual pulse characteristics. The detected ions may be ions ionized from a sample being investigated or analyzed by a mass spectrometry technique. In operation 3204, one or more pulse characteristic distributions are generated based on the characteristics of the pulses. For example, the generated pulse characteristic distribution may be generated with respect to a particular pulse characteristic, such as pulse height. Thus, the pulse characteristic distribution generated in operation 3204 may be a pulse height distribution similar to the pulse height distribution depicted in FIG. 2 and discussed above.

[0089] In operation 3206, ion peaks having adjacent charge states are identified based on the pulse characteristic distribution. Identifying peaks having adjacent charge states may include determining an estimated or coarse charge state based on the pulse characteristic distribution. The coarse charge state identification may be accurate only to a range of possible charge states for at least one peak of the pair, where at least one charge state from this range is adjacent to the charge state identified for the second peak.

[0090] An example of ion peaks with adjacent charge states is shown in exemplary plot 1100 in Figure 11. A first peak is located at an m / z position of (m / z) 1 and a second peak has an m / z position of (m / z) 2. The charge states of the peaks may be estimated based on the pulse characteristic distribution generated in operation 3204.

[0091] In operation 3208, the charge state of the ions forming the peak may also be determined based on the following equation: [ka]

[0092] Equation 1 expresses the relationship between the m / z position of the second peak (z2) with the non-charge carrier mass of the ion (M), the charge state of the second peak ((m / z)2), and the mass of the charge carrier (X). Equation 2 expresses the relationship between the m / z position of the first peak ((m / z)1) with the non-charge carrier mass of the ion (M), the charge state of the second peak (z2), and the mass of the charge carrier (X). Notably, Equation 2 assumes that the charge state difference between the first and second peaks is 1. Therefore, in other embodiments where the estimated charge state difference between the first and second peaks is a value other than 1, the 1 in Equation 2 is replaced with that value.

[0093] Based on Equations 1 and 2, Equation 3 expresses the relationship between the m / z position of the second peak ((m / z)2), the m / z position of the first peak ((m / z)1), and the charge state for the ions in the second peak (z2) along with the mass of the charge carrier (M). Each of these values ​​is either measured by a detector or is known from the ionized sample and / or sample preparation process. For example, the common charge carrier is a proton, which has a mass of approximately 1 atomic mass unit (AMU). Thus, the charge state of the ions forming the second peak (z2) may be determined using Equation 3. Based on the determined charge state for the ions forming the second peak (z2), the charge state for the ions forming the first peak (z1) may be determined. The determined charge state may be a refined charge state that is compared to the initially estimated or determined crude charge state.

[0094] The refined charge state should be an integer or an integer nearby, such as within a threshold of an integer. If not, the coarse charge state identification or the refined charge state identification may be incorrect. Thus, the coarse and / or refined charge state identification may only be accepted if the refined identified charge state is an integer within a threshold. If not, the coarse charge state may be re-estimated, and the method is performed again with the revised coarse charge state. Additionally, to potentially increase the confidence in the assignment, a third peak with an adjacent charge state may be identified. The third peak forms a pair with at least one of the first two peaks, and the charge state identification for the common peak matches in both pairs.

[0095] FIG. 3F depicts another exemplary method 3300 for charge state assignment. Method 3300 utilizes an image charge detector. Unlike an ADC detector, an image charge detector detects the oscillation of ions in a mass analyzer. FIG. 12 depicts an exemplary plot of a transient time-domain signal measured by an image charge detector, including components from each of a plurality of ions oscillating in a mass analyzer. To resolve the transient time-domain signal measured by the image charge detector into individual components, the transient time-domain signal is transformed into a frequency-domain signal. Transformation methods include, but are not limited to, Fourier transform or wavelet transform. Peaks in the frequency-domain signal correspond to individual ions of the plurality of ions oscillating in the mass analyzer. The frequency-domain peaks are transformed into m / z peaks using well-known formulas that depend on the specific type of mass analyzer to generate a mass spectrum.

