Longitudinal quality control monitoring

Abstract

A computer-implemented method for longitudinally monitoring a quality of data obtained from a measurement system configured to analyse a sample comprises the steps of receiving measurement data of a sample from a measurement system, subjecting the measurement data to data analysis, whereby at least one data parameter associated with the measurement data is obtained, and associating the data parameter with at least one reference data parameter associated with reference measurement data, and monitoring a quality of the measurement data based on the association of the data parameter with the reference data parameter. The method further comprises the step of receiving at least one operating parameter being associated with an operation of the measurement system and associating the operating parameter with at least one reference operating parameter being associated with a reference operation of the measurement system and monitoring the quality of the measurement data based on their association. Additionally or alternatively, receiving at least one sample parameter being associated with the sample and associating the sample parameter with at least one reference sample parameter being associated with the reference sample and monitoring the quality of the based on their association.

Description

[0001] TITLE

[0002] LONGITUDINAL QUALITY CONTROL MONITORING

[0003] TECHNICAL FIELD

[0004] The present invention relates to a computer-implemented method for monitoring the quality of data obtained from a measurement system configured to analyse a sample according to claim 1 , to a computer program product causing a computer to carry out said method according to claim 20, and to a measurement system for analysing a sample comprising a processor that carries out said method according to claim 21 .

[0005] PRIOR ART

[0006] In the analysis of a sample by a measurement system such as in the analysis of biological tissue samples by mass spectrometry, a large amount of information of the tissue samples is obtained. At the same time, the measurement is subject to a number of possible interferences, which can lead to distortions and to errors in the information gained.

[0007] On the one hand, these interferences can result from the performance of the measurements or the data analysis itself. Often, however, other influences also play a decisive role, such as environmental influences, from the sample preparation, or from an operation or handling of the measurement system.

[0008] Early detection and elimination of such interferences is crucial for obtaining measurement data of good quality.

[0009] SUMMARY OF THE INVENTION

[0010] It is an object of the present invention to provide a method for improved quality control of measurement data. This object is achieved with a computer-implemented method according to claim 1. That is, a computer-implemented method for longitudinally monitoring a quality of data obtained from a measurement system configured to analyse a sample is provided. The method comprises the steps of i) Receiving measurement data of a sample from a measurement system that has carried out at least one measurement on the sample, ii) Subjecting the measurement data to data analysis, whereby at least one data parameter associated with the measurement data is obtained, iii) Associating the data parameter with at least one reference data parameter associated with reference measurement data received from at least one reference measurement that has been carried out for at least one refence sample by the measurement system, and iv) Monitoring a quality of the measurement data based on the association of the data parameter with the reference data parameter. The method further comprises the step of v) Receiving at least one operating parameter being associated with an operation of the measurement system and associating the operating parameter with at least one reference operating parameter being associated with a reference operation of the measurement system, and wherein the monitoring of the quality of the measurement data is further based on the association of the operating parameter with the reference operating parameter, and / or vi) Receiving at least one sample parameter being associated with the sample and associating the sample parameter with at least one reference sample parameter being associated with the reference sample, and wherein the monitoring of the quality of the measurement data is further based on the association of the sample parameter with the reference sample parameter.

[0011] That is to say, the present invention is based on the insight that an improved quality monitoring can be provided by taking into account one or more operating parameters associated with the operation of the measurement system and / or one or more sample parameters being associated with the sample. These operational parameter(s) and / or sample parameter(s) are considered in addition to the regular quality monitoring based on the measurement data obtained from the measurement system and the reference measurement data.

[0012] To this end it is a longitudinal monitoring being performed, i.e. the measurements and the reference measurements are carried out by the same measurement system. More generally speaking, the (reference) operating parameter(s) and the data parameter(s) are associated with or obtained from the same measurement system, respectively. Additionally or alternatively, the operating parameter and the reference operating parameter being associated with one another are preferably of a same type. For instance, operating parameter and reference operating parameter of a same type can be firmware versions of the measurement system, serial numbers of the measurement system, data acquisition software versions of the measurement system, hardware components of the measurement system, measurement configurations of the measurement system, tuning variables of the measurement system, readback values of the measurement system, data acquisition methods being performed by the measurement system when carrying out the measurement on the sample and the reference measurement on the reference sample, etc., see also further below.

[0013] It should be noted that the monitoring can be based on single parameters such as on a single (reference) data parameter, a single (reference) operating parameter and / or a single (reference) sample parameter. However, it is likewise conceivable that the monitoring is based on two or more, in particular on a plurality of these parameters. Thus, statements made with respect to a single parameter preferably likewise apply to two or more parameters and vice versa.

[0014] The data parameter and the reference data parameter, and / or the operating parameter and the reference operating parameter, and / or the sample parameter and the reference sample parameter are preferably individually assessed and associated with one another. That is, it is preferred that no correlation among these different parameters is performed.

[0015] The data parameter can be preferably derived from measurement data of a single measurement that has been carried out by the measurement system on the sample. Alternatively, the data parameter can be derived from measurement data of two or more measurements that have been carried out on the sample. That is, the present method allows a monitoring of the quality of measurement data from one measurement or from multiple measurements.

[0016] Likewise, it is conceivable that the at least one reference data parameter is obtained from a single measurement or from multiple measurements of the reference sample. Moreover, in the event of two or more reference measurements, it is conceivable that a single reference sample is measured several times or that several reference samples of a same type of samples are measured several times. The data analysis of the measurement data and the data analysis of the reference measurement data is preferably carried out separately.

