Correction of mass spectral data

Abstract

A method of mass / ion mobility spectrometry comprising: a) separating and then mass / ion mobility analysing a sample to obtain a list of sample peaks in which a mass-to charge ratio or mobility is assoc

Description

CROSS-REFERENCE TO RELATED APPLICATION This application claims priority from and the benefit of United Kingdom patent application No. 2405275.5 filed on 12 April 2024. The entire contents of this application are incorporated herein by reference. FIELD OF THE INVENTION The present invention relates generally to mass spectrometers and in particular to methods and mass spectrometers that obtain mass spectral data as a function of retention time in a separator device, and then correct the mass to charge ratio measurements. BACKGROUND It is well known that the accuracy of mass to charge ratio measurements obtained by a mass spectrometer can change over time due to variations in factors that are external to the mass spectrometer. For example, the ambient temperature in which the mass spectrometer is located may vary with time, which may cause components of the spectrometer such as metalwork to expand or contract, and / or may affect the power supply and detection electronics. The accuracy of the mass to charge ratio measurements may also change over time due to variations in factors that are internal to the mass spectrometer. For example, fluctuations in the total ion current through the mass spectrometer may affect the accuracy of the mass to charge ratio measurements. For instance, high concentrations of ions in ion-optical components can cause space-charge effects that particularly affect mass measurements such as, for example, in mass analysers in which ions are trapped. It is desirable to measure how the accuracy of the mass to charge ratio measurements varies over time, or the factors that lead to them, and then adjust the mass spectral data or parameters of the mass spectrometer to compensate for this. Various techniques for measuring and correcting the error in the mass to charge ratio measurement are known. For example, it is known to add a reference standard to the sample that, when ionised, has a known mass to charge ratio, to measure the mass to charge ratio of the reference standard, and to compare the measured value to the known value so as to determine the accuracy of the mass measurement. However, adding reference standards to a sample may interfere with mass spectral data relating to other species present in the sample. This can affect the accuracy of the mass correction and can also interfere with the mass to measurements of the sample. Alternatively, rather than adding a reference standard to the sample to be mass analysed, it is known to periodically interrupt the mass analysis of the sample so as to mass analyse the reference standard. However, interrupting the mass analysis of the sample in this way can result in the loss of mass spectral data, which can become particularly problematic if only a relatively short analysis time is available, e.g. if the sample is eluting quickly from a chromatography device. Also, the mechanism used to switch the mass spectrometer between analysing the sample and the reference standard can add complexity, cost and maintenance to the mass spectrometer. The above-described known techniques that use reference standards also suffer from the disadvantage that it is necessary to ensure that there is a supply of a suitable reference standard. A known alternative to adding a reference standard to a sample to be mass analysed is to use a species that is known to be in the sample, or in the solvent carrying the sample, as a reference species. For example, it may be known that a contaminant is present in the solvent that produces a known mass to charge ratio when ionised. The mass spectrometer measures the mass to charge ratio of this species, and compares the measured value to the known value so as to determine the accuracy of the mass measurement. However, the presence of such species is generally undesirable and their concentration may be different in different batches of solvent or in different samples. Also, the abundance of such species in the measured mass spectra is uncontrolled, and their ionisation may be suppressed by analyte species. It is also known to correct the mass to charge ratios detected when mass analysing a chromatographically separated sample by matching the measured mass to charge ratios to mass to charge ratios in a library that have the same retention time. However, the matching process between the library and sample data can be complicated by differences in the retention times of a given species in the library and sample data. These differences in retention time may be caused by various factors, such as the library data and sample data being obtained by different mass spectrometers, variations in the conditions under which the sample is acquired, variations between samples, variations in the volume of sample loaded for mass analysis, variations in the solvent used to carry the sample, variations in the LC separation column, variations in the chromatographic conditions such as the gradient or column geometry, variations in the timing of the sample injection, and many other factors. These differences in retention times of a given species in the library and sample data may limit the quality of the correction, such as by limiting the retention time resolution of the correction, which can be particularly problematic when it is desired to correct for variations in the mass accuracy that occur over a short timescale. It is therefore desired to provide an alternative technique for correcting the mass to charge ratio measurements in mass spectral data. SUMMARY From a first aspect the present invention provides a method of mass spectrometry comprising: a) separating a sample in a separator device and then mass analysing the sample so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the separator device; b) providing a list of library peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device; c) dividing the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and dividing the list of library peaks, according to retention time, into multiple library peak sub-lists; d) peak-matching at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; e) determining an error in the mass to charge ratio of the first peak in the first sample peak sub-list based on the mass to charge ratios of the peaks in the first pair of matched peaks; and f) adjusting the mass to charge ratio of at least the first peak in the first sample peak sub-list using said error in mass to charge ratio so as to provide an adjusted first sample peak sub-list. As the peak-matching step matches a peak in the first sample peak sub-list with a peak in the first library peak sub-list, it is able to consider matching peaks that may have a difference in retention times, but whilst restricting the difference to the range of retention times that the sub-lists have. As such, embodiments of the present invention allow the matching of sample peaks to library peaks under a wide range of chromatographic and mass spectral conditions. It will be appreciated that each retention time in the list of sample peaks is the retention time in the separator device of the analyte species that gave rise to the ion species that was detected during mass analysis so as to obtain the sample peak. Similarly, each retention time in the list of library peaks represents the retention time in the, or a, separator device of an analyte species that gives rise to an ion species that forms the library peak. A matrix may be analysed in order to construct the library. The sample being analysed may contain the matrix, or a similar matrix (e.g. of similar biological origin). The peak-matching may be performed by matching peaks that have mass to charge ratios within a selected mass to charge ratio error value. The method is performed using a mass spectrometer having the separator device and a mass analyser for performing said mass analysis. It will be appreciated that the mass spectrometer may have one or more processors and electronic circuitry that are configured to perform the method steps described, e.g. automatically. The separator device may be a liquid or gas chromatography separation device. Alternatively, the separator device may be an ion separator device, such as an ion mobility separator. Step e) described above may comprise determining an error in the retention time of the first peak in the first sample peak sub-list based on a difference in the retention times of the peaks in the first pair of matched peaks; and optionally step f) may comprise adjusting the retention time of at least the first peak in the first sample peak sub-list using said error in retention time so as to provide the adjusted first sample peak sub-list. The method may comprise: g) peak-matching at least a second peak in a second of the sample peak sub-lists with at least a second respective peak in a second of the library peak sub-lists so as to obtain at least a pair of matched peaks; h) determining an error in the mass to charge ratio of the second peak in the second sample peak sub-list based on the mass to charge ratios of the peaks in the pair of matched peaks; and i) adjusting the mass to charge ratio of at least the second peak in the second sample peak sub-list using the error in mass to charge ratio. The method may comprise determining an error in the retention time of the second peak in the second sample peak sub-list based on a difference in the retention times of the peaks in the pair of matched peaks; and optionally adjusting the retention time of at least the second peak in the second sample peak sub-list using the error in retention time. During peak-matching step d), and / or during peak-matching step g), only sample peaks that have a peak intensity or area over a pre-selected threshold value may be taken into account during the peak matching; and / or only the N most abundant sample peaks may be taken into account during the peak matching, where N is an integer having a value that is <150, <100, <80, <60, <50, <40, <30, <20, <10 or <5. The peak matching step may also, or alternatively, be required to meet one or more other peak-matching criterion. For example, the method may identify sample peaks that are likely to have large errors in their mass accuracy and exclude these peaks from being matched to library peaks in the peak matching step. For instance, the method may identify sample peaks that have been detected when the ion detector or mass analyser is saturated or when there is interference from another species, and may exclude these peaks from the peak matching step. The method may only proceed from peak-matching step d) to step errordetermining step e), and / or may only proceeds from peak-matching step g) to step error determining step h), if a pre-selected minimum number of peaks P are matched in that peak-matching step. For example, it may be required to match >5, >10, >20, >30, >40 or >50 peaks in order for the method to proceed to step e) and / or to proceed to step h). When determining if peaks match in peak-matching step d), and / or in peakmatching step g), the method may not match the retention time of the sample peak to the retention time of the library peak. The step of dividing the list of sample peaks (step c) may comprise dividing the list of library peaks at a selected quantile of retention times in that list (e.g. at the 50% quantile), or at the median or mean retention time of the peaks in that list. Similarly, step c) may comprise dividing the list of sample peaks at the selected quantile of retention times in that list (e.g. at the 50% quantile), or at the median or mean retention time of the peaks in that list. Step c) may comprise arranging the peaks in the list of sample peaks according to retention time and then dividing this list such that the first sample peak sub-list comprises or consists of peaks in a first quantile of the list of sample peaks arranged according to retention time; and arranging the peaks in the list of library peaks according to retention time, and then dividing this list such that the first library peak sub-list comprises or consists of the same quantile of the list of library peaks arranged according to retention time that is the same as the first quantile. The step of dividing the list of sample peaks may provide a second sample peak sub-list that comprises or consists of peaks in a second, different quantile of the sample peak list arranged according to retention time. Similarly, step of dividing the list of library peaks may provide a second library peak sub-list that comprises or consists of peaks in said second quantile of the library peak list arranged according to retention time. Peak-matching step d) may comprise peak-matching a plurality of peaks in the first sample peak sub-list with a plurality of respective peaks in the first library peak sub-list so as to obtain a plurality of pairs of matched peaks; error-determining step e) may comprise determining an error in the mass to charge ratio of the sample peak in each of the plurality of pairs of peaks, based on the mass to charge ratios of the peaks in that pair of matched peaks, so as to obtain a plurality of mass to charge ratio error values; averaging these error values so as to obtain an average error value, and then performing said step of adjusting the mass to charge ratios of sample peaks in the first sample peak sub-list using the average error value. The method may comprise determining an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in retention time of the peaks in that pair of matched peaks, so as to obtain a plurality of retention time error values; averaging the retention time error values so as to obtain an average error value, and then performing said step of adjusting the retention times of sample peaks in the first sample peak sub-list using the average retention time error value. Peak-matching step d) may comprise peak-matching a plurality of peaks in the first sample peak sub-list with a plurality of respective peaks in the first library peak sub-list so as to obtain a plurality of pairs of matched peaks at different respective retention times; wherein error-determining step e) comprises determining an error in the mass to charge ratio of the sample peak in each of the plurality of pairs of peaks, based on the mass to charge ratios of the peaks in that pair of matched peaks, so as to obtain a relationship of mass to charge ratio error value as a function of retention time; and wherein step f) comprises adjusting the mass to charge ratios of sample peaks in the first sample peak sub-list using said relationship of mass to charge ratio error value as a function of retention time. The method may comprise determining an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in the retention times of the peaks in that pair of matched peaks, so as to obtain a relationship of retention time error value as a function of retention time, and adjusting the retention times of the sample peaks in the first sample peak sub-list using said relationship of retention time error value as a function of retention time. As described above, embodiments peak-match the second sample peak sub-list to the second library peak sub-list and adjust the mass to charge ratios (and optionally retention times) of the sample peaks in the second sample peak sub-list. In these embodiments, the method of adjusting the mass to charge ratios (and optionally retentions times) of the sample peaks in the second sample peak sub-list may be performed in a corresponding manner to that described above in relation to adjusting the mass to charge ratios (and optionally retentions times) of the sample peaks in the first sample peak sublist. Accordingly, peak-matching step g) may comprise peak-matching a plurality of peaks in the second sample peak sub-list with a plurality of respective peaks in the second library peak sub-list so as to obtain a plurality of pairs of matched peaks; and errordetermining step h) may comprise: determining an error in the mass to charge ratio of the sample peak in each of the plurality of pairs of peaks, based on the mass to charge ratios of the peaks in that pair of matched peaks, so as to obtain a plurality of mass to charge ratio error values; averaging these error values so as to obtain an average error value, and then performing said step i) of adjusting the mass to charge ratios of sample peaks in the second sample peak sub-list using the average error value. Additionally, or alternatively, the method may determine an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in retention time of the peaks in that pair of matched peaks, so as to obtain a plurality of retention time error values; average the retention time error values so as to obtain an average error value, and then perform said step of adjusting the retention times of sample peaks in the second sample peak sub-list using the average retention time error value. Peak-matching step g) may comprise peak-matching