[0096] For image charge detectors, the intensity of the frequency-domain signal or peak is therefore proportional to the charge state of the underlying ion, similar to how the pulse described above is proportional to charge state. Thus, the intensity or other characteristics of the frequency-domain (FD) peak may be used to generate a distribution similar to the pulse characteristic distribution discussed above. A distribution generated from the characteristics of the FD peak may be referred to as an FD peak characteristic distribution or an FD peak intensity distribution, where the intensity of the FD peak is used as the characteristic focus. The FD peak characteristic distribution may then be used in substantially the same manner as the pulse characteristic distribution to determine the charge state.

[0097] 3F, in operation 3302, a transient time-domain signal induced on an image charge detector of the mass analyzer by oscillation of a plurality of ions in the mass analyzer is detected. In operation 3304, the transient time-domain signal is converted into a plurality of frequency-domain (FD) peaks. Each frequency-domain peak may correspond to an ion of the plurality of ions.

[0098] In operation 3306, one or more FD peak characteristic distributions are generated. For example, the generated FD peak characteristic distributions may be generated with respect to a particular FD peak characteristic, such as intensity. In operation 3308, an identification of a charge state of one or more ions in the plurality of ions detected in operation 3302 is generated based on the one or more FD peak characteristic distributions generated in operation 3306. Identifying the charge state based on the FD peak characteristic distribution may be performed using any of the methods described herein using the pulse characteristic distribution. For example, the FD peak characteristic distribution may be used instead of the pulse characteristic distribution.

[0099] In operation 3310, a mass spectrum is generated for the detected ions. The mass spectrum may be generated based on the charge states identified in operation 3308. For example, overlapping peaks with known charge states may be resolved or otherwise represented in the mass spectrum. For example, because the charge states for the ions forming the mass spectrum are known or identified in operation 3308, a deconvoluted mass spectrum may be generated for the detected ions. For a deconvoluted mass spectrum, one axis of the mass spectrum may be mass rather than mass / charge (m / z). In operation 3312, compounds or amounts of compounds may be identified in the sample corresponding to the detected ions. The compounds or amounts of compounds may be identified from the mass spectrum generated in operation 304 and / or from the charge states identified in operation 3312.

[0100] FIG. 13 depicts an exemplary system 1300 including an image charge detector 1318. The system of FIG. 13 includes a mass spectrometer 1310 and computing components including a memory and a processor 1320. The computing elements of the system, such as the processor 1320 and memory, may be included in the mass spectrometer itself, located adjacent to the mass spectrometer, or located remotely from the mass spectrometer. Generally, the computing elements of the system may be in electronic communication with the detector 1318 such that the computing elements are able to receive signals generated from the detector 1318. The processor 1320 may include multiple processors and may include any type of suitable processing components to process the signals and generate the results discussed herein.

[0101] The mass spectrometer 1310 includes a mass analyzer 1317. The mass analyzer 1317 includes an image charge detector 1318. The image charge detector 1318 generates an oscillatory signal or a transient time-domain signal for the detected ions with an amplitude proportional to the ion charge state. The mass analyzer 1317 can be any type of mass analyzer capable of detecting ions using an image charge detector, including, but not limited to, an electrostatic linear ion trap (ELIT), an FT-ICR, or an Orbitrap mass analyzer. The mass analyzer 1317 is shown in FIG. 13 as an ELIT, and the image charge detector 1318 is shown as the pickup electrode of the ELIT.

[0102] The mass analyzer 1317 detects a transient time-domain signal 1319 induced on an image charge detector 1318 by the oscillation of the ions in the mass analyzer 1317. The ions are transmitted to the mass analyzer 1317 by the mass analyzer 1310. The processor 1320 converts the transient time-domain signal 1319 into a plurality of frequency-domain pulses or peaks 1321, each frequency-domain signal corresponding to an ion of the plurality of ions. The processor 1320 converts the transient time-domain signal 1319 into a plurality of frequency-domain peaks 1321 using, for example, a Fourier transform.