[0017] To this end the reference data parameter is preferably obtained from one or more reference measurements that have been carried out by the measurement system on the reference sample prior to the actual measurements of the actual sample, i.e. the sample of interest. Such a reference measurement could be a measurement being performed before the measurement system is put into operation and / or a measurement right after an inspection of the measurement system, for instance. Consequently, these reference measurements are preferably measurements by the measurement system where the measurement data have been checked and have been rated as good. At this point it is noted that the quality monitoring according to the present invention can be performed directly on the measurement data, wherein no specific quality control (QC) samples are required. QC samples are typically used, e.g. once per week. They are typically only used to check a measurement system but not a sample parameter associated with the sample as it is the case in the present invention. However, QC samples can of course likewise be used in the method according to the invention.

[0018] Moreover, the reference data parameter and the reference operating parameter and / or the reference sample parameter are preferably provided in a single storage space. That is, all reference parameters, i.e. the reference data parameter and the reference operating parameter and / or the reference sample parameter, are preferably stored in a single storage space, for instance in a local storage of the measurement system and / or on a cloud-based database, e.g. in a Digital Twin.

[0019] Likewise, the data parameter and the operating parameter and / or the sample parameter can be provided in a single storage space.

[0020] To this end it is particularly preferred that these parameters and these reference parameters are all provided in a common, single storage space.

[0021] The data analysis preferably comprises a statistical analysis. The data parameter preferably is a statistical parameter derived from the measurement data. The reference data parameter preferably is a statistical parameter derived from the reference measurement data. Hence, the data analysis preferably comprises a statistical analysis, for instance the application of at least one statistical algorithm to the measurement data and the reference measurement data, respectively.

[0022] Conceivable statistical parameters are an average, a standard deviation, a median, or a kth- percentile such as the 10thor 90thpercentile, etc. being derived from the measurement data and the reference measurement data, respectively.

[0023] The data parameter being associated with the reference data parameter preferably means that the data parameter is compared to the reference data parameter. Thus, for example, the data parameter can be a statistical parameter such as a median derived from the measurement data that is compared with a reference statistical parameter such as the median derived from the reference measurement data, and wherein the quality of the measurement data is then assessed based on the outcome of said comparison.

[0024] A matching association preferably exists in the case of identical statistical parameters determined for the data parameter and the reference parameter, for instance the median derived from the measurement data corresponding to the median derived from the reference measurement data.

[0025] However, it is likewise conceivable that the reference data parameter is a threshold against which the data parameter is compared during the association and wherein a matching or mismatching association is determined if the data parameter is below or above the threshold, for instance. For example, a matching association could be determined if the median derived from the measurement data is below or above the median derived from the reference measurement data.

[0026] The threshold can be a threshold value such as a static fixed value, or a dynamic value derived from reference measurement.

[0027] The association preferably takes into account tolerances, for instance measurement tolerances such as noise. That is, a comparison between the data parameter and the reference data parameter preferably considers tolerances such as measurement tolerances.

[0028] An outcome of the association, i.e. a matching or a mismatching, preferably is a Boolean such as true or false.

[0029] The operating parameter and the reference operating parameter are preferably associated with at least one of: a firmware version of the measurement system, a name of the measurement system, a serial number of the measurement system, a data acquisition software version of the measurement system, a hardware component of the measurement system, a measurement configuration of the measurement system, a tuning variable of the measurement system, a readback value of the measurement systems, or a data acquisition method being performed by the measurement system when carrying out the measurement on the sample and the reference measurement on the reference sample, respectively.

[0030] That is, examples of conceivable (reference) operating parameters are a firmware version of the measurement system, i.e. a collection of software numbers for firmware parts. Firmware is a type of machine-level software permanently embedded into a read-only memory of the measurement system. It is configured to provide instructions on how the measurement system is supposed to operate. In simpler terms, firmware gives low-level control over the hardware of the measurement system, making it the “software of hardware”. Examples of firmware versions of the measurement system are: I4IP-12.282.12.277; IPPT- 12.282.12.277; IPEX-12.282.12.277; FXM3-0.0.1 .6; MXMC-0.0.4.6; MXIF-0.0.2.0; MXI2- NOT_PRESENT; MXRF-0.0.1 .1 ; RFXS-0.1 .3.0; RFXD-NOT_PRESENT.

[0031] Another example of conceivable (reference) operating parameters are the name of the measurement system, for instance timsTOF fleX MALDI-2, i.e. a matrix-assisted laser desorption / ionization imaging measurement system being combined with laser-induced post-ionization.

[0032] Another example of conceivable (reference) operating parameters are a serial number of the measurement system, e.g. 1877407.00678.

[0033] Another example of conceivable (reference) operating parameters are a data acquisition software version of the measurement system, e.g. 4.1.12.

[0034] Another example of conceivable (reference) operating parameters is a hardware component such as a structural or physical part of the measurement system.

[0035] Examples of hardware components of the measurement system are a cartridge or a pump in the event of the measurement system comprising, for instance, a mass spectrometer. The (reference) operating parameter is preferably at least one numerical value and / or at least one alphanumeric character that represents the hardware component such as a type or serial number of the cartridge or the pump.

[0036] Another example of conceivable (reference) operating parameters is a measurement configuration that preferably reflects a matching of the measurement system with a known standard or reference and / or reflects at least one of an accuracy, deviation, adjustment, or uncertainty of the measurement system.

[0037] Examples of measurement configurations of the measurement system are results of a calibration of the measurement system. For instance, in the event of the measurement system comprising a mass spectrometer conceivable measurement configurations are results of a m / z calibration or of an ion mobility calibration or of a digitizer calibration.

[0038] An m / z calibration preferably provides as a calibration result the coefficients for a transformation between time-of-flight and m / z value. The calculation is preferably performed based on a measurement with a well-known substance.