a plurality of peaks in the second sample peak sub-list with a plurality of respective peaks in the second library peak sub-list so as to obtain a plurality of pairs of matched peaks at different respective retention times; and error-determining step h) may comprise determining an error in the mass to charge ratio of the sample peak in each of the plurality of pairs of peaks, based on the mass to charge ratios of the peaks in that pair of matched peaks, so as to obtain a relationship of mass to charge ratio error value as a function of retention time; and wherein step i) comprises adjusting the mass to charge ratios of sample peaks in the second sample peak sub-list using said relationship of mass to charge ratio error value as a function of retention time. The method may comprise determining an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in the retention times of the peaks in that pair of matched peaks, so as to obtain a relationship of retention time error value as a function of retention time, and adjusting the retention times of the sample peaks in the second sample peak sub-list using said relationship of retention time error value as a function of retention time. After the mass to charge ratios (and optionally retention times) of the peaks in the sample peak sub-lists have been adjusted, the method may then divide each adjusted sample peak sub-list, according to retention time, into multiple further sub-lists with the aim of seeking to match peaks in those sub-lists with library peaks and then adjust the mass to charge ratios (and optionally retentions times) of peaks in those sub-lists. Accordingly, the method may comprise: dividing the adjusted first sample peak sublist, according to retention time, into multiple further sample peak sub-lists, and dividing the first library peak sub-list, according to retention time, into multiple further library peak sublists; peak-matching at least a first peak in a first of said further sample peak sub-lists with at least a first respective peak in a first of said further library peak sub-lists so as to obtain at least a first pair of matched peaks; determining an error in the mass to charge ratio of the first peak in the first further sample peak sub-list based on the mass to charge ratios of the peaks in the first pair of matched peaks; and adjusting the mass to charge ratio of at least the first peak in the first further sample peak sub-list using said error in mass to charge ratio. The method may comprise determining an error in the retention time of the first peak in the first further sample peak sub-list based on a difference in the retention times of the peaks in the first pair of matched peaks, and adjusting the retention time of at least the first peak in the first further sample peak sub-list using said error in retention time. Of course, it is not necessarily only the adjusted first sample peak sub-list that is subjected to these steps. For example, corresponding steps may be performed on the second sample peak sub-list. Also, the sub-lists that are divided from the first and / or second sub-lists may be subjected to corresponding steps. Accordingly, more generally, the method may further comprise performing a cycle that comprises: j) dividing one of the sample peak sub-lists, according to retention time, into multiple further sample peak sub-lists, and dividing a corresponding library peak sublist, according to retention time, into multiple further library peak sub-lists; k) peak-matching at least a first peak in a first of said further sample peak sub-lists with at least a first respective peak in a first of said further library peak sub-lists so as to obtain at least a first pair of matched peaks; I) determining an error in the mass to charge ratio of the first peak in the first further sample peak sub-list based on the mass to charge ratios of the peaks in the first pair of matched peaks; and m) adjusting the mass to charge ratio of at least the first peak in the first further sample peak sub-list using said error in mass to charge ratio; wherein the cycle of steps j) to I) is repeatedly performed, and wherein each time the cycle is performed the sample peak sub-list that is divided in step j) is the adjusted sample peak sub-list produced in step m) of the preceding cycle. The likelihood of a sample peak being matched to multiple library peaks, or vice versa, decreases with each further cycle, as the number of peaks being compared in the peak-matching process is reduced. This is because each sub-list has a smaller retention time range than its parent list and therefore contains fewer peaks. It will therefore be appreciated that the iterative nature of the method causes the peak matching to improve in each further cycle. Step I) may also comprise determining an error in the retention time of the first peak in the first further sample peak sub-list based on a difference in the retention times of the peaks in the first pair of matched peaks, and step m) may comprise adjusting the retention time of at least the first peak in the first further sample peak sub-list using said error in retention time. As described above, the first time said cycle is performed, step j) may be performed on the adjusted first sample peak sub-list from step f). When step j) is performed it will produce multiple further sample peak sub-lists. Although steps k) to m) have been described as being performed on a first of these further sample peak sub-lists, corresponding steps may also be performed on a second of these further sample peak sub-lists that has been created in step j) of the cycle. In other words, each time the cycle is performed it will produce multiple further sub-lists and each of these further sub-lists that has been produced may be subjected to the remaining steps in said cycle so as to produce an adjusted sample peak sub-list. Each of those adjusted sample peak sub-lists may then be subjected to said cycle. For example, step j) of the cycle may be performed on the adjusted second sample peak sub-list from step i) (of claim 3). This will produce multiple further sample peak sublists, each of which may be subjected to steps k) to m) of the cycle so as to produce an adjusted sample peak sub-lists. Each of those adjusted sample peak sub-lists may then be subjected to said cycle. Peak-matching step k) may only match peaks that differ in mass to charge ratio by less than or equal to a mass to charge ratio tolerance value; wherein the mass to charge ratio tolerance value is calculated by: estimating the maximum rate at which the mass accuracy of the mass spectrometer varies with retention time, determining the range of retention times in the first further sample peak sub-list and multiplying this value by the maximum rate at which the mass accuracy varies with retention time. The likelihood of a sample peak being matched to multiple library peaks, or vice versa, decreases with each further cycle, as the number of peaks being compared in the peak-matching process is reduced. This is because each sub-list has a smaller retention time range than its parent list and therefore the mass to charge ratio tolerance value decreases in each cycle. Optionally, any given time that the cycle is being performed the method only proceeds from peak-matching step k) to error-determining step I) if a pre-selected minimum number of peaks P in a sample peak sub-list are matched to peaks in a library peak sub-list during peak-matching step k). For example, it may be required to match >5, >10, >20, >30, >40 or >50 peaks in order for the method to proceed to step I). Optionally, if said pre-selected minimum number of peaks P in the sample peak sub-list are not matched to peaks in a library peak sub-list then the method does not proceed from peak-matching step k) to error-determining step I), and instead the retention time range for that sample peak sub-list is assigned a final mass to charge ratio error and / or a final retention time error corresponding to the mass to charge ratio error and / or retention time error, respectively, that were determined for the sample peak sub-list that it was divided from. Optionally, any given time that the cycle is being performed the method only proceeds from dividing step j) to peak-matching step k) if it is determined that greater than or equal to a threshold number of peaks are present in the first further sample peak sublist. For example, as described above, any given time that the cycle is being performed the method may only proceed from peak-matching step k) to error-determining step I) if a preselected minimum number of peaks P in a sample peak sub-list are matched to peaks in a library peak sub-list, and the method may require that at least twice this number of peaks are present in the first further sample peak sub-list in order to proceed from dividing step j) to peak-matching step k). If it is determined that fewer than said threshold number of peaks are present in the first further sample peak sub-list then the method may not proceed from dividing step j) to peak matching step k), and instead the retention time range for that sample peak sub-list is assigned a final mass to charge ratio error and / or final retention time error corresponding to the mass to charge ratio error and retention time error, respectively, that were determined for the sample peak sub-list that it was divided from. Accordingly, said cycle may be performed in full on every sample peak sub-list that is generated unless it is: a) not possible to match said pre-selected minimum number of peaks P in the sample peak sub-list to peaks in a corresponding library peak sub-list, in which case the retention time range for that sample peak sub-list is assigned the mass to charge ratio error and / or retention time error that was determined for the sample peak sublist that it was divided from; and / or b) determined that dividing the sample peak sub-list at step j) produces said first further sample peak sub-list having fewer than a threshold number of peaks present, in which case the retention time range for said first further sample peak sub-list is assigned the mass to charge ratio error and / or retention time error that was determined for the sample peak sub-list that it was divided from. The cycle may be repeatedly performed until a final mass to charge ratio error, and / or a final retention time error, is assigned to the retention time range of all of the sample peak sub-lists that are produced. The method may comprise correcting the mass to charge ratios and / or retention times of sample peaks in the list of sample peaks using the final mass to charge ratio errors and / or the final retention time errors associated with retention times that correspond to the retention times of the sample peaks in the list of sample peaks; or using the final mass to charge ratio errors and retention time errors of the sample peak sub-lists to obtain calibration data that is representative of a calibration curve or line of mass to charge ratio error as a function of retention time, and using said calibration data to correct the mass to charge ratios and retention times of sample peaks in the list of sample peaks. It will be appreciated that according to the embodiments described herein, the original list of sample peaks may be subdivided into a greater number of sample peak sublists over a first range of retention time values than over a second, different but equally sized range of retention time values, if more sample peaks in the first range are able to be matched to library peaks than for the second range. The method described above is performed using a mass spectrometer. It will be appreciated that the mass spectrometer may have one or more processors and electronic circuitry that are configured to perform the method steps described herein, e.g. automatically. Although the method has been described as including the step of separating a sample in a separator device and then mass analysing the sample so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is correlated with a retention time in the separator device, the invention extends to performing the method when the list of sample peaks is simply provided. Accordingly, the present invention provides a computer implemented method of correcting the mass to charge ratios and retention times of sample peaks in mass spectral data, the method comprising: providing a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device; providing a list of library peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device; dividing the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and dividing the list of library peaks, according to retention time, into multiple library peak sub-lists; peak-matching at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; determining an error in the mass to charge ratio of the first peak in the first sample peak sub-list based the mass to charge ratios of the peaks in the first pair of matched peaks; and adjusting the mass to charge ratio of at least the first peak in the first sample peak sub-list using said error in mass to charge ratio so as to provide an adjusted first sample peak sub-list. This method may comprise any of the features described above in relation to the first aspect of the present invention, except without having to include the step of separating a sample in a separator device and then mass analysing the sample so as to obtain the list of sample peaks. The present invention also provides a computer-readable medium comprising instructions which, when executed on a computer, cause the computer to carry out the method described above. The present invention also provides a mass spectrometer arranged and configured to perform the method described herein. Accordingly, the present invention provides a mass spectrometer comprising: a separator device for separating a sample; a mass analyser for mass analysing the separated sample so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the separator device; control circuitry configured to: access a list of library peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device; divide the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and divide the list of library peaks, according to retention time, into multiple library peak sub-lists; peak-match at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; determine an error in the mass to charge ratio of the first peak in the first sample peak sublist based on the mass to charge ratios of the peaks in the first pair of matched peaks; adjust the mass to charge ratio of at least the first peak in the first sample peak sub-list using said error in mass to charge ratio so as to provide an adjusted first sample peak sublist. The mass spectrometer may include a memory on which the list of library peaks is stored. The mass spectrometer may comprise one or more processors and electronic circuitry configured to perform any of the method steps described herein, such as those described in relation to the first aspect of the invention. Although the mass spectrometer has been described as having a separator device and mass analyser for obtaining a list of sample peaks, the invention extends more generally to an apparatus for correcting mass spectral data, to which the list of sample peaks is simply provided. Accordingly, the present invention also provides an apparatus for correcting the mass to charge ratio of mass spectral data, comprising control circuitry configured to: access a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device; access a list of library peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device; divide the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and divide the list of library peaks, according to retention time, into multiple library peak sub-lists; peak-match at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; determine an error in the mass to charge ratio of the first peak in the first sample peak sub-list based on the mass to charge ratios of the peaks in the first pair of matched peaks; and adjust the mass to charge ratio of at least the first peak in the first sample peak sub-list using said error in mass to charge ratio so as to provide an adjusted first sample peak sub-list. The apparatus may have any of the features described above in relation to the mass spectrometer, except without having to include the separator device and mass analyser for obtaining a list of sample peaks. The apparatus may include a memory on which the lists of sample peaks and library peaks are stored. The methods, spectrometer and apparatus described herein may identify ion species, or analyte molecules from which the ion species are derived, using the corrected mass to charge ratios and retention times. The first aspect of the invention has been described as mass analysing the sample so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the separator device, and then errors in the mass to charge ratios are corrected. However, rather than mass analysing the sample, it is contemplated that the ion mobility of the ions may be analysed and the method may correct the error in ion mobility that