[0103] The processor 1320 may compare the intensity of each frequency-domain peak of the plurality of frequency-domain peaks 1321 to two or more different predetermined intensity ranges corresponding to two or more different charge state ranges. The processor 1320 may store each frequency-domain peak in one of two or more data sets 1322 corresponding to the two or more predetermined intensity ranges based on the comparison. The processor 1320 may create a mass spectrum based on the frequency-domain peaks and / or the identified charge states discussed herein.

[0104] In various embodiments, the processor 1320 transforms the transient time-domain signal 1319 into a plurality of frequency-domain peaks 1321, compares the intensity of each frequency-domain peak to two or more different predetermined intensity ranges, and stores each frequency-domain peak during acquisition in one of two or more data sets 1322. In an alternative embodiment, the processor 1320 transforms the transient time-domain signal 1319 into a plurality of frequency-domain peaks 1321, compares the intensity of each frequency-domain peak to two or more different predetermined intensity ranges, and stores each frequency-domain peak after acquisition in one of two or more data sets 1322.

[0105] As explained above, if multiple copies of the same ion oscillate simultaneously in the mass analyzer 1317, the measured intensity may not be proportional to the charge state. As a result, in various embodiments, the mass analyzer 1310 transmits ions to the mass analyzer 1317 such that the mass analyzer 1317 contains only a single ion of a particular m / z and charge state at any given time.

[0106] In various embodiments, the system of Figure 13 further includes an ion source device 1311. The ion source device 1311 can be, for example, an electrospray ion source (ESI) device. Although the ion source device 1311 is shown in Figure 13 as part of the mass spectrometer 1310, it can also be a separate device.

[0107] In addition, the mass spectrometer 1310 further includes a dissociation device, which can be, but is not limited to, an ExD device 1315 or a CID device 1313. The dissociation device can be used, for example, for top-down protein analysis.

[0108] In top-down protein analysis, the ion source device 1311 ionizes proteins in a sample to generate precursor ions for the proteins in the ion beam. The dissociation device dissociates the precursor ions in the ion beam to generate product ions with different charge states in the ion beam. The mass spectrometer 1310 transmits the product ions to the mass analyzer 1317, such that the product ions are transmitted by the mass spectrometer 1310 to the mass analyzer 1317, as described above.

[0109] In various embodiments, the processor 1320 is used to control or provide instructions for the ion source device 1311 and the mass spectrometer 1310, and to analyze collected data. The processor 1320 controls or provides instructions by controlling, for example, one or more voltage, current, or pressure sources (not shown).

[0110] FIG. 4 depicts another exemplary method 400 for charge state assignment. Method 400 may be particularly useful when the isotope distribution is not well resolved for at least some peaks. Operations 401-404 may be substantially identical to operations 3010-3040. In operation 405, a charge state for a feature may be estimated. A feature may be formed from multiple peaks sharing similar pulse characteristic distributions, and grouping of peaks to form a feature may be performed as discussed above for operations 3050-3060. In operation 406, for every possible charge state estimated in operation 405, neighboring peaks may be identified, and likelihood may be scored for the neighboring peaks. In operation 407, the most likely candidate may be selected based on the scoring performed in operation 407. A detected event may then be assigned to a feature in operation 408, and a feature charge state may be assigned to the detected event based on the charge state of the feature.

[0111] In another embodiment, the method in Figures 3A-3E may be combined with the method in Figure 4. In such an embodiment, peaks with fully resolved isotopic patterns may be processed using the strategy described in method 3000 of Figure 3D. For the remaining peaks with substantially unresolved isotopic patterns, the strategy from Figure 4 may be used. In such a combination, an extra step may be employed to define whether a found peak is isotopically resolved. This step may, for example, compare the peak width of the detected feature with characteristic peak widths for the present mass spectrometer for resolved and unresolved features.