[0039] An ion mobility preferably provides as a calibration result the coefficients for a transformation between countervoltage and mobility value. The calculation is preferably performed based on a measurement with a well-known substance.

[0040] A digitizer calibration preferably provides as a calibration result an adjustment of a baseline of the measurement system, in particular of a component thereof, to zero by measuring the offset in Volts. In other words, the digitizer calibration preferably determines a digital noise and an offset of the measurement system, in particular of a component thereof.

[0041] The (reference) operating parameter is preferably at least one numerical value and / or at least one alphanumeric character that represents the measurement configuration such as a type or a coefficient or a score of a m / z calibration or a type or a coefficient or a score of an ion mobility calibration or an offset or a noise suppression threshold of a digitizer calibration of the measurement system.

[0042] Another example of conceivable (reference) operating parameters is a tuning variable that is preferably configured to tune at least one component, in particular at least one hardware component of the measurement system.

[0043] Examples of tuning variables of the measurement system are for instance a TOF voltage or a detector voltage in the event the measurement system comprising a mass spectrometer. The (reference) operating parameter is preferably at least one numerical value that represents the tuning variable such a numerical value corresponding to the TOF voltage or to the detector voltage of the measurement system.

[0044] Another example of conceivable (reference) operating parameters is a readback value that is preferably detectable by at least one component, in particular by at least one hardware component of the measurement system.

[0045] Examples of readback values of the measurement system are a temperature within the measurement system, a power consumption of a turbo pump, or a vacuum pressure within the measurement system.

[0046] The (reference) operating parameter is preferably at least one numerical value and / or at least one alphanumeric character that represents the readback value such a numerical value corresponding to the TOF voltage or to the detector voltage or to the vacuum pressure of the measurement system.

[0047] Another example of conceivable (reference) operating parameters are a data acquisition method being performed by the measurement system when carrying out the measurement on the sample such as an acquisition mode, a polarity, a mass range and / or a name of the data acquisition method, for instance AmberGen-HiPLEX-IHC-20um.m. “AmberGen- HiPLEX-IHC-20um.m” is the name of a default acquisition method, which users get from Bruker for the respective mass spectrometer. They can use the method when measuring this type of data.

[0048] It should be noted that these examples are not exhaustive, but many other (reference) operating parameters are conceivable. For instance, another conceivable reference operating parameter could correspond to a scheduled date for cleaning the measurement system or a time duration between a last cleaning and a new cleaning of the measurement system and the operating parameter could be the actual date of cleaning the measurement system or the actual time duration between the actual last cleaning and the actual new cleaning of the measurement system.

[0049] Hence, the operating parameter and the reference operating parameter preferably comprises or consists of one or more alphanumeric characters and / or numerical values and / or numbers that are associated with an operation of the measurement system. An example of a number associated with the measurement system could be a voltage of a detector of the measurement system, etc., see explanations and examples made above.

[0050] The reference operating parameter is preferably defined and / or provided by the manufacturer of the measurement system.

[0051] The operating parameter is preferably determined by the measurement system. That is, the measurement system is preferably configured to read-out the operating parameter.

[0052] The operating parameter being associated with the reference operating parameter preferably means that the operating parameter is compared to the reference operating parameter. In particular, the alphanumeric characters of the operating parameter and the alphanumeric characters of the reference operating parameter are preferably compared with each other, wherein a matching association exists in the case of identical alphanumeric characters and a mismatching association exists in the case of differing alphanumeric characters. Additionally or alternatively, in the event of the operating parameter being at least one numerical value and / or number, the reference operating parameter can be a threshold such as a threshold value that preferably takes into account tolerances.

[0053] The sample parameter is preferably associated with a type of the sample and / or a preparation of the sample such as a sample preparation protocol according to which the sample has been prepared before being measured by the measurement system. The reference sample parameter is preferably associated with a type of the reference sample and / or a preparation of the reference sample such as a reference sample preparation protocol according to which the reference sample has been prepared before being measured by the measurement system.

[0054] The sample and the reference sample preferably are the same, i.e. of a same type. For instance, the sample and the reference sample preferably are from the same biological sample, such as a same tissue type, from a same preservation state, etc. Examples for tissue types are brain, lung, tonsil, etc. Examples of a preservation state are FFPE (formalin-fixed, paraffin-embedded), FF (fresh frozen), etc.

[0055] The (reference) sample being associated with a (reference) sample preparation preferably is an association of one or more steps of a sample preparation protocol. That is, the (reference) sample is preferably prepared according to a pre-determined (reference) sample preparation protocol. Conceivable steps of a (reference) sample preparation protocol and thus conceivable (reference) sample parameters are a type and / or conditions of (reference) sample collection, a type and / or conditions of (reference) sample enrichment, etc. For instance, the reference sample parameters could be particular types of chemicals and / or concentrations of chemicals and / or purity grades of chemicals and / or mixing ratios of chemicals and / or reaction conditions such as time parameters, e.g. incubation duration and / or shaking durations, temperature, light conditions such as illumination duration or light sensitivity, quantities such as working volumes, equipment parameters such as container types, shaking speed, filter size, centrifugation force, etc., resulting in a proper sample preparation of the reference sample. These reference sample parameters are thus preferably also used as the sample parameters for preparing the actual sample properly as well. That is, it is preferred that the actual sample is prepared using these particular chemicals and / or reaction conditions, for instance.

[0056] In more general terms, the (reference) sample parameters preferably relate to information necessary to reproduce the preparation of the (reference) sample that is ultimately subjected to analysis by the measurement system.

[0057] Hence, the sample parameter and the reference sample parameter in each case preferably comprise or consist of one or more numerical values and / or alphanumeric characters that are associated with the preparation of the sample and of the reference sample and / or with the type of the sample and the reference sample, respectively. For instance, the (reference) sample parameter could be the name of a sample preparation protocol or the name of the tissue type.