is detected. Accordingly, from a second aspect the present invention provides a method of ion mobility spectrometry comprising: a) separating a sample in a separator device and then ion mobility analysing the sample so as to obtain a list of sample peaks in which a mobility for each peak is associated with a retention time in the separator device; b) providing a list of library peaks in which a mobility for each peak is associated with a retention time in a separator device; c) dividing the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and dividing the list of library peaks, according to retention time, into multiple library peak sub-lists; d) peak-matching at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; e) determining an error in mobility of the first peak in the first sample peak sub-list based on the mobilities of the peaks in the first pair of matched peaks; and f) adjusting the mobility of at least the first peak in the first sample peak sub-list using said error in mobility so as to provide an adjusted first sample peak sub-list. The second aspect of the invention may have any of the features described in relation to the first aspect of the invention, except wherein mass to charge ratio is replaced by mobility. For example, the method is performed using a mobility spectrometer having the separator device and a mobility analyser for performing said mobility analysis. It will be appreciated that the spectrometer may have one or more processors and electronic circuitry that are configured to perform the method steps described, e.g. automatically. The peak-matching may be performed by matching peaks that have mobilities within a selected mobility error value. The separator device may be a liquid or gas chromatography separation device, or less preferably an ion separator device that separates ions by mass to charge ratio or mobility. Step e) described above may comprise determining an error in the retention time of the first peak in the first sample peak sub-list based on a difference in the retention times of the peaks in the first pair of matched peaks; and optionally step f) may comprise adjusting the retention time of at least the first peak in the first sample peak sub-list using said error in retention time so as to provide the adjusted first sample peak sub-list. The method may comprise: g) peak-matching at least a second peak in a second of the sample peak sub-lists with at least a second respective peak in a second of the library peak sub-lists so as to obtain at least a pair of matched peaks; h) determining an error in the mobility of the second peak in the second sample peak sub-list based on the mobilities of the peaks in the pair of matched peaks; and i) adjusting the mobility of at least the second peak in the second sample peak sub-list using the error in mobility. The method may comprise determining an error in the retention time of the second peak in the second sample peak sub-list based on a difference in the retention times of the peaks in the pair of matched peaks; and optionally adjusting the retention time of at least the second peak in the second sample peak sub-list using the error in retention time. For each occurrence of the peak-matching step (i.e. when peak-matching each of the first sample peak sub-list and the second sample-peak sub-list), the peak-matching step may only consider peaks as being matched if they differ in mobility by less than or equal to a selected maximum error in mobility. The selected maximum error in mobility may be the maximum error in a mobility measurement that the spectrometer can make. This value may be estimated or experimentally determined. During peak-matching step d), and / or during peak-matching step g), only sample peaks that have a peak intensity or area over a pre-selected threshold value may be taken into account during the peak matching; and / or only the N most abundant sample peaks may be taken into account during the peak matching, where N is an integer having a value that is <150, <100, <80, <60, <50, <40, <30, <20, <10 or <5. The peak matching step may also, or alternatively, be required to meet one or more other peak-matching criterion. For example, the method may identify sample peaks that are likely to have large errors in their mobility accuracy and exclude these peaks from being matched to library peaks in the peak matching step. For instance, the method may identify sample peaks that have been detected when the ion detector or mobility analyser is saturated or when there is interference from another species, and may exclude these peaks from the peak matching step. The method may only proceed from peak-matching step d) to step errordetermining step e), and / or only proceeds from peak-matching step g) to step errordetermining step h), if a pre-selected minimum number of peaks P are matched in that peak-matching step. For example, it may be required to match >5, >10, >20, >30, >40 or >50 peaks in order for the method to proceed to step e) and / or to proceed to step h). When determining if peaks match in peak-matching step d), and / or in peakmatching step g), the method may not match the retention time of the sample peak to the retention time of the library peak. The step of dividing the list of sample peaks (step c) may comprise dividing the list of library peaks at a selected quantile of retention times in that list (e.g. at the 50% quantile), or at the median or mean retention time of the peaks in that list. Similarly, step c) may comprise dividing the list of sample peaks at the selected quantile of retention times in that list (e.g. at the 50% quantile), or at the median or mean retention time of the peaks in that list. Step c) may comprise arranging the peaks in the list of sample peaks according to retention time and then dividing this list such that the first sample peak sub-list comprises or consists of peaks in a first quantile of the list of sample peaks arranged according to retention time; and arranging the peaks in the list of library peaks according to retention time, and then dividing this list such that the first library peak sub-list comprises or consists of the same quantile of the list of library peaks arranged according to retention time that is the same as the first quantile. The step of dividing the list of sample peaks may provide a second sample peak sub-list that comprises or consists of peaks in a second, different quantile of the sample peak list arranged according to retention time. Similarly, step of dividing the list of library peaks may provide a second library peak sub-list that comprises or consists of peaks in said second quantile of the library peak list arranged according to retention time. Peak-matching step d) may comprises peak-matching a plurality of peaks in the first sample peak sub-list with a plurality of respective peaks in the first library peak sub-list so as to obtain a plurality of pairs of matched peaks; and error-determining step e) comprises determining an error in the mobility of the sample peak in each of the plurality of pairs of peaks, based on the mobilities of the peaks in that pair of matched peaks, so as to obtain a plurality of mobility error values; averaging these error values so as to obtain an average error value, and then performing said step of adjusting the mobilities of sample peaks in the first sample peak sub-list using the average error value. The method may comprise determining an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in retention time of the peaks in that pair of matched peaks, so as to obtain a plurality of retention time error values; averaging the retention time error values so as to obtain an average error value, and then performing said step of adjusting the retention times of sample peaks in the first sample peak sub-list using the average retention time error value. Peak-matching step d) may comprise peak-matching a plurality of peaks in the first sample peak sub-list with a plurality of respective peaks in the first library peak sub-list so as to obtain a plurality of pairs of matched peaks at different respective retention times. Error-determining step e) may comprise determining an error in the mobility of the sample peak in each of the plurality of pairs of peaks, based on the mobilities of the peaks in that pair of matched peaks, so as to obtain a relationship of mobility error value as a function of retention time; and step f) may comprise adjusting the mobilities of sample peaks in the first sample peak sub-list using said relationship of mobility error value as a function of retention time. The method may comprise determining an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in the retention times of the peaks in that pair of matched peaks, so as to obtain a relationship of retention time error value as a function of retention time, and adjusting the retention times of the sample peaks in the first sample peak sub-list using said relationship of retention time error value as a function of retention time. As described above, embodiments peak-match the second sample peak sub-list to the second library peak sub-list and adjust the mobilities (and optionally retention times) of the sample peaks in the second sample peak sub-list. In these embodiments, the method of adjusting the mobilities (and optionally retentions times) of the sample peaks in the second sample peak sub-list may be performed in a corresponding manner to that described above in relation to adjusting the mobilities (and optionally retentions times) of the sample peaks in the first sample peak sub-list. Accordingly, peak-matching step g) may comprise peak-matching a plurality of peaks in the second sample peak sub-list with a plurality of respective peaks in the second library peak sub-list so as to obtain a plurality of pairs of matched peaks; and errordetermining step h) may comprise: determining an error in the mobility of the sample peak in each of the plurality of pairs of peaks, based on the mobilities of the peaks in that pair of matched peaks, so as to obtain a plurality of mobility error values; averaging these error values so as to obtain an average error value, and then performing said step i) of adjusting the mobilities of sample peaks in the second sample peak sub-list using the average error value. Additionally, or alternatively, the method may determine an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in retention time of the peaks in that pair of matched peaks, so as to obtain a plurality of retention time error values; average the retention time error values so as to obtain an average error value, and then perform said step of adjusting the retention times of sample peaks in the second sample peak sub-list using the average retention time error value. Peak-matching step g) may comprise peak-matching a plurality of peaks in the second sample peak sub-list with a plurality of respective peaks in the second library peak sub-list so as to obtain a plurality of pairs of matched peaks at different respective retention times; and error-determining step h) may comprise determining an error in the mobility of the sample peak in each of the plurality of pairs of peaks, based on the mobilities of the peaks in that pair of matched peaks, so as to obtain a relationship of mobility error value as a function of retention time; and wherein step i) comprises adjusting the mobilities of sample peaks in the second sample peak sub-list using said relationship of mobility error value as a function of retention time. The method may comprise determining an error in the retention time of the sample peak in each of the plurality of pairs of peaks, based on a difference in the retention times of the peaks in that pair of matched peaks, so as to obtain a relationship of retention time error value as a function of retention time, and adjusting the retention times of the sample peaks in the second sample peak sub-list using said relationship of retention time error value as a function of retention time. After the mobilities (and optionally retention times) of the peaks in the sample peak sub-lists have been adjusted, the method may then divide each adjusted sample peak sublist, according to retention time, into multiple further sub-lists with the aim of seeking to match peaks in those sub-lists with library peaks and then adjust the mobilities (and optionally retentions times) of peaks in those sub-lists. Accordingly, the method may comprise: dividing the adjusted first sample peak sublist, according to retention time, into multiple further sample peak sub-lists, and dividing the first library peak sub-list, according to retention time, into multiple further library peak sublists; peak-matching at least a first peak in a first of said further sample peak sub-lists with at least a first respective peak in a first of said further library peak sub-lists so as to obtain at least a first pair of matched peaks; determining an error in the mobility of the first peak in the first further sample peak sub-list based on the mobilities of the peaks in the first pair of matched peaks; and adjusting the mobility of at least the first peak in the first further sample peak sub-list using said error in mobility. The method may comprise determining an error in the retention time of the first peak in the first further sample peak sub-list based on a difference in the retention times of the peaks in the first pair of matched peaks, and adjusting the retention time of at least the first peak in the first further sample peak sub-list using said error in retention time. Of course, it is not necessarily only the adjusted first sample peak sub-list that is subjected to these steps. For example, corresponding steps may be performed on the second sample peak sub-list. Also, the sub-lists that are divided from the first and / or second sub-lists may be subjected to corresponding steps. Accordingly, more generally, the method further comprises performing a cycle that comprises: j) dividing one of the sample peak sub-lists, according to retention time, into multiple further sample peak sub-lists, and dividing a corresponding library peak sub-list, according to retention time, into multiple further library peak sub-lists; k) peak-matching at least a first peak in a first of said further sample peak sub-lists with at least a first respective peak in a first of said further library peak sub-lists so as to obtain at least a first pair of matched peaks; I) determining an error in the mobility of the first peak in the first further sample peak sub-list based on the mobilities of the peaks in the first pair of matched peaks; and m) adjusting the mobility of at least the first peak in the first further sample peak sub-list using said error in mobility; wherein the cycle of steps j) to I) is repeatedly performed, and wherein each time the cycle is performed the sample peak sub-list that is divided in step j) is the adjusted sample peak sub-list produced in step m) of the preceding cycle. The likelihood of a sample peak being matched to multiple library peaks, or vice versa, decreases with each further cycle, as the number of peaks being compared in the peak-matching process is reduced. This is because each sub-list has a smaller retention time range than its parent list and therefore contains fewer peaks. It will therefore be appreciated that the iterative nature of the method causes the peak matching to improve in each further cycle. Step I) may also comprise determining an error in the retention time of the first peak in the first further sample peak sub-list based on a difference in the retention times of the peaks in the first pair of matched peaks, and step m) may comprise adjusting the retention time of at least the first peak in the first further sample peak sub-list using said error in retention time. As described above, the first time said cycle is performed, step j) may be performed on the adjusted first sample peak sub-list from step f). When step j) is performed it will produce multiple further sample peak sub-lists. Although steps k) to m) have been described as being performed on a first of these further sample peak sub-lists, corresponding steps may also be performed on a second of these further sample peak sub-lists that has been created in step j) of the cycle. In other words, each time the cycle is performed it will produce multiple further sub-lists and each of these further sub-lists that has been produced may be subjected to the remaining steps in said cycle so as to produce an adjusted sample peak sub-list. Each of those adjusted sample peak sub-lists may then be subjected to said cycle. For example, step j) of the cycle may be performed on the adjusted second sample peak sub-list from step i). This will produce multiple further sample peak sub-lists, each of which may be subjected to steps k) to m) of the cycle so as to produce an adjusted sample peak sub-lists. Each of those adjusted sample peak sub-lists may then be subjected to said cycle. Peak-matching step k) may only match peaks that differ in mobility by less than or equal to a selected amount. For example, peak-matching step