[0112] For all these methods, three positive outcomes can generally be envisioned and may be practically the same in their usefulness. First, using information about individual detected events (e.g., charge state), the deconvoluted mass spectrum can be calculated using the formula (m / zm p ) * Z, where Z is the determined charge state, and m pis the mass of a proton. Second, this information may be used to construct a set of spectra for each charge state covering either the entire m / z range or a segment of the m / z range. Third, this information may be used to generate a list of each individual feature with an m / z and z assigned to it. Compounds and / or amounts of compounds present in the analyzed sample may then be determined or generated based on the identified features.

[0113] For all described embodiments, the step of assigning a charge state to an individual detection event can be replaced by or include the step of assigning a probability of that detection event arising from an ion forming a feature, and thus being associated with the charge state assigned to that feature. Such probabilities can be calculated, for example, using a Bayesian structure. In a subsequent step of matching representative mass spectra (either in whole or in part), the proportional contribution from the detection event can then be distributed, as appropriate, among the features it represents along with its individual position within the mass spectrum.

[0114] While the present teachings will be described in conjunction with various embodiments, it is not intended that the present teachings be limited to such embodiments. On the contrary, the present teachings encompass various alternatives, modifications, and equivalents, as will be appreciated by those skilled in the art.

[0115] For example, aspects of the disclosure are described above with reference to block diagrams and / or operational illustrations of methods, systems, and computer program products according to aspects of the disclosure. The functions / acts noted in the blocks may occur out of the order shown in any flowchart. For example, two blocks shown in succession may, in fact, be executed substantially in parallel, or the blocks may sometimes be executed in the reverse order, depending on the functionality / acts involved. Furthermore, as used in this specification and in the claims, the phrase “at least one of element A, element B, or element C” is intended to convey any of element A, element B, element C, elements A and B, elements A and C, elements B and C, and elements A, B, and C.

[0116] The summary and illustrations of one or more aspects provided herein are not intended to limit or restrict in any way the scope of the claimed disclosure. The aspects, examples, and details provided herein are deemed sufficient to convey proprietary rights and to enable others to make and use the claimed disclosure in its best mode. The claimed disclosure should not be construed as limited to any aspect, example, or detail provided herein. Various features (both structural and methodological), whether shown and described in combination or separately, are intended to be selectively included or omitted to produce embodiments with a particular set of features. Given the summary and illustrations herein, those skilled in the art may envision variations, modifications, and alternative aspects that fall within the spirit of the general inventive concept in its broader aspects embodied herein without departing from the broader scope of the claimed disclosure.

Claims

1. 1. A method for classifying the charge state of detected ions, said method comprising: generating a pulse for each ion of the plurality of ions detected by the detector, each pulse having a pulse characteristic; generating a first pulse characteristic distribution of the generated pulses; generating a charge state identification of one or more ions of the plurality of ions based on the first pulse characteristic distribution; A method comprising:

2. The method of claim 1 , wherein the first pulse characteristic distribution is a plot of probability versus pulse characteristic.

3. The method of any one of claims 1 to 2, wherein the pulse characteristic is at least one of pulse height, pulse width, or pulse area.

4. The method of claim 3 , wherein the pulse characteristic is a pulse height, the pulse height being a maximum voltage of the pulse.

5. The method of any one of claims 1 to 4, wherein the detector is an electron multiplying detector, the detector being configured to primarily detect single ion events.

6. The method of any one of claims 1 to 5, wherein generating the charge state identification comprises comparing the first pulse characteristic distribution with a reference pulse characteristic distribution.

7. 7. The method of claim 6, wherein the plurality of ions detected by the detector are produced from ionization of a sample, and the reference pulse characteristic distribution is identified based on known properties of the sample, the known properties of the sample comprising an isotope pattern or charge state distribution of compounds in the sample.

8. The method of any one of claims 1 to 7, wherein the generated identification comprises a probability of the charge state.

9. The method of any one of claims 1 to 8, further comprising generating a deconvoluted mass spectrum for the detected ions based on the identification of the charge states, wherein one axis of the mass spectrum is mass rather than mass / charge (m / z).