[0058] The sample parameter being associated with the reference sample parameter preferably means that the sample parameter is compared to the reference sample parameter. In particular, the numerical values and / or alphanumeric characters of the sample parameter and the numerical values and / or the alphanumeric characters of the reference sample parameter are preferably compared with each other, wherein a matching association exists in the case of identical numerical values and / or alphanumeric characters and a mismatching association exists in the case of differing numerical values and / or alphanumeric characters.

[0059] A mismatching association between the operating parameter and the reference operating parameter preferably is indicative of at least one of: an inferior condition of the measurement system, a wrong data acquisition method, an outdated firmware version of the measurement system, an outdated data acquisition software version of the measurement system, or a different handling of the measurement system.

[0060] An example of an inferior condition of the measurement system is a contaminated measurement system. A desired or targeted condition of the measurement system in this case would be an uncontaminated or clean measurement system, for instance. In this case, the association or comparison between the operating parameter and the reference operating could be a difference in date or time duration regarding the scheduled cleaning and the actual cleaning of the measurement system.

[0061] An example of a wrong data acquisition method is a data acquisition method deviating from the data acquisition method that has been used for acquiring the reference measurement data. This acquisition method needs to be suitable for the sample.

[0062] An example of a different handling of the measurement system is a measurement system where the time duration between the last calibration and the measurement is too large etc.

[0063] A mismatching association between the sample parameter and the reference sample parameter is preferably indicative of an inferior sample preparation and / or of an incorrect sample type being used.

[0064] Examples of an inferior sample preparation are chemicals that have expired, chemicals that have been stored incorrectly, one or more sample preparation steps that deviate from the sample preparation protocol, pipetting errors that have incurred in the sample preparation, etc. Hence, a desired or target sample preparation could be samples that have been prepared with high quality chemicals and under correct reaction conditions such as time, concentration, temperature, illumination, etc.

[0065] Examples of an incorrect sample type being used are for instance brain tissue being used as the sample whereas the reference sample was lung tissue, a FF sample being used whereas the reference sample was FFPE, etc.

[0066] A mismatching association between the data parameter and the reference data parameter preferably is indicative of at least one of: an inferior condition of the measurement system, a wrong data acquisition method, an outdated firmware version of the measurement system, an outdated data acquisition software version of the measurement system, a different handling of the measurement system, or an inferior sample preparation.

[0067] The operating parameter is preferably associated with the reference operating parameter only in the event of a mismatching association between the data parameter and the reference data parameter.

[0068] Additionally or alternatively, the sample parameter is preferably associated with the reference sample parameter in the event of a matching association of the data parameter and the reference data parameter as well as in the event of a mismatching association.

[0069] That is, it is preferred that an association of the operating parameter with the reference operating parameter is only performed in the event of a mismatching association of the data parameter and the reference data parameter, whereas an association of the sample parameter with the reference sample parameter is performed in any case.

[0070] At least one output signal is preferably generated in the event of a mismatching association between the data parameter and the reference data parameter. Additionally or alternatively, at least one output signal is preferably generated in the event of a mismatching association between the operating parameter and the reference operating parameter. Additionally or alternatively, at least one output signal is preferably generated in the event of a mismatching association between the sample parameter and the reference sample parameter. The output signal is preferably communicated to a user of the measurement system. Additionally or alternatively, the output signal preferably comprises indications of the mismatching association such as the indication of the mismatching data parameter and / or the mismatching operating parameter and / or the mismatching sample parameter. Additionally or alternatively, the output signal is preferably configured as an input signal to the measurement system and configured to control an operation of the measurement system.

[0071] That is, in the event of a mismatching association between the parameter(s) and the reference parameter(s), the method preferably generates an output signal which, for instance, is communicated to the user and thus informs the user on the mismatching. For instance, the output signal can be configured as an automatic notification or warning to the user. Additionally or alternatively, the output signal could comprise a recommendation or a proposed solution to the user. For example, in the event of a mismatching operating parameter relating to an outdated data acquisition software version or firmware version of the measurement system, the output signal could notify the user that the data acquisition software version or the firmware version has changed. In the event of mismatching data parameters such as a high mass deviation between the data parameter and the reference data parameter (see also further below), the output signal could notify the user about a too long duration since the last calibration and that could recommend performing a calibration of the measurement system or checking the calibration parameters. In the event of mismatching data parameters such as low intensities of selected mass tags of the data parameter as compared to the reference data parameter, the output signal could recommend checking the quality for the mass tag chemicals being used. In the event of mismatching data parameters such as low Total Ion Chromatogram (TIC) deviations between the data parameter and the reference data parameter and low intensities of all mass tags, the output signal could recommend checking the last date of cleaning the measurement system. Etc.

[0072] It should be noted that such an output signal can likewise be generated and communicated in the event of a matching association between the parameter(s) and the reference parameter(s).

[0073] The measurement data and the reference measurement data preferably originate from a measurement system comprising a mass spectrometer, optionally connected to an LC system. The measurement data and the reference measurement data preferably correspond to mass spectral data comprising m / z-values and associated intensities.

[0074] That is, the method according to the invention is preferably used to monitor a quality of data obtained from a mass spectrometer, wherein the measurement data being monitored preferably is at least one mass spectrum.