k) may only match peaks that differ in mobility by less than or equal to a mobility tolerance value; wherein the mobility tolerance value is calculated by: estimating the maximum rate at which the mobility accuracy of the mobility analyser varies with retention time, determining the range of retention times in the first further sample peak sub-list and multiplying this value by the maximum rate at which the mobility accuracy varies with retention time. The likelihood of a sample peak being matched to multiple library peaks, or vice versa, decreases with each further cycle, as the number of peaks being compared in the peak-matching process is reduced. This is because each sub-list has a smaller retention time range than its parent list and therefore the mobility tolerance value decreases in each cycle. Any given time that the cycle is being performed the method may only proceed from peak-matching step k) to error-determining step I) if a pre-selected minimum number of peaks Pina sample peak sub-list are matched to peaks in a library peak sub-list during peak-matching step k). For example, it may be required to match >5, >10, >20, >30, >40 or >50 peaks in order for the method to proceed to step I). If said pre-selected minimum number of peaks P in the sample peak sub-list are not matched to peaks in a library peak sub-list then the method does not proceed from peakmatching step k) to error-determining step I), and instead the retention time range for that sample peak sub-list is assigned a final mobility error and / or a final retention time error corresponding to the mobility error and / or retention time error, respectively, that were determined for the sample peak sub-list that it was divided from. Any given time that the cycle is being performed the method may only proceed from dividing step j) to peak-matching step k) if it is determined that greater than or equal to a threshold number of peaks are present in the first further sample peak sub-list. For example, as described above, any given time that the cycle is being performed the method may only proceed from peak-matching step k) to error-determining step I) if a pre-selected minimum number of peaks P in a sample peak sub-list are matched to peaks in a library peak sub-list, and the method may require that at least twice this number of peaks are present in the first further sample peak sub-list in order to proceed from dividing step j) to peak-matching step k). If it is determined that fewer than said threshold number of peaks are present in the first further sample peak sub-list then the method may not proceed from dividing step j) to peak matching step k), and instead the retention time range for that sample peak sub-list may be assigned a final mobility error and optionally final retention time error corresponding to the mobility error and / or retention time error, respectively, that were determined for the sample peak sub-list that it was divided from. Accordingly, said cycle may be performed in full on every sample peak sub-list that is generated unless it is: a) not possible to match said pre-selected minimum number of peaks P in the sample peak sub-list to peaks in a corresponding library peak sub-list, in which case the retention time range for that sample peak sub-list is assigned the mobility error and / or retention time error that was determined for the sample peak sub-list that it was divided from; and / or b) determined that dividing the sample peak sub-list at step j) produces said first further sample peak sub-list having fewer than a threshold number of peaks present, in which case the retention time range for said first further sample peak sub-list is assigned the mobility error and / or retention time error that was determined for the sample peak sub-list that it was divided from. The cycle may be repeatedly performed until a final mobility error, and / or a final retention time error, is assigned to the retention time range of all of the sample peak sublists that are produced. The method may comprise correcting the mobilities and / or retention times of sample peaks in the list of sample peaks using the final mobility errors and / or the final retention time errors associated with retention times that correspond to the retention times of the sample peaks in the list of sample peaks; or the method may use the final mobility errors and retention time errors of the sample peak sub-lists to obtain calibration data that is representative of a calibration curve or line of mobility error as a function of retention time, and then use said calibration data to correct the mobilities and retention times of sample peaks in the list of sample peaks. It will be appreciated that according to the embodiments described herein, the original list of sample peaks may be subdivided into a greater number of sample peak sublists over a first range of retention time values than over a second, different but equally sized range of retention time values, if more sample peaks in the first range are able to be matched to library peaks than for the second range. The method described above is performed using a mobility spectrometer. It will be appreciated that the spectrometer may have one or more processors and electronic circuitry that are configured to perform the method steps described herein, e.g. automatically. Although the method has been described as including the step of separating a sample in a separator device and then mobility analysing the sample so as to obtain a list of sample peaks in which a mobility for each peak is correlated with a retention time in the separator device, the invention extends to performing the method when the list of sample peaks is simply provided. Accordingly, the present invention provides a computer implemented method of correcting the mobilities of sample peaks in spectral data, the method comprising: providing a list of sample peaks in which a mobility for each peak is associated with a retention time in a separator device; providing a list of library peaks in which a mobility for each peak is associated with a retention time in a separator device; dividing the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and dividing the list of library peaks, according to retention time, into multiple library peak sub-lists; peak-matching at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; determining an error in the mobility of the first peak in the first sample peak sub-list based the mobilities of the peaks in the first pair of matched peaks; and adjusting the mobility of at least the first peak in the first sample peak sub-list using said error in mobility so as to provide an adjusted first sample peak sub-list. This method may comprise any of the features described above in relation to the second aspect of the present invention, except without having to include the step of separating a sample in a separator device and then mobility analysing the sample so as to obtain the list of sample peaks. The present invention also provides a computer-readable medium comprising instructions which, when executed on a computer, cause the computer to carry out the method described above. The present invention also provides a spectrometer arranged and configured to perform the method described above. Accordingly, the present invention provides a mobility spectrometer comprising: a separator device for separating a sample; a mobility analyser for analysing the ion mobilities of the separated sample so as to obtain a list of sample peaks in which a mobility for each peak is associated with a retention time in the separator device; and control circuitry configured to: access a list of library peaks in which a mobility for each peak is associated with a retention time in a separator device; divide the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and divide the list of library peaks, according to retention time, into multiple library peak sub-lists; peak-match at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; determine an error in the mobility of the first peak in the first sample peak sub-list based on the mobilities of the peaks in the first pair of matched peaks; and adjust the mobility of at least the first peak in the first sample peak sub-list using said error in mobility so as to provide an adjusted first sample peak sub-list. The spectrometer may include a memory on which the list of library peaks is stored. The spectrometer may comprise one or more processors and electronic circuitry configured to perform any of the method steps described above, such as those described in relation to the second aspect of the invention. Although the spectrometer has been described as having a separator device and mobility analyser for obtaining a list of sample peaks, the invention extends more generally to an apparatus for correcting spectral data, to which the list of sample peaks is simply provided. Accordingly, the present invention also provides an apparatus for correcting the mobility of spectral data, comprising control circuitry configured to: access a list of sample peaks in which a mobility for each peak is associated with a retention time in a separator device; access a list of library peaks in which a mobility for each peak is associated with a retention time in a separator device; divide the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and divide the list of library peaks, according to retention time, into multiple library peak sub-lists; peak-match at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks; determine an error in the mobility of the first peak in the first sample peak sub-list based on the mobilities of the peaks in the first pair of matched peaks; and adjust the mobility of at least the first peak in the first sample peak sub-list using said error in mobility so as to provide an adjusted first sample peak sub-list. The apparatus may have any of the features described above in relation to the mobility spectrometer, except without having to include the separator device and mobility analyser for obtaining a list of sample peaks. The apparatus may include a memory on which the lists of sample peaks and library peaks are stored. The methods, spectrometer and apparatus described above may identify ion species, or analyte molecules from which the ion species are derived, using the corrected mobilities and retention times. BRIEF DESCRIPTION OF THE DRAWINGS Various embodiments will now be described, by way of example only, and with reference to the accompanying drawings in which: Fig. 1 shows an embodiment of a mass spectrometer according to the present invention; Fig. 2 shows a flow chart illustrating a method according to an embodiment of the present invention; Fig. 3A shows mass spectral data corresponding to a list of library peaks in the graph on the left, and mass spectral data corresponding to a list of sample peaks in the graph on the right; Fig. 3B shows graphs of the mass spectral shown in Fig. 3A, except wherein the sample peaks have been corrected, and each of the lists of library and sample peaks has been divided into sub-lists; Fig. 3C shows graphs of the mass spectral shown in Fig. 3B, except wherein the sample peaks have been corrected, and each of the sub-lists of library and sample peaks has been divided into further sub-lists; Fig. 3D shows graphs of the mass spectral shown in Fig. 3C, except wherein the sample peaks have been corrected, and each of the further sub-lists of library and sample peaks has been divided into yet further sub-lists; Fig. 4A shows an example of how the mass to charge ratio error varies as a function of retention time for the mass analysis of E. coli bacteria; and Fig. 4B shows an example of how the retention time error varies as a function of retention time for the mass analysis of the E. coli bacteria. DETAILED DESCRIPTION Fig. 1 shows an embodiment of a mass spectrometer according to the present invention. The spectrometer comprises a sample separator such as a liquid chromatography device 2, an ion source 4, a first mass filter 6, a fragmentation or reaction device 8, and a mass analyser 10. The mass analyser may be a Time of Flight mass analyser, although it will be appreciated that other types of mass analyser may be used, such as an electrostatic ion trap mass analysers (e.g. an Orbitrap mass analyser). It will also be appreciated that additional ion-optical devices may be provided and / or that the mass filter 6 and / or fragmentation or reaction device 8 may be omitted or at least deactivated so as to simply operate as ion guides, at least in some modes of operation. As described above, a sample is injected into the sample separator 2, e.g. liquid chromatography device, which has different retention times for different analytes and so causes the different analytes to elute over different respective time periods. The eluting analytes are ionised by the ion source 4, such as an electrospray ion source for example, so as to produce precursor ions. It will be appreciated that other types of ion source may be used instead. The precursor ions pass to the first mass filter 6, which may either be set to transmit a restricted range of mass to charge ratios at any given time, e.g. corresponding to a single precursor ion species, or may be operated as an ion guide so as to transmit a wide range of mass to charge ratios. The precursor ions that are transmitted by the first mass filter 6 may pass to the fragmentation or reaction device 8 and be fragmented, or reacted (e.g. with reactant ions or molecules), so as to form fragment or product ions. These fragment or product ions are then transmitted to the mass analyser 10 for mass analysis. It is contemplated that the ions need not be fragmented or reacted, at least in one mode, and that the precursor ions may be mass analysed. It is known to use a mass spectrometer having a separator device and mass analyser to separate and mass analyse species in a sample, thereby obtaining mass to charge ratio values for species present in the sample. However, the accuracy of the mass to charge ratio measurements may vary depending on various factors. As such, it may be desired to correct the mass to charge ratios detected by the mass spectrometer. Fig. 2 shows a flow chart illustrating a method according to an embodiment of the present invention. As represented by step 20 in Fig. 2, a mass spectrometer having a separator device and mass analyser, such as that described in relation to Fig. 1, is used to separate and mass analyse an analytical sample so as to obtain mass spectral data for species present in the sample. The mass spectral data is peak-detected (and optionally deisotoped and / or de-charged) in order to determine the mass to charge ratio (or molecular weight) of each mass peak, and this is then associated with the respective retention time in the separator device of the species that gave rise to the peak. The retention time for a species may be determined based on the time that its mass peak was detected relative to the time at which the sample began to be separated by the separator device. This retention time could be the centroid or apex value of the chromatographic elution profile for the species, or a representative value produced by any chromatographic peak detection method. The peak detection may be carried out directly in the retention time-m / z plane, or sequentially in m / z and then retention time (or vice-versa). Accordingly, step 20 provides a list S of mass peaks detected in the sample, wherein the mass to charge ratio for each sample peak is associated with the retention time in the separator device of the analyte species that gave rise to the ion species. As described above, the accuracy of the mass to charge ratio measurements made at step 20 may vary depending on various factors, and so it may be desired to correct the mass to charge ratios detected by the mass spectrometer. As will be described below, a library of reference mass spectral data is used when making such corrections. As shown at step 22, the method accesses a library L of stored mass spectral peaks for ion species, in which the accurate mass to charge ratio for each library peak is associated with the retention time in a separator device of the analyte species that gave rise to the ion species. The library L of mass spectral data may be created by measuring the mass to charge ratio of each ion species and its respective retention time, optionally under standardised conditions. For example, the retention time value for each species may be standardised during the creation of the library data. This may be done by measuring the retention time of at least one species for which a standardised retention time is available, and the difference between the measured retention time and the standardised retention time for this species may be used to calibrate the retention times measured for the other species when creating the library. Similarly, the mass to charge ratio for each species may be standardised by measuring the mass to charge ratio of at least one reference species that already has a known mass to charge ratio, determining the difference between the