10. 10. The method of claim 1, wherein the plurality of ions are grouped into different groups using m / z domain, and the identification of charge state based on the first pulse characteristic distribution is performed for each group.

11. The grouping step includes: generating a mass spectrum based on the detected ions; identifying a first peak in the mass spectrum, the first peak having a mass-to-charge (m / z) value; grouping a plurality of ions within a mass-to-charge (m / z) range based on the m / z value of the first peak; The method of claim 10, comprising:

12. The grouping step comprises: selecting a first subset of the plurality of ions into a first intensity band and a second subset of the plurality of ions into a second intensity band; generating a first mass spectrum for the first intensity band; generating a second mass spectrum for the second intensity band; identifying a first peak in at least one of the first mass spectrum and the second mass spectrum, the first peak having a mass-to-charge (m / z) value; grouping a plurality of ions within a mass-to-charge (m / z) range based on the m / z value of the first peak; The method of claim 10, comprising:

13. The method further comprising generating a second pulse characteristic distribution for the plurality of ions within the m / z range; generating the identification determining that a plurality of ions forming the first pulse characteristic distribution have a first charge state; determining that a plurality of ions forming the second pulse characteristic distribution have a second charge state; The method of claim 10, comprising:

14. A method described in any one of claims 11 or 12, wherein the method further comprises determining, based on the identification of the charge state, that one or more isotopes correspond to the one or more ions forming the first peak.

15. The method of any one of claims 10 to 14, wherein generating the charge state identification comprises comparing the first pulse characteristic distribution with a reference pulse characteristic distribution.

16. 16. The method of claim 15, wherein the plurality of ions detected by the detector are produced from ionization of a sample, and the reference pulse characteristic distribution is identified based on known properties of the sample, the known properties of the sample comprising an isotope pattern or charge state distribution of compounds in the sample.

17. The method of any one of claims 10 to 16, wherein the generated identification comprises a probability of the charge state.

18. The method of any one of claims 1 to 17, wherein the method is carried out as part of a top-down protein analysis.

19. The method comprising: Identifying a second peak; and determining a consensus m / z distance based on at least the first peak and the second peak; identifying the first peak and the second peak as forming a feature; and further comprising 15. The method of claim 11, wherein generating the identification of the charge state is based on the consensus distance.

20. The method of claim 19, wherein identifying that the first peak and the second peak form the feature includes comparing a pulse characteristic distribution of the first peak with a pulse characteristic distribution of the second peak, and selecting a plurality of peaks having substantially the same pulse characteristic distribution.

21. The method described in claim 20, wherein comparing the pulse characteristic distribution of the first peak with the pulse characteristic distribution of the second peak is performed by calculating a Euclidean distance between the pulse characteristic distribution of the first peak and the pulse characteristic distribution of the second peak, and comparing the Euclidean distance with a predetermined threshold.

22. A method described in any one of claims 19 to 21, wherein the method further comprises identifying missing peaks corresponding to the features based on the consensus distance.

23. A method according to any one of claims 19 to 21, further comprising generating a deconvoluted mass spectrum for the detected ions based on the identification of the charge states, wherein one axis of the mass spectrum is mass rather than m / z.

24. A method according to any preceding claim, wherein the pulse characteristic is a maximum voltage of the pulse.

25. 1. A mass spectrometry system, comprising: a detector configured to generate a pulse for each ion detected by the detector; a processor; Memory that stores instructions Equipped with The instructions, when executed by the processor, generating a pulse for each ion of a plurality of ions impinging on the electron multiplying detector, each pulse having a pulse characteristic; generating a pulse characteristic distribution of the generated pulses; generating an identification of a charge state of one or more ions of the plurality of ions based on the pulse characteristic distribution; a mass spectrometry system configured to cause the mass spectrometry system to perform a set of operations including:

26. 26. The mass spectrometry system of claim 25, wherein the detector is an electron multiplying detector.

27. The mass analysis system according to any one of claims 25 to 26, further comprising an ion source device, a dissociation device, and a mass analyzer.

Images