[0075] To this end various mass spectrometers and mass spectrometry techniques are conceivable. For instance, the mass spectrometer can comprise a time-of-flight (TOF) analyzer, for instance with an axial TOF arrangement, or a trapped ion mobility time-of-flight (timsTOF) mass spectrometry. For example, the measurement data can be obtained from a mass spectrometry imaging (MSI), in particular using MALDI (matrix assisted laser desorption / ionization) as ionisation technique and a TOF analyser such as the timsTOF. For example, the measurement system could be configured for targeted MALDI imaging or general MALDI imaging. Targeted MALDI imaging means, that only m / z-values are considered, which have been defined before performing the measurement. Those can be e.g. the m / z-values relating to certain lipids, metabolites, or peptides. General MALDI imaging means here the opposite of targeted imaging, so without focus on mass tags. However, it should be noted that various other mass spectrometers and mass spectrometry techniques or ionization techniques such as ESI (electrospray ionization), etc., are likewise conceivable and are well-known in the art. Since the measurements and the reference measurements are preferably carried out by the same measurement system, it is thus preferred that the reference measurement data correspond to mass spectral data comprising m / z-values and associated intensities of the reference sample as well.

[0076] It is particularly preferred that the (reference) measurement data correspond to several (reference) mass spectra and that the (reference) data parameters are derived from either all (reference) mass spectra or from a selection thereof, i.e. from targeted m / z-values.

[0077] The data analysis of the measurement data and the reference measurement data in each case preferably comprises the derivation of a Total Ion Chromatogram (TIC) by summing up all intensities of all m / z-signals of each mass spectrum as a function of the recording time. The data parameter and the reference data parameter preferably is in each case a statistical parameter derived from the respective TIC.

[0078] The data parameter being a statistical parameter preferably is at least one of: an average, a standard deviation, a k-th percentile such as the 10thor 90thpercentile, a minimum value, or a maximum value derived from said TIC.

[0079] Likewise, the reference data parameter is preferably a statistical parameter being derived from a Total Ion Chromatogram (TIC) obtained by summing up all intensities of all m / z- signals of each mass spectrum of the reference measurement data as a function of the recording time. The reference data parameter being a statistical parameter preferably is at least one of: an average, a standard deviation, a k-th percentile such as the 10thor 90thpercentile, a minimum value, or a maximum value derived from said TIC.

[0080] To this end it is particularly preferred that said reference data parameters derived from the TIC of the reference measurement data are a threshold value, against which the data parameter is compared during the association of the data parameter with the reference data parameter.

[0081] As mentioned earlier, the threshold value can be a static fixed value, or a dynamic value derived from the reference measurement.

[0082] The minimum value is understood as the lowest TIC value detected in the data analysis of the (reference) measurement data. The maximum value is understood as the highest TIC value detected in the data analysis of the (reference) measurement data.

[0083] The data analysis of the measurement data and the reference measurement data in each case preferably comprises the determination of a mass deviation for targeted m / z values. The data parameter and the reference data parameter preferably is in each case a statistical parameter derived from the respective mass deviation.

[0084] That is, the data analysis of the (reference) measurement data preferably comprises the determination of a mass deviation for targeted m / z-values, and wherein the (reference) data parameter is a statistical parameter derived from said mass deviation.

[0085] The mass deviation can be defined or determined in several ways:

[0086] Mass deviation [ppm] = (m / z measured - m / z theoretical) * 106 / m / z theoretical

[0087] Mass deviation [Da] = m / z measured - m / z theoretical

[0088] Mass deviation [mDa] = (m / z measured - m / z theoretical) * 1000

[0089] In the above equations, "rn / zmeasured" refers to the m / z-values determined from the (reference) measurement data.

[0090] "m / ztheoreticai" refers to the theoretical m / z-values that are calculated from the theoretical masses of the atoms constituting the chemical structure of the sample.

[0091] The (reference) data parameter being a statistical parameter derived from the mass deviation preferably is at least one of: an average, a standard deviation, a median, a k-th percentile such as the 10thor 90thpercentile, a minimum value, or a maximum value derived from said mass deviation and / or one or more fitted values obtained from fitting one or more parametrized curves or parametrized lines to the mass deviation. The fitted values are preferably derived from a linear regression being performed on the mass deviation. In particular, the fitted values preferably are at least one of a first and last value, a slope, a distance or a correlation of a linear regression.

[0092] That is, mass deviations of different mass spectra can drift over time, which drift can be approximated as being a linear drift. Therefore, the (reference) data parameters can be derived from fitting a linear regression line to the mass deviations of the different mass spectra, whereby (reference) data parameters being regression coefficients are obtained.

[0093] For instance, if one or more regression coefficients associated with said mass drift are outside a reference data parameter in the form of a mass tolerance, a mismatching association is determined. The mass tolerance is understood as the maximal difference between the measured mass, i.e. m / zmeasured, and the theoretical mass, i.e. m / ztheoreticai.

[0094] That is, rn / Zmeasured can deviate from m / ztheoreticai. If, for instance, the deviation is below the mass tolerance, m / zmeasured can be considered as a statistical parameter in the statistical analysis. If not it is ignored. Therefore, the mass tolerance is preferably used as a filter and is part of the statistical analysis.

[0095] To this end it is particularly preferred to determine mass deviations for each selected mass tag.

[0096] It is furthermore preferred that the statistical parameter is derived from the mass deviation only in the event that the mass deviation is below a predetermined or preset mass deviation tolerance. In other words, it is preferred that only signals with a mass deviation below a mass deviation tolerance are used for determining the statistical parameter. That is, a data filter is preferably applied.

[0097] The data analysis of the measurement data and the reference measurement data in each case preferably comprises the determination of intensities for targeted m / z values. The data parameter and the reference data parameter preferably is in each case a statistical parameter derived from the respective intensities.

[0098] That is, the data analysis of the (reference) measurement parameter preferably comprises the determination of the intensities for targeted m / z-values, and wherein the (reference) data parameter is a statistical parameter derived from said intensities.