measured mass to charge ratio and the known mass to charge ratio for this species, and calibrating the mass to charge ratios measured for the other species when creating the library. Embodiments of the present invention extend to generating the library L, e.g. by mass analysing a single injection of a given sample type, or by mass analysing multiple injections. Alternatively, embodiments of the present invention may simply access a preconstructed library L. It is also contemplated that some or all of the peaks in the library may have theoretical values for mass to charge ratio and / or retention time, rather than values that have been obtained from experimental measurements. As represented by step 24 in Fig. 2, in order to correct the mass to charge ratio measurements obtained at step 20, embodiments of the present invention obtain or estimate the maximum error in a mass to charge ratio measurement that the mass spectrometer can make in step 20, e.g. in units of ppm. In other words, embodiments obtain or estimate what is the worst mass accuracy that the mass spectrometer can have, e.g. in units of ppm. The maximum rate at which the mass accuracy of the mass spectrometer may change with retention time under standard or specified environmental conditions, Rmax, is also determined or estimated. For example, the value of Rmax may be expressed in parts per million per minute (ppm / min). This data is then used in correcting the mass to charge ratio measurements from step 20, as will be described below. As depicted by step 28, at least some of the sample peaks in the list of sample peaks S from step 20 are then matched to at least some of the library peaks in the list of library peaks L from step 22, based on the maximum error in mass to charge ratio measurement that was estimated or determined at step 24. The retention times of the peaks may not be taken into account during this peak matching step 28. When using a TOF mass analyser the dominant cause of error in a mass to charge ratio measurement is a mass drift due to a gain or scaling of the whole mass to charge ratio axis. For example, a +5 ppm drift in mass to charge ratio corresponds to multiplying the whole mass to charge ratio axis by 1.000005, and similarly a -5ppm drift in mass to charge ratio corresponds to multiplying the whole mass to charge ratio axis by 0.999995. As such, the maximum error may be modelled as a gain (i.e. a proportional change, e.g. in ppm). In such cases a library peak may be considered to be matched to a sample peak if the value of the difference in mass to charge ratio between these peaks divided by the mass to charge ratio of the library peak is within the maximum error. It is anticipated that many incorrect matches may be produced at peak-matching stage 28. More specifically, a given library peak may be matched to several sample peaks, and vice-versa, due to said maximum error being relatively large and there potentially being a relatively large number of peaks present in the list of library peaks L and list of sample peaks S since these lists include peaks for a relatively large range of retention times. Accordingly, only some of the sample peaks detected in step 20 may be taken into account during the peak matching step 28. For example, only the sample peaks having an ion abundance (i.e. peak intensity or area) that is over a pre-selected threshold value may be taken into account during this peak matching step. Alternatively, or additionally, only the N most abundant sample peaks from step 20 may be taken into account during the peak matching step, where N is an integer having a value that is <150, <100, <80, <60, <50, <40, <30, <20, <10 or <5. Alternatively, only a quantile of the sample peaks from step 20 may be taken into account during the peak matching step, where the peaks that are taken into account are the most abundant of the sample peaks. For example, <1 %, <5% or <10% of the sample peaks from step 20 may be taken into account during the peak matching step, where the peaks that are taken into account are the most abundant of the sample peaks. The peak matching step 28 may also, or alternatively, be required to meet one or more other minimum peak-matching criterion. For example, the method may identify sample peaks from step 20 that are likely to have large errors in their mass accuracy and exclude these peaks from being matched to library peaks in the peak matching step 28. For instance, the method may identify sample peaks that have been detected when the ion detector or mass analyser is saturated or when there is interference from another species, and may exclude these peaks from the peak matching step. Additionally, or alternatively, to each of the above requirements for step 28, the method may be required to match a pre-selected minimum number of peaks P before it can proceed to step 30 in Fig. 2, e.g. it may be required to match >5, >10, >20, >30, >40 or >50 peaks before it can proceed to step 30. Assuming that at least P peaks are matched in peak-matching step 28, then the method proceeds to step 30, in which the mass spectrometer uses the matched peaks to determine the error in the mass to charge ratios for the sample peaks. As described above, when using a TOF mass analyser the dominant cause of error in a mass to charge ratio measurement is a mass drift due to a gain or scaling of the whole mass to charge ratio axis. Accordingly, the error in a mass to charge ratio measurement may be determined to be a gain value that that the sample peak must be multiplied by in order to obtain the mass to charge ratio of the matched library peak. In such cases the error, i.e. gain value, may be determined as the difference in mass to charge ratio between the matched peaks divided by the mass to charge ratio of the library peak. The present invention allows for the determination of the error in the mass to charge ratios for the sample peaks in other manners. For instance, the error may be determined to be represented by a mathematical function that maps the mass to charge ratios of the sample peaks to the mass to charge ratios of their respective library peaks. For example, the error in the mass to charge ratios for the sample peaks may be expressed as a parameterised function of x, where x=sqrt(m / z), e.g. the error may be expressed as a polynomial in x e.g. a+bx, a+bx+cx2 etc. where the parameters a, b, and c etc. are determined by fitting the function to the list of matched peaks. The fitting procedure used should be tolerant to outliers. A Bayesian method may be used to determine these parameters, which may therefore have associated prior probability distributions. Preferably the error is calculated using a plurality of pairs of matched peaks. For example, the mass to charge ratio error values for a plurality of pairs of matched peak may be averaged, e.g. by taking the median or mean value, so as to obtain a value for the error in the mass to charge ratio for the peaks measured in the sample. It is preferred to use a median or other average value (such as a mean within an inter-quantile range) as this helps to reject outliers in the data. A similar process to that described above in relation to determining the error in mass to charge ratio is performed by the mass spectrometer at step 30 to determine the error in the retention time for the sample peaks. This error may be calculated, for example, as the difference in retention time between the retention time of a sample peak and the retention time of the library peak that it was matched to. It will be appreciated that the difference in retention time is a vector and not merely a magnitude, i.e. if the measured sample peak has a longer retention time than its matched library peak then the error value may be considered to be a negative value as it must be subtracted from the retention time of the sample peak to obtain the library value, whereas if the measured peak has a shorter retention time than its matched library peak then the error value may be considered to be positive value as it must be added to the retention time of the measured sample peak to obtain the library value. Preferably the error is calculated using a plurality of pairs of matched peaks. For example, the difference in retention time between the two peaks in each pair of peaks that has been matched may be calculated so as to obtain a retention time error value for each of these pairs. These retention time error values may then be averaged, e.g. by taking the median or mean value, so as to obtain a value for the error in the retention time for the sample peaks, which may be used to correct the sample peaks as discussed below. It is preferred to use a median or other average value (such as mean within an inter-quantile range), as this helps to reject outliers in the data. In order to account for situations where the error in mass to charge ratio changes with retention time, embodiments are contemplated in which the mass spectrometer determines how the mass to charge ratio error varies as a function of retention time. A preliminary retention time dependence may be established by determining the error in mass to charge ratio for each pair of matched peaks (e.g. in one of the manners discussed above), obtaining the retention time related to each pair of matched peaks (e.g. the retention time of the library peak, or less preferably the retention time of the sample peak), and then using these values to provide a relationship of mass to charge ratio error as a function of retention time. This technique may therefore provide a calibration line or curve of mass to charge ratio error as a function of retention time, which may be used to correct the sample peaks as discussed below in relation to step 32. Fig. 3A shows mass spectral data corresponding to the library peaks L in the graph on the left, and also the mass spectral data corresponding to the sample peaks S in the graph on the right. Each graph represents the relative intensity of the peaks, which is normalised by the most intense peak, as a function of retention time. In this example, step 30 determined that the average error in the mass to charge ratio for the sample peaks was 10.7 ppm and the average error in the retention time was 0.25 mins. Referring back to Fig. 2, as shown by step 32, the error in the mass to charge ratio and the error in the retention time, that were determined in step 30, are then used to adjust the mass to charge ratio and retention time of each sample peak. For example, in the embodiment in which step 30 determines the error in mass to charge ratio as a gain value, step 32 may adjust the mass to charge ratio for each measured sample peak by multiplying it by the gain value. Alternatively, in the embodiment in which step 30 determines the error in mass to charge ratio as a mathematical function, step 32 may adjust the mass to charge ratio for each measured sample peak using the function. As mentioned above, step 32 uses the error in the retention time determined in step 30 to adjust the retention time of each sample peak. The retention time for each measured peak is adjusted by summing it with the average (e.g. median) error in retention time obtained at step 30. If the retention time error was determined to be a negative value then each sample peak will be shifted to a shorter retention time, whereas if the retention time error was determined to be a positive value then each sample peak will be shifted to a longer retention time. Referring again to the example in Fig. 3A, it will be appreciated that the mass to charge ratios and retention times of the sample peaks in the graph on the right will be adjusted using the mass to charge ratio error and retention time error determined at step 30. Referring back to Fig. 2, the method then proceeds to step 34, at which the library peak list L and the sample peak list S are each divided, according to retention time, into multiple sub-lists. Each peak list may be split into two peak lists so as to give two library peak sub-lists L1 ,L2 and two sample peak sub-lists S1 ,S2. For example, each peak list L,S may be split at a particular quantile of retention time (e.g. at the 50% quantile) or at the median or mean retention time of the peaks in the list, although other splits may be used. It is contemplated that the library peak list L may be selected to be divided into two sublists at a particular retention time, and that the sample peak list S is divided into two sublists at a retention time corresponding to said particular retention time summed with the error in retention time that was determined at step 30. However, preferably, each peak list L,S is split at a particular quantile of retention time. It may be assumed that when obtaining both the library peak list and the sample peak list the analyte species will elute from the chromatography device in substantially the same order. If each peak list L,S is split at the same quantile of retention time, then the method remains accurate even if a different chromatography condition is used when obtaining the sample peak list to that which is used when obtaining the library peak list. For example, if the 50% quantile is used to divide the lists then the subdivision will occur at the midpoint of each list of peaks, as ordered by retention time, and not at the halfway point in time of the experiment. In other words, the sample peak list S may be arranged in order of retention time and the list may be divided such that a first percentage of the peaks, which have the lowest retention times, are located in a first sample peak sub-list S1 and the remaining percentage of the peaks, which have higher retention times, are located in a second sample peak sub-list S2. The library peak list L may also be arranged in order of retention time and the list may be divided such that a percentage of the peaks, that is the same as the first percentage, and which have the lowest retention times, are located in a first library peak sub-list L1 and the remaining percentage of the peaks, which have higher retention times, are located in a second library peak sub-list L2. For example, the sample peak list S may be divided such that 50% of the peaks are allocated to the first sample peak sub-list S1 and 50% of the peaks are allocated to second sample peak sub-list S2, and the library peak list L may be divided such that 50% of the peaks are allocated to first library peak sub-list L1 and 50% of the peaks are allocated to second library peak sub-list L2. Fig. 3B shows graphs of the mass spectral shown in Fig. 3A, except wherein the sample peaks (in the graph on the right) have been corrected according to step 32. The vertical lines 40 and 41 represent the division of the library peak list L and the sample peak list S, respectively, by retention time as described above in relation to step 34 of Fig. 2. The peaks on the left side of vertical line 40 may be considered to be library peak sub-list L1 and the peaks on the right of the vertical line 40 may be considered to be library peak sub-list L2. Similarly, the peaks on the left side of vertical line 41 may be considered to be the sample peak sub-list S1 and the peaks on the right of the vertical line 41 may be considered to be sample peak sub-list S2. Although the mass to charge ratios of the sample peaks have been corrected according to step 32, there will still be a residual mass to charge ratio error for each peak in sub-lists S1 and S2, partly due to mass drift during the retention time ranges of these sublists. Referring back to Fig. 2, in order to account for this, the method then proceeds from step 34 to step 36. Step 36 determines a mass to charge ratio tolerance value for each of the sample peak sub-lists S1 ,S2 that were generated at step 34, where this tolerance value is indicative of the estimated maximum amount that the mass accuracy may change by during the range of retention times in that sub-list. The mass to charge ratio tolerance value for each sub-list may be calculated by determining the range of retention times in that sub-list (i.e. the maximum retention time minus the minimum retention time) and multiplying this value by the maximum rate at which the mass accuracy varies with retention time, Rmax, which is obtained from step 24. For example, if the sub-lists S1 and S2 have retention time ranges of 30 mins and 50 mins respectively, and the maximum amount that the mass accuracy may change during each sub-list is 1ppm / min, then the mass to charge ratio tolerance value for sample sub-list S1 is 30 ppm and the mass to charge ratio tolerance value for sample sub-list S2 is 50 ppm. However, it is contemplated that alternative methods for calculating the mass to charge ratio tolerance value may be used. For example, the tolerance value may be scaled up so as to avoid missing peak matches, and the tolerance value may be changed as a function of retention time. The tolerance may also incorporate information about the expected precision of measurement of the position of individual peaks. For example, the tolerance may increase as peak intensity decreases. The method then loops back to step 28, which is performed in the same manner as discussed above, except wherein instead of peak-matching the original library peak list L and the original sample peak list S, the method instead peak-matches peaks from each sample sub-list S1 ,S2 with peaks in their corresponding library sub-list lists L1,L2 based on the mass to charge ratio tolerance value that was determined at step 36. In other words, a sample peak in sample sub-list S1 is matched to a library peak in library sub-list L1 if the sample peak differs from the mass to charge ratio of the library peak by an amount that is equal to or less than the mass to charge ratio tolerance value for sub-list S1 that was obtained at step 36. Similarly, a sample peak in sample sub-list S2 is matched to a library peak in library sub-list L2 if the sample peak differs from the mass to charge ratio of the library peak by an amount that is equal to or less than the mass to charge ratio tolerance value for sub-list S2 that was obtained at step 36. The retention times of the peaks may not be taken into account during this peak matching process. As described previously in relation to the first time step 28 is performed, it is anticipated that when peak matching step 28 is performed on the sub-lists, a given library peak may be matched to multiple sample peaks, and vice-versa. However, the likelihood of such multiple matches for the same peak is reduced as the number of peaks being compared has been reduced since each sub-list has a smaller retention time range than its parent list, and the mass to charge ratio tolerance value may be smaller than the maximum error in mass to charge ratio that was estimated or determined at step 24. The method then uses the matched peaks to determine the error in the mass to charge ratio for the mass peaks in each sample sub-list S1,S2, using a process corresponding to that described above in step 30. For example, as described above, the error in mass to charge ratio may be determined to be a gain value or a mathematical function. As will be appreciated, each corresponding pair of sample and library sub-lists, i.e. S1 ,L1 and S2,L2, may have its own mass to charge ratio error value or function. These errors may be different for different pairs of sub-lists, i.e. the error for the pair S1,L1 may be different for the pair S2,L2. The method may also use the matched peaks to determine the error in the retention time for the peaks in each sample sub-list S1,S2, using a process corresponding to that described above in step 30. That is, the error may be calculated, for example, as the difference in retention time between the retention time of a mass peak measured in the sample sub-list S1,S2 and the retention time of the mass peak in the library sub-list L1,L2 that it was matched to. As will be appreciated, each corresponding pair of sample and library sub-lists, i.e. S1,L1 and S2,L2, will have its own average error in retention time (or mass-retention error profile) and these errors may be different for different pairs of sublists, i.e. the error for the pair S1 ,L1 may be different for the pairS2,L2. Referring back to the example in Fig. 3B, it will be appreciated that the mass to charge ratios of peaks in the sub-list S1 on the left side of the vertical line 41 will be matched to peaks in the library sub-list L1 on the left of vertical line 40, and the mass to charge ratio error and retention time error are determined from these matched peaks. In the depicted example, the mass to charge ratio error was determined to be 11.0 ppm and the retention time error was determined to be 0.59 mins. Similarly, the mass to charge ratios of peaks in the sub-list S2 on the right side of the vertical line 41 will be matched to peaks in the library sub-list L2 on the right of vertical line 40, and the mass to charge ratio error and retention time error are determined from these matched peaks. In the depicted example, the mass to charge ratio error was determined to be 10.5 ppm and the retention time error was determined to be 0.20 mins. The mass to charge ratio and retention time of each peak in each sample sub-list S1,S2 is then adjusted in a corresponding manner to that described previously in relation to step 32. That is, the errors in mass to charge ratio and retention time determined in step 30 for each sample sub-list S1 ,S2 are used to adjust the mass to charge ratio and retention time of each peak in that sample sub-list. For example, in the embodiment in which step 30 determines the error in mass to charge ratio as a gain value, step 32 may adjust the mass to charge ratio for each measured sample peak by multiplying it by the gain value. Alternatively, in the embodiment in which step 30 determines the error in mass to charge ratio as a mathematical function, step 32 may adjust the mass to charge ratio for each measured sample peak using the function. As mentioned above, step 32 uses the error in the retention time determined in step 30 to adjust the retention time of each sample peak. The retention time for each of these peaks is adjusted by summing it with the average (e.g. median) retention time error for sample sub-list S1 determined in step 30. The mass to charge ratio for each peak in the sample sub-list S2 may be adjusted in a corresponding manner to that described above in relation to adjusting each peak in the sample sub-list S1. For example, the mass to charge ratio for each peak in the sample sub-list S2 may be adjusted by multiplying it by a gain value determined for sample sub-list S2 in step 30, or by using a mathematical function determined for sample sub-list S2 in step 30. The retention time for each of these peaks is adjusted by summing it with the average retention time error for sample sub-list S2 determined in step 30. Referring again to the example in Fig. 3B, the mass to charge ratio and retention time of each peak in the sample sub-list S1 (on the left of the vertical line 41) are adjusted using the determined mass to charge ratio error and retention time error for sample sub-list S1. Similarly, the mass to charge ratios of peaks in the sub-list S2 (on the right side of the vertical line 41) are adjusted using the determined mass to charge ratio error and retention time error for the sample sub-list S2. The process described above in relation to step 34 is then repeated for each sample sub-list S1,S2 so as to provide multiple new sample sub-lists derived from each sample sub-list S1 ,S2. For example, the library peak sub-list L1 and its corresponding sample peak sub-list S1 are each divided, according to retention time, into multiple new sub-lists. Each sub-list may be split into two new sub-lists to give two new library peak sub-lists L1a,L1b and two sample peak sub-lists S1a,S1b. For example, each peak sub-list may be split at the 50% quantile of retention time or the median or mean retention time of the peaks in the list, or at another retention time. Similarly, the library peak sub-list L2 and its corresponding sample peak sub-list S2 are each divided, according to retention time, into multiple new sub-lists. Each sub-list may be split into two new sub-lists to give two new library peak sub-lists L2a,L2b and two sample peak sub-lists S2a,S2b. For example, each peak sub-list may be split at the 50% quantile of retention time or the median or mean retention time of the peaks in the list, or at another retention time. Fig. 3C shows graphs of the mass spectral data shown in Fig. 3B, except wherein the sample peaks have been corrected according to step 32. The vertical line 42 represents the division of the library peak sub-list L1 into the two further sub-lists L1a,L1b, and the vertical line 43 represent the division of the library peak sub-list L2 into the two further sub-lists L2a,L2b. Similarly, the vertical line 44 represents the division of the sample peak sub-list S1 into the two further sub-lists S1a,S1b, and the vertical line 45 represent the division of the sample peak sub-list S2 into the two further sub-lists S2a,S2b As depicted by arrow 35 in Fig. 2, if it is determined that fewer than a threshold number of peaks are present in any one of the new sample sub-lists S1a,S1b,S2a,S2b, then the method does not proceed to step 36 for that new sample sub-list. Instead, as indicated at step 38, the retention time range for that new sub-list is assigned the mass to charge ratio error and retention time error that was determined at step 30 for the sample sub-list S1 or S2 that it was divided from. For example, if it is determined that fewer than 2xN peaks are present in any one of the new sample sub-lists S1a,S1b,S2a,S2b (where, as described above, N is the number of most abundant sample peaks that are considered in the peak matching step), then the method may proceed to step 38. Alternatively, or additionally, when determining if fewer than the threshold number of peaks are present in any one of the new sample sub-lists, only peaks that have an intensity or area above a preselected (non-zero) minimum value may be taken into account. On the other hand, if it is determined that greater than or equal to the threshold number of peaks are present in any one of the new sample sub-lists S1 a,S1 b,S2a,S2b then the method proceeds to step 36 for that new sample sub-list. For each new sample sub-list that proceeds to step 36, a mass to charge ratio tolerance value is determined in a corresponding manner to that described above in relation to step 36. The mass to charge ratio tolerance value for each new sub-list may be calculated by determining the range of retention times in that new sub-list (i.e. the maximum retention time minus the minimum retention time) and multiplying this value by the maximum rate at which the mass accuracy varies with retention time, Rmax, which is obtained from step 24. The cycle described above in relation to steps 28-36 may then be repeatedly cycled through, based on each new sample sub-list that is generated the previous time that step 34 is performed. However, for any given sample sub-list, when it is not possible to peak match a pre-selected number of peaks P in the sample sub-list with peaks in the corresponding library sub-list at step 28, then the method may proceed to step 38 for that sub-list. In this instance, the retention time range for that sub-list is assigned the mass to charge ratio error and retention time error that was determined the previous time that step 30 was performed on the sample sub-list that it was divided from. Similarly, if after step 34 it is determined that fewer than a threshold number of peaks are present in any one of the sample sub-lists, then the method may not proceed to step 36 for that sample sub-list but may instead proceed to step 38. In this instance, the retention time range for that sub-list is assigned the mass to charge ratio error and retention time error that was determined at step 30 for the sample sub-list that it was divided from. For example, for each of the sample sub-lists S1a,S1b,S2a,S2b created at step 34 that have been subjected to step 36, the method may loop back to step 28. Step 28 is then performed in the same manner as discussed above, except wherein instead of peakmatching the sample sub-list S1,S2 with peaks in their corresponding library sub-list lists L1,L2, the method instead peak-matches peaks from each new sample sub-list S1a,S1b,S2a,S2b with peaks in their corresponding library sub-list list L1a,L1b,L2a,L2b based on the mass to charge ratio tolerance value that was just determined at step 36. For instance, a sample peak in new sample sub-list S1a is matched to a library peak in library sub-list L1a if the sample peak differs from the mass to charge ratio of the library peak by an amount that is equal to or less than the mass to charge ratio tolerance value for that sub-list S1a that was obtained at step 36. The likelihood of a sample peak being matched to multiple library peaks, or vice versa, is relatively low as the number of peaks being compared has been reduced since each sub-list has a smaller retention time range than its parent list, and the mass to charge ratio tolerance value is smaller than the mass to charge ratio tolerance value used when matching peaks in the parent list. It will therefore be appreciated that the iterative nature of the method causes the peak matching to improve. If it is possible to peak match a pre-selected number of peaks P in any given new sample sub-list S1a,S1b,S2a,S2b with peaks in its corresponding library sub-list L1a,L1b,L2a,L2b, the method conducts step 30 on that new sample sub-list. Referring back to the example in Fig. 3C, the mass to charge ratios of peaks in the sample sub-list S1a on the left of vertical line 44 were matched to peaks in the library sublist L1a on the left of vertical line 42, and the mass to charge ratio error was determined to be 11.1 ppm and the retention time error was determined to be 0.78 mins. The mass to charge ratios of peaks in the sample sub-list S1b between vertical lines 41 and 44 were matched to peaks in the library sub-list L1b between vertical lines 40 and 42, and the mass to charge ratio error was determined to be 10.9 ppm and the retention time error was determined to be 0.32 mins. The mass to charge ratios of peaks in the sample sub-list S2a between vertical lines 41 and 45 were matched to peaks in the library sub-list L2a between vertical lines 40 and 43, and the mass to charge ratio error was determined to be 10.8 ppm and the retention time error was determined to be 0.24 mins. The mass to charge ratios of peaks in the sample sub-list S2b on the right of vertical line 45 were matched to peaks in the library sub-list L2b on the right of vertical line 43, and the mass to charge ratio error was determined to be 10.2 ppm and the retention time error was determined to be 0.07 mins. Referring back to Fig. 2, the method may then go on to steps 32 and 34. Fig. 3D shows an example where these steps have been performed on all of the sub-lists S1a,S1b,S2a,S2b,L1a,L1b,L2a,L2b. Fig. 3D shows graphs of the mass spectral data shown in Fig. 3C, except wherein the sample peaks have been corrected according to step 32. The vertical line 46 represents the division of the library peak sub-list L1a into the two further sub-lists, the vertical line 47 represent the division of the library peak sub-list L1b into the two further sub-lists, the vertical line 48 represent the division of the library peak sub-list L2a into the two further sub-lists, and the vertical line 49 represent the division of the library peak sublist L2b into the two further sub-lists. Similarly, the vertical line 50 represents the division of the sample peak sub-list S1a into the two further sub-lists, the vertical line 51 represent the division of the sample peak sub-list S1b into the two further sub-lists, the vertical line 52 represent the division of the sample peak sub-list S2a into the two further sub-lists, and the vertical line 53 represent the division of the sample peak sub-list S2b into the two further sub-lists. The method may then perform step 36 on each sub-list, and may loop back again to perform steps 28 and 30. In the example depicted in Fig. 3D, the mass to charge ratios of peaks in the sample sub-list on the left side of the vertical line 50 were matched to peaks in the library sub-list on the left of vertical line 46, and the mass to charge ratio error was determined to be 11.2 ppm and the retention time error was determined to be 0.91 mins. The mass to charge ratios of peaks in the sample sub-list between vertical lines 44 and 50 were matched to peaks in the library sub-list between vertical lines 42 and 46, and the mass to charge ratio error was determined to be 11.1 ppm and the retention time error was determined to be 0.66 mins. The mass to charge ratios of peaks in the sample sub-list between vertical lines 44 and 51 were matched to peaks in the library sub-list between vertical lines 42 and 47, and the mass to charge ratio error was determined to be 10.8 ppm and the retention time error was determined to be 0.55 mins. The mass to charge ratios of peaks in the sample sub-list between vertical lines 41 and 51 were matched to peaks in the library sub-list between vertical lines 40 and 47, and the mass to charge ratio error was determined to be 10.9 ppm and the retention time error was determined to be 0.27 mins. The mass to charge ratios of peaks in the sample sub-list between vertical lines 41 and 52 were matched to peaks in the library sub-list between vertical lines 40 and 48, and the mass to charge ratio error was determined to be 11.1 ppm and the retention time error was determined to be 0.23 mins. The mass to charge ratios of peaks in the sample sub-list between vertical lines 45 and 52 were matched to peaks in the library sub-list between vertical lines 43 and 48, and the mass to charge ratio error was determined to be 10.5 ppm and the retention time error was determined to be 0.25 mins. The mass to charge ratios of peaks in the sample sub-list between vertical lines 45 and 53 were matched to peaks in the library sub-list between vertical lines 43 and 49, and the mass to charge ratio error was determined to be 10.5 ppm and the retention time error was determined to be 0.06 mins. The mass to charge ratios of peaks in the sample sub-list on the right of vertical line 53 were matched to peaks in the library sub-list on the right of vertical line 49, and the mass to charge ratio error was determined to be 9.7 ppm and the retention time error was determined to be 0.12 mins. As described above, when the looped method is being performed, if it is determined that fewer than the pre-selected number of peaks P in any given sample sub-list are matched with peaks in its corresponding library sub-list at step 28, then the method does not proceed to step 30 for that sample sub-list. Instead, as depicted by arrow 36, the method proceeds to step 38 at which the retention time range for that sub-list may be assigned the mass to charge ratio error and retention time error that was determined at step 30 for the sub-list that it was divided from. Also, as described above and depicted by arrow 35 in Fig. 2, if it is determined that fewer than a threshold number of peaks are present in any one of the sample sub-lists, then the method does not proceed to step 36 for that sample sub-list. Instead, as indicated at step 38, the retention time range for that sublist is assigned the mass to charge ratio error and retention time error that was determined at step 30 for the sample sub-list that it was divided from. Eventually the method will proceed to step 38 for all sample sub-lists. A final mass to charge ratio error and retention time error is therefore obtained for the retention time range of each sample sub-list that proceeds to step 38. It will be appreciated that the original sample peak list S may have been subdivided into a greater number of sub-lists over a first range of retention time values than over a second, different but equally sized range of retention time values, if more peaks in the first range are able to be matched to library peaks than for the second range. The cumulative mass to charge ratio error and cumulative retention time error that has been applied to each mass peak in the original sample peak list S in order to get to step 38 may then be calculated, and may be plotted as a function of retention time, e.g. as shown in Figs. 4A and 4B. Fig. 4A shows an example of how the mass to charge ratio error (in ppm) varies as a function of retention time (in mins) for the mass analysis of peptides in a tryptic lysate of E. coli bacteria. Fig. 4B shows an example of how the retention time error (in mins) varies as a function of retention time (in mins) for the mass analysis of the E. coli bacteria. The cumulative mass to charge ratio error and cumulative retention time error that has been applied to each mass peak may then be used to correct their respective sample peaks. Alternatively, the values of the mass to charge ratio error as a function of retention time may be smoothed or otherwise simplified so as to obtain calibration data that is representative of a calibration curve or line of mass to charge ratio error as a function of retention time. This smoothed or simplified calibration relationship may then be used to correct the sample peaks. Although the present invention has been described with reference to preferred embodiments, it will be understood by those skilled in the art that various changes in form and detail may be made without departing from the scope of the invention as set forth in the accompanying claims. For example, although step 34 has been described as dividing each list of peaks into two sub-lists, it is contemplated that this step may instead divide each list of peaks into more the two sub-lists. Embodiments have been described in which the overall output of the method is a set of sample peaks having corrected mass to charge ratios. However, the method described also provides information that allows the retention times of the sample peaks to be standardised, regardless of whether or not that information is then used to adjust the retention times of the sample peaks from their original values. It is therefore contemplated that the overall output of the method may be a set of sample peaks having corrected retention times, and that the mass to charge ratios of those sample peaks may or may not be corrected. Correcting the retention times of the sample peaks in this manner is advantageous, for example, so as to allow the sample peaks to be searched against a library of spectral data that includes retention time values. Additionally, or alternatively, realigning the retention times of the sample peaks in the manner described allows improved tracking of species across multiple samples, which may be important in many applications, particularly in experiments that aim to identify changes in abundance of a component between samples. The method may therefore be repeated on multiple samples, and the realigned peaks from those different samples that have substantially the same retention time may be compared, e.g. to determine how the abundance (e.g. intensity or peak area) of the peak has changed. The method may be used to align the retention times in two or more samples, e.g. by designating the peaks from one sample as the library peaks, even if there is nothing (a priori) that singles out that sample as being special, and then treating the peaks from the other sample as the sample peaks in the above-described method. It is envisaged that more than two samples could be processed simultaneously, e.g. with peak matching etc. taking place across all corresponding sub-lists. It is contemplated that the method described herein may be performed in a postprocessing step after the sample has been fully mass analysed by step 20 and all of the mass spectral data has been obtained. Alternatively, the method may be cycled through in real-time during the mass analysis of the sample, i.e. on the fly. When the method is performed in real-time during the mass analysis of the sample, it will be appreciated that at any given retention time during the mass analysis the sample peak list S obtained at step 20 will be incomplete as it will contain only peaks that have been detected up until that time. Those peaks will be subjected to the looped method described in relation to Fig. 2. When the mass analysis of the sample has progressed further to a later retention time, more sample peaks will have been detected and these peaks will be subjected to the looped method described in relation to Fig. 2. This may be repeated until the mass analysis of the sample is complete. In the embodiments described above, each sub-list may be treated independently in that the mass to charge ratio correction, and optionally the retention time correction, that is determined and applied to a given sub-list is independent of the corresponding correction(s) that is determined and applied to other sub-lists that have the same parent list. However, it is alternatively contemplated that, following the matching process at each level (i.e. following step 28), a degree of additional consistency may be required between corrections that are applied to sub-lists that have the same parent list. This consistency could be enforced by seeking to determine a single mass to charge ratio correction, and / or retention time correction, function across the retention time range that covers all of the sublists that have the same parent list, which is consistent with the peak matches determined for each sub-list. The lowest level sub-list available for each retention time range would be used for this purpose where the subdivision process has already been terminated according to step 38. The function may be a simple piecewise linear interpolation or a higher order spline, e.g. cubic spline, through the points determined in step 28 as the matched data points. Additional constraints on the slope and / or curvature of this spline may be imposed to add smoothness or for consistency with known characteristics of the instrumentation, such as maximum mass to charge ratio accuracy drift (e.g. in ppm / min). Optionally, this smoothing process may be applied only after the final stage of subdivision to obtain the final overall corrections to be applied to the data. As described hereinabove, the dominant change in mass to charge ratio calibration during an experiment may be a single parameter such as a gain describing a stretching of the mass to charge ratio axis; e.g. m / z -> g*(m / z) where g is a function of retention time. More generally, however, there may be more than one parameter describing the change in mass to charge ratio calibration as a function of retention time. For example, there may be a change corresponding to gain and offset related in ion arrival time during a TOF mass analyser experiment. This can be described by two parameters, e.g. a + b*sqrt(m / z) where a and b are independent functions of retention time. As above, the size, slope or curvature of these parameters considered as functions of retention time may be restricted or controlled. Even more generally, the change may be an nth degree polynomial in sqrt(m / z) with n+1 coefficients that are all independent functions of retention time. Many other functions of mass to charge ratio are possible including splines, such as cubic splines. In embodiments described above, peak lists are recursively divided into two sublists. This is advantageous from the point of view of speed and algorithmic simplicity, but the boundary that divides any two sub-lists persists through all subsequent stages of the processing and this could, for example, lead to a situation in which a library peak and its counterpart sample peak are placed on different sides of a boundary between sub-lists at an early stage of the processing. For example, a sample peak may be arranged in sample peak sub-list S1, when it is actually a correct match for a library peak present in library peak sub-list L2. The algorithm as described would not be able to peak match these two peaks, as sub-list S1 is not compared to sub-list L2 in the peak matching step 28, and it also cannot recover a correct match for this peak at a later stage in the processing. There will typically be sufficiently many peaks remaining in the sub-lists to ensure peak matching that enables the mass to charge ratio and retention time corrections that are required. However, in order to avoid this issue, it is envisaged that the process may be repeated but wherein the boundary between sub-lists is adjusted at each cycle of processing, or even completely recalculated. For example, relatively small random changes may be made to the position of the list boundaries at each cycle of processing, e.g. such that a few peaks are effectively moved from one sub-list to an adjacent sub-list. This may help to reduce any artefacts resulting from the issue described above. Alternatively, sub-lists could be freshly determined during each cycle of the processing, which also allows complete freedom in the number of sub-lists to be used at each cycle. This would also allow a slower increase in the number of sub-lists at each cycle compared with the exponential growth in the embodiments described above. Although this may increase processing time it could potentially improve the quality of the final result in some scenarios. Embodiments have been described in which in which a sample is separated in a LC or GC chromatography device and then mass analysed so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the chromatography device. However, rather than the mass to charge ratio for each peak being associated with a retention time in the chromatography device, it is contemplated that the ionised sample may be separated in an ion separator, such as an ion mobility separator, and then mass analysed so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in (i.e. elution from) the ion separator device. In such embodiments the list of library peaks that is provided has a mass to charge ratio for each peak that is associated with a retention time in (i.e. elution from) the ion separator device. These embodiments may operate in a corresponding manner to those described above, except wherein the retention time in (i.e. elution from) the ion separator device is used in each step that has been described as using the retention time in the chromatography device. It is also contemplated that the method may comprise separating the sample in a LC or GC chromatography device, ionising the sample, separating the resulting ions in an ion separator such as an ion mobility separator, and then mass analysing the ions so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the chromatography device and a retention time in (i.e. elution from) the ion separator device. In such embodiments the list of library peaks that is provided has a mass to charge ratio for each peak that is associated with a retention time in the chromatography device and a retention time in (i.e. elution from) the ion separator device. These embodiments may operate in a corresponding manner to those described above, except wherein both the retention time in the chromatography device and the retention time in (i.e. elution from) the ion separator device are used in each step that has been described as using the retention time in the chromatography device. Embodiments have been described for correcting errors in mass to charge ratios. However, alternatively, the ion mobility of the ions may be analysed and the method may correct the error in ion mobility that is detected. Such methods correspond to those described hereinabove, except rather than mass analysing the sample so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the chromatography device, the method comprises mobility analysing the sample so as to obtain a list of sample peaks in which a mobility for each peak is associated with a retention time in the chromatography device. In such embodiments the list of library peaks that is provided has a mobility for each peak that is associated with a retention time in the chromatography device. These embodiments may operate in a corresponding manner to those described above, except wherein the mobility is used in each step that has been described as using the mass to charge ratio. As described above, embodiments of the invention determine the error in retention time between the matched library and sample peaks. This error may be used to provide diagnostic information about the performance of the chromatography system. For example, the onset of degradation of the performance of a chromatography separation column or a pump associated with the chromatography device may be detected early by detecting that the errors in retention time between the matched library and sample peaks has increased. This information may be signalled to the user or instrument servicing provider, e.g. via an electronic display and optionally via cloud-based diagnostics. Transient problems that affect a single sample injection may also be highlighted in the same manner. More specifically, the method described herein associates sample peak data to library peak data that has been acquired under close to ideal conditions and / or that has been standardised and corrected as far as possible. These associations can therefore be used to provide information regarding non-idealities or changes in performance of the instrument that is used to perform the analysis of the sample that may arise over time. For example, the error in the mass to charge ratio (or mobility) of the sample peak relative to the library peak, or the rate of change of this mass to charge ratio (or mobility) error with retention time may be monitored and used for diagnostic purposes. If the spectrometer determines that the error or rate of change of the error exceeds a selected threshold or tolerance then a warning may be sent and / or displayed, e.g. that servicing of the instrument is required. Similarly, the error in retention time of the sample peak relative to the library peak, or the rate of change of this error with retention time may be monitored and used for diagnostic purposes in a corresponding manner to that described above. Since the method disclosed herein creates large numbers of probable associations between peaks in the library and peaks in the sample, other characteristics of the data may be monitored including, but not limited to, instrument resolution, peak width, peak shape, relative abundance of peaks, isotopic distributions, sensitivity, detector performance etc. Such characteristics of the peaks may be monitored an any or all analytic dimensions including chromatographic, mass spectral and ion mobility. Any changes detected in one or more of these characteristics may be diagnostic of changes in instrument performance that might eventually lead to loss of instrument performance or specification, particularly if they trend with time (i.e. over the analysis of multiple samples). This detected change may be used to trigger some kind of warning to the user, manufacturer and / or servicing company and potentially trigger some kind of intervention such as an engineer visit, an instrument service, or replacement of parts in the instrument etc.

Claims

1. A method of mass spectrometry comprising:a) separating a sample in a separator device and then mass analysing the sample so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the separator device;b) providing a list of library peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device;c) dividing the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and dividing the list of library peaks, according to retention time, into multiple library peak sub-lists;d) peak-matching at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks;e) determining an error in the mass to charge ratio of the first peak in the first sample peak sub-list based on the mass to charge ratios of the peaks in the first pair of matched peaks; andf) adjusting the mass to charge ratio of at least the first peak in the first sample peak sub-list using said error in mass to charge ratio so as to provide an adjusted first sample peak sub-list.

2. The method of claim 1, wherein the separator device is a liquid or gas chromatography separation device.

3. The method of claim 1 or 2, comprising:g) peak-matching at least a second peak in a second of the sample peak sub-lists with at least a second respective peak in a second of the library peak sub-lists so as to obtain at least a pair of matched peaks;h) determining an error in the mass to charge ratio of the second peak in the second sample peak sub-list based on the mass to charge ratios of the peaks in the pair of matched peaks; andi) adjusting the mass to charge ratio of at least the second peak in the second sample peak sub-list using the error in mass to charge ratio.lethod of any preceding claim, wherein during peak-matching step d), and / or during peak-matching step g), only sample peaks that have a peak intensity or area over a pre-selected threshold value are taken into account during the peak matching; and / or wherein only the N most abundant sample peaks are taken into account during the peak matching, where N is an integer having a value that is <150, <100, <80, <60, <50, <40, <30, <20, <10 or <5.

5. The method of any preceding claim, wherein the method only proceeds from peakmatching step d) to step error-determining step e), and / or only proceeds from peakmatching step g) to step error-determining step h), if a pre-selected minimum number of peaks P are matched in that peak-matching step.

6. The method of any preceding claim, wherein when determining if peaks match in peak-matching step d), and / or in peak-matching step g), the method does not match the retention time of the sample peak to the retention time of the library peak.

7. The method of any preceding claim, wherein step c) comprises arranging the peaks in the list of sample peaks according to retention time and then dividing this list such that the first sample peak sub-list comprises or consists of peaks in a first quantile of the list of sample peaks arranged according to retention time; and arranging the peaks in the list of library peaks according to retention time, and then dividing this list such that the first library peak sub-list comprises or consists of the same quantile of the list of library peaks arranged according to retention time that is the same as the first quantile.

8. The method of any preceding claim, wherein peak-matching step d) comprises peak-matching a plurality of peaks in the first sample peak sub-list with a plurality of respective peaks in the first library peak sub-list so as to obtain a plurality of pairs of matched peaks; and error-determining step e) comprises determining an error in the mass to charge ratio of the sample peak in each of the plurality of pairs of peaks, based on the mass to charge ratios of the peaks in that pair of matched peaks, so as to obtain a plurality of mass to charge ratio error values; averaging these error values so as to obtain an average error value, and then performing said step of adjusting the mass to charge ratios of sample peaks in the first sample peak sub-list using the average error value.

9. The method of any preceding claim, wherein peak-matching step d) comprises peak-matching a plurality of peaks in the first sample peak sub-list with a plurality ofaks in the first library peak sub-list so as to obtain a plurality of pairs of matched peaks at different respective retention times;wherein error-determining step e) comprises determining an error in the mass to charge ratio of the sample peak in each of the plurality of pairs of peaks, based on the mass to charge ratios of the peaks in that pair of matched peaks, so as to obtain a relationship of mass to charge ratio error value as a function of retention time; andwherein step f) comprises adjusting the mass to charge ratios of sample peaks in the first sample peak sub-list using said relationship of mass to charge ratio error value as a function of retention time.

10. The method of any preceding claim, further comprising performing a cycle that comprises:j) dividing one of the sample peak sub-lists, according to retention time, into multiple further sample peak sub-lists, and dividing a corresponding library peak sub-list, according to retention time, into multiple further library peak sub-lists;k) peak-matching at least a first peak in a first of said further sample peak sub-lists with at least a first respective peak in a first of said further library peak sub-lists so as to obtain at least a first pair of matched peaks;I) determining an error in the mass to charge ratio of the first peak in the first further sample peak sub-list based on the mass to charge ratios of the peaks in the first pair of matched peaks; andm) adjusting the mass to charge ratio of at least the first peak in the first further sample peak sub-list using said error in mass to charge ratio;wherein the cycle of steps j) to I) is repeatedly performed, and wherein each time the cycle is performed the sample peak sub-list that is divided in step j) is the adjusted sample peak sub-list produced in step m) of the preceding cycle.

11. The method of claim 10, wherein peak-matching step k) only matches peaks that differ in mass to charge ratio by less than or equal to a mass to charge ratio tolerance value; wherein the mass to charge ratio tolerance value is calculated by: estimating the maximum rate at which the mass accuracy of the mass spectrometer varies with retention time, determining the range of retention times in the first further sample peak sub-list and multiplying this value by the maximum rate at which the mass accuracy varies with retention time.

12. The method of claim 10 or 11, wherein any given time that the cycle is being performed the method only proceeds from peak-matching step k) to error-determining stepicted minimum number of peaks Pina sample peak sub-list are matched to peaks in a library peak sub-list during peak-matching step k).

13. The method of claim 12, wherein if said pre-selected minimum number of peaks P in the sample peak sub-list are not matched to peaks in a library peak sub-list then the method does not proceed from peak-matching step k) to error-determining step I), and instead the retention time range for that sample peak sub-list is assigned a final mass to charge ratio error and / or a final retention time error corresponding to the mass to charge ratio error and / or retention time error, respectively, that were determined for the sample peak sub-list that it was divided from.

14. The method of any one of claims 10 to 13, wherein any given time that the cycle is being performed the method only proceeds from dividing step j) to peak-matching step k) if it is determined that greater than or equal to a threshold number of peaks are present in the first further sample peak sub-list.

15. The method of claim 14, wherein if it is determined that fewer than said threshold number of peaks are present in the first further sample peak sub-list then the method does not proceed from dividing step j) to peak matching step k), and instead the retention time range for that sample peak sub-list is assigned a final mass to charge ratio error and / or final retention time error corresponding to the mass to charge ratio error and retention time error, respectively, that were determined for the sample peak sub-list that it was divided from.

16. The method of claim 13 or 15, where the cycle is repeatedly performed until a final mass to charge ratio error, and / or a final retention time error, is assigned to the retention time range of all of the sample peak sub-lists that are produced.

17. The method of claim 16, comprising correcting the mass to charge ratios and / or retention times of sample peaks in the list of sample peaks using the final mass to charge ratio errors and / or the final retention time errors associated with retention times that correspond to the retention times of the sample peaks in the list of sample peaks; or using the final mass to charge ratio errors and retention time errors of the sample peak sub-lists to obtain calibration data that is representative of a calibration curve or line of mass to charge ratio error as a function of retention time, and using said calibration data to correct the mass to charge ratios and retention times of sample peaks in the list of sample peaks.

18. A computer implemented method of correcting the mass to charge ratios and retention times of sample peaks in mass spectral data, the method comprising:providing a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device;providing a list of library peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device;dividing the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and dividing the list of library peaks, according to retention time, into multiple library peak sub-lists;peak-matching at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks;determining an error in the mass to charge ratio of the first peak in the first sample peak sub-list based the mass to charge ratios of the peaks in the first pair of matched peaks; andadjusting the mass to charge ratio of at least the first peak in the first sample peak sub-list using said error in mass to charge ratio so as to provide an adjusted first sample peak sub-list.

19. A computer-readable medium comprising instructions which, when executed on a computer, cause the computer to carry out the method of claim 18.

20. A mass spectrometer comprising:a separator device for separating a sample;a mass analyser for mass analysing the separated sample so as to obtain a list of sample peaks in which a mass to charge ratio for each peak is associated with a retention time in the separator device;control circuitry configured to:access a list of library peaks in which a mass to charge ratio for each peak is associated with a retention time in a separator device;divide the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and divide the list of library peaks, according to retention time, into multiple library peak sub-lists;peak-match at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks;determine an error in the mass to charge ratio of the first peak in the first sample peak sub-list based on the mass to charge ratios of the peaks in the first pair of matched peaks; andadjust the mass to charge ratio of at least the first peak in the first sample peak sub-list using said error in mass to charge ratio so as to provide an adjusted first sample peak sub-list.

21. A method of ion mobility spectrometry comprising:a) separating a sample in a separator device and then ion mobility analysing the sample so as to obtain a list of sample peaks in which a mobility for each peak is associated with a retention time in the separator device;b) providing a list of library peaks in which a mobility for each peak is associated with a retention time in a separator device;c) dividing the list of sample peaks, according to retention time, into multiple sample peak sub-lists, and dividing the list of library peaks, according to retention time, into multiple library peak sub-lists;d) peak-matching at least a first peak in a first of the sample peak sub-lists with at least a first respective peak in a first of the library peak sub-lists so as to obtain at least a first pair of matched peaks;e) determining an error in mobility of the first peak in the first sample peak sub-list based on the mobilities of the peaks in the first pair of matched peaks; andf) adjusting the mobility of at least the first peak in the first sample peak sub-list using said error in mobility so as to provide an adjusted first sample peak sub-list.

Images