[0099] That is, for each mass tag in each (reference) mass spectrum it is preferred that the intensities are subjected to the data analysis in order to determine the (reference) data parameters being statistical parameters.

[0100] The (reference) data parameter being a statistical parameter preferably is at least one of: an average, a standard deviation, a median, a kth-percentile such as the 10thor 90thpercentile, or a percentage of mass spectra containing a target mass tag, i.e. a target m / z- value. For example, a percentage of mass spectra containing a target m / z-value can be determined by counting a number of (reference) mass spectra, where the target mass tag has been observed, i.e. the signal was detected, and divide it by the number of all mass spectra.

[0101] It is preferred that the statistical parameter is derived from the intensities only in the event that the intensities are above a predetermined or preset intensity threshold. The intensity threshold can be set by a user of the measurement system. That is, a data filter is preferably applied in this case as well.

[0102] In the following, some conceivable examples for mismatching data parameters and reference data parameters are given, that enables the recognition of bad data quality according to the invention:

[0103] Different absolute values between the data parameters and the reference data parameters such as:

[0104] - TIC - data parameter in the form of standard deviation below threshold defined by the reference data parameter;

[0105] - Mass deviations for m / z values - data parameters in the form of 10th and 90th percentile are above threshold defined by the reference parameter;

[0106] - Mass deviations for m / z values - data parameters in the form of minimum and maximum value are above threshold defined by the reference data parameter;

[0107] - Mass deviations for m / z values - data parameters in the form of 1st and last value of fitted linear regression line are outside set tolerance defined by the reference data parameter;

[0108] - Intensities for m / z values - percentage of spectra containing m / z value

[0109] Moreover, in the following some conceivable examples for an output signal in the event of a mismatching association are given:

[0110] Example 1 :

[0111] Matching association between operating parameter(s) and reference operating parameter(s);

[0112] Mismatching association between data parameter and reference data parameter: TIC deviation is low, e.g. in the event of a standard deviation of the TIC being below a threshold; When the standard deviation for the TIC is low, it indicates that all measured intensities are low.

[0113] Mismatching association triggers notification: "Hardware and sample preparation needs to be checked."

[0114] Example 2:

[0115] Matching association between operating parameter(s) and reference operating parameter(s);

[0116] Mismatching association between data parameter and reference data parameter: Intensity for some but not all m / z values are low, e.g. in the event of the measured intensities being below a threshold.

[0117] Mismatching association triggers notification: "Some antibodies need to be checked."

[0118] Example 3:

[0119] Mismatching association between operating parameter(s) and reference operating parameter(s);

[0120] Mismatching association between data parameter and reference data parameter: Intensity for all m / z values are low, e.g. in the event of the measured intensities being below a threshold.

[0121] Mismatching associations trigger notification: "Operating parameters need to be checked." Or “new reference measurement is required, e.g. when software version or firmware version have been changed.”

[0122] Hence, said combination of the QC results provided by the data parameters and the operating and / or the sample parameters allow to list differences between the last QC with good result and the one with the bad result, e.g. because of changed software versions, calibration errors, tuning of the hardware device.

[0123] Possible causes can then be reported to the user. As a consequence, the user can identify bad quality data at an early state, e.g. before said data is further processed.

[0124] In another aspect, a computer program product comprising instructions is provided which, when carried out by a computer, causes the computer to carry out the method as described above.

[0125] In another aspect, a measurement system for analysing a sample, preferably a mass spectrometer, is provided. The measurement system comprises a processor comprising instructions which, when carried out by the processor, causes the processor to carry out the method as described above.

[0126] Any explanations made herein regarding the method preferably likewise apply to the computer program product configured to carry out the method and to the measurement system comprising the processor configured to carry out the method and vice versa.

[0127] BRIEF DESCRIPTION OF THE DRAWINGS

[0128] Preferred embodiments of the invention are described in the following with reference to the drawings, which are for the purpose of illustrating the present preferred embodiments of the invention and not for the purpose of limiting the same. In the drawings,

[0129] Fig. 1 shows a schematic of the method for monitoring a quality of data according to the invention;

[0130] Fig. 2 shows an example of data quality being monitored by the method according to the invention.

[0131] DESCRIPTION OF PREFERRED EMBODIMENTS

[0132] Aspects of the method for monitoring a quality of data according to the invention are now illustrated with reference to the figures.

[0133] That is, and as schematically indicated in figure 1 , the method according to the invention preferably starts with receiving measurement data of a sample from a measurement system that has carried out at least one measurement on the sample. Said measurement data is subjected to data analysis, whereby at least one data parameter associated with the measurement data is obtained. Moreover, reference measurement data of a reference sample that has been acquired by said measurement system is received and subjected to the data analysis as well, whereby at least one reference data parameter associated with the reference measurement data is obtained.

[0134] In addition, at least one operating parameter being associated with an operation of the measurement system and at least one reference operating parameter being associated with a reference operation of the measurement system are received and associated with one another. Additionally or alternatively, at least one sample parameter being associated with the sample and at least one reference sample parameter being associated with the reference sample are received and associated with one another.

[0135] The quality of the measurement data is monitored based on the association of the data parameter and the reference data parameter and in addition based on the association of i) the operating parameter with the reference operating parameter and / or ii) the association of the sample parameter with the reference sample parameter.

[0136] In the event of a mismatching association, at least one output signal is generated and communicated to a user of the measurement system or input to the measurement system for controlling the measurement system.

[0137] Figure 2 illustrates an example. Here, the sample is a tissue, in particular a human colorectal cancer tissue.

[0138] The sample is prepared according to a protocol developed by AmberGen as described in Yagnik, Liu, Rothschild, Lim, (2021). Highly Multiplexed Immunohistochemical MALDI-MS Imaging of Biomarkers in Tissue. Anal. Chem. 32, 977-988, doi: 10.1021 / jasms.0c00473. The workflow includes tissue preparation with antigen retrieval and staining with antibodies of choice with photocleavable mass-tags as described in Yagnik et al. as well. The result of the sample preparation are slides which can be mounted on a slide adapter for using in the measurement system. In particular, the sample was prepared on IntelliSlides as commercially sold by Bruker and introduced into a measurement system here in the form of a neofleX MS instrument, an axial TOF instrument.

[0139] In particular, the measurement system carried out measurements on the sample using the following data acquisition:

[0140] • MS data, positive mode

[0141] • Default method was used: positive reflector mode, m / z range 800 - 2200, laser frequency 10 kHz

[0142] • MALDI HiPLEX-IHC data

[0143] • flexControl 5.0 software

[0144] • DataAnalysis 6.0 software for creating this plot

[0145] The thus received measurement data was then subjected to a data analysis according to the invention, wherein data parameters were derived from the measurement data. The outcome of said data analysis is depicted in figure 2, which shows a plot comprising a selection of 4 MS spectra out of > 100.000 and 3 mass tags.

[0146] From each single MS spectrum, the Total Ion Chromatogram (TIC) value is extracted which is used in the data analysis for a statistical calculation, whereby data parameters in the form of statistical parameters were derived.

[0147] Moreover, for each mass tag, the data analysis performed statistics on intensity and mass accuracy, whereby further data parameters being statistical parameters were derived.

[0148] The statistical parameters which have been determined (the reference statistical parameters are thresholds that are indicated in brackets):

[0149] TIC

[0150] • an average (> 5000) (matching: 10000, mismatching: 2000)

[0151] • a standard deviation (> 2000) (matching: 3000, mismatching: 500)

[0152] • a k-th percentile such as the 10thor 90thpercentile (10th: > 2500, 90th: this value was only tracked) (10thmatching: 3000, 10thmismatching: 2000)

[0153] • a minimum value (this value was only tracked)

[0154] • a maximum value (this value was only tracked)

[0155] Mass deviation for each target m / z value:

[0156] • an average (< 100 ppm) (matching: 10, mismatching: 150)

[0157] • a standard deviation (< 100 ppm) (matching: 70, mismatching:120)

[0158] • a median (< 100 ppm) (matching: 15, mismatching: 160)

[0159] • a k-th percentile such as the 10thor 90thpercentile (10th: this value was only tracked, 90th: < 80 ppm) (90thmatching: 60, mismatching: 90)

[0160] • a minimum value (this value was only tracked)

[0161] • or a maximum value (this value was only tracked)

[0162] • fitted mass deviation for first and / or last spectrum outside the given mass tolerance + 400 ppm (matching: -200 / 200 ppm, mismatching: -450 / 350 ppm)

[0163] Intensities for each target m / z value:

[0164] • an average (> 3000) (matching: 5000, mismatching: 2500)

[0165] • a standard deviation (< 1000) (matching: 600, mismatching: 2000)

[0166] • a median (< 3000) (matching: 4500, mismatching: 2000)

[0167] • a kth-percentile such as the 10thor 90thpercentile (10th: this value was only tracked, 90th: > 2000) (90thmatching: 3000, mismatching: 1000) • a percentage of mass spectra containing a target mass tag (> 10 %) (matching: 25, mismatching: 7)

[0168] Although not shown in the plot, also operating parameters were extracted from the measurement system and associated with reference operating parameters. In particular, the flexControl 5.0 software operating parameter was associated with the reference operating parameter being flexControl 4.0 software, which resulted in mismatching.

[0169] Moreover, the sample parameter "Sample preparation protocol “SOP version 1” from 2024- 01-04 for tissue type “brain”" was associated with reference sample parameters.

[0170] Example matching:

[0171] - Sample preparation protocol “SOP versionl” from 2024-01-04 for tissue type “brain” Examples mismatching:

[0172] - Sample preparation protocol “SOP version2” from 2024-01-04 for tissue type “brain”

[0173] - Sample preparation protocol “SOP versionl” from 2024-04-08 for tissue type “brain”

[0174] - Sample preparation protocol “SOP versionl” from 2024-01-04 for tissue type “liver”

Claims

23CLAIMS1. A computer-implemented method for longitudinally monitoring a quality of data obtained from a measurement system configured to analyse a sample, wherein the method comprises the steps of:- Receiving measurement data of a sample from a measurement system that has carried out at least one measurement on the sample,- Subjecting the measurement data to data analysis, whereby at least one data parameter associated with the measurement data is obtained, and- Associating the data parameter with at least one reference data parameter associated with reference measurement data received from at least one reference measurement that has been carried out for at least one refence sample by the measurement system, and- Monitoring a quality of the measurement data based on the association of the data parameter with the reference data parameter, characterised in that the method further comprises the step of:- Receiving at least one operating parameter being associated with an operation of the measurement system and associating the operating parameter with at least one reference operating parameter being associated with a reference operation of the measurement system, and wherein the monitoring of the quality of the measurement data is further based on the association of the operating parameter with the reference operating parameter, and / orReceiving at least one sample parameter being associated with the sample and associating the sample parameter with at least one reference sample parameter being associated with the reference sample, and wherein the monitoring of the quality of the measurement data is further based on the association of the sample parameter with the reference sample parameter.

2. The method according to claim 1 , wherein the data parameter being associated with the reference data parameter corresponds to a comparison of the data parameter to the reference data parameter, and / or wherein the operating parameter being associated with the reference operating parameter corresponds to a comparison of the operating parameter to the referenceoperating parameter, and / or wherein sample parameter being associated with the reference sample parameter corresponds to a comparison of the sample parameter to the reference sample parameter.

3. The method according to any one of the preceding claims, wherein the measurements and the reference measurements are carried out by the same measurement system, and / or wherein the operating parameter and the reference operating parameter are of a same type.

4. The method according to any one of the preceding claims, wherein the data analysis comprises a statistical analysis, wherein the data parameter is a statistical parameter derived from the measurement data, and wherein the reference data parameter is a statistical parameter derived from the reference measurement data.

5. The method according to any one of the preceding claims, wherein a matching association between the data parameter and the reference data parameter is determined if the data parameter and the reference data parameter are identical to one another, preferably while taking into account tolerances such as measurement tolerances.

6. The method according to any one of the preceding claims, wherein the reference data parameter is a threshold against which the data parameter is compared during the association, and wherein a matching or mismatching association is determined if the data parameter is below or above the threshold, and wherein the threshold preferably is a threshold value such as a static fixed value or a dynamic value derived from the reference measurement.

7. The method according to any one of the preceding claims, wherein the operating parameter and the reference operating parameter are associated with at least one of: a firmware version of the measurement system, a name of the measurement system, a serial number of the measurement system, a data acquisition software version of the measurement system, a hardware component of the measurement system, a measurement configuration of the measurement system, a tuning variable of the measurement system, a readback value of the measurement systems, or a data acquisition method being performedby the measurement system when carrying out the measurement on the sample and the reference measurement on the reference sample, respectively.

8. The method according to any one of the preceding claims, wherein the operating parameter and the reference operating parameter comprise or consist of one or more numerical values and / or alphanumeric characters, wherein the numerical values and / or the alphanumeric characters of the operating parameter and the numerical values and / or alphanumeric characters of the reference operating parameter are compared with each other, and wherein a matching association is determined in the case of identical alphanumeric characters and / or identical numerical values and a mismatching association is determined in the case of differing alphanumeric characters and / or numerical values.

9. The method according to any one of the preceding claims, wherein the sample parameter is associated with a sample preparation protocol according to which the sample has been prepared before being measured by the measurement system and / or a type of the sample, and / or wherein the reference sample parameter is associated with a reference sample preparation protocol according to which the reference sample has been prepared before being measured the measurement system and / or a type of the reference sample.

10. The method according to any one of the preceding claims, wherein the sample parameter and the reference sample parameter comprise or consist of one or more numerical values and / or alphanumeric characters, wherein the numerical values and / or the alphanumeric characters of the sample parameter and the numerical values and / or alphanumeric characters of the reference sample parameter are compared with each other, and wherein a matching association is determined in the case of identical alphanumeric characters and / or identical numerical values and a mismatching association is determined in the case of differing alphanumeric characters and / or numerical values.11 . The method according to any one of the preceding claims, wherein a mismatching association between the operating parameter and the reference operating parameter is indicative of at least one of: an inferior condition of the measurement system, a wrong data acquisition method, an outdated firmware version of the measurement system, an outdated data acquisition software version of the measurement system, or a different handling of the26 measurement system.

12. The method according to any one of the preceding claims, wherein a mismatching association between the sample parameter and the reference sample parameter is indicative of an inferior sample preparation and / or of an incorrect sample type being used.

13. The method according to any one of the preceding claims, wherein a mismatching association between the data parameter and the reference data parameter is indicative of at least one of: an inferior sample preparation, an inferior condition of the measurement system, a wrong data acquisition method, an outdated firmware version of the measurement system, an outdated data acquisition software version of the measurement system, a different handling of the measurement system.

14. The method according to any one of the preceding claims, wherein the operating parameter is associated with the reference operating parameter only in the event of a mismatching association between the data parameter and the reference data parameter, and / or wherein the sample parameter is associated with the reference sample parameter in the event of a matching association and a mismatching association between the data parameter and the reference data parameter.

15. The method according to any one of the preceding claims, wherein at least one output signal is generated in the event of a mismatching association between the data parameter and the reference data parameter and / or between the operating parameter and the reference operating parameter and / or the sample parameter and the reference sample parameter, and wherein at least one of:- the output signal is communicated to a user of the measurement system,- the output signal comprises indications of the mismatching association such as the indication of the mismatching data parameter and / or the mismatching operating parameter and / or the mismatching sample parameter, or- the output signal is configured as an input signal to the measurement system and configured to control an operation of the measurement system.

16. The method according to any one of the preceding claims, wherein the measurement data and the reference measurement data originate from a measurement system comprising a mass spectrometer, and wherein the measurement data and the27 reference measurement data correspond to mass spectral data comprising m / z-values and associated intensities.

17. The method according to claim 16, wherein the data analysis of the measurement data and the reference measurement data in each case comprises the derivation of a Total Ion Chromatogram (TIC) by summing up all intensities of all m / z-signals of each mass spectrum as a function of the recording time, and wherein the data parameter and the reference data parameter is in each case a statistical parameter derived from the respective TIC.

18. The method according to claim 16 or 17, wherein the data analysis of the measurement data and the reference measurement data in each case comprises the determination of a mass deviation for targeted m / z values, and wherein the data parameter and the reference data parameter is in each case a statistical parameter derived from the respective mass deviation.

19. The method according to any one of claims 16 to 18, wherein the data analysis of the measurement data and the reference measurement data in each case comprises the determination of intensities for targeted m / z values, and wherein the data parameter and the reference data parameter is in each case a statistical parameter derived from the respective intensities.

20. A computer program product comprising instructions which, when carried out by a computer, causes the computer to carry out the method according to any one of the preceding claims.

21. A measurement system for analysing a sample, preferably a mass spectrometer, wherein the measurement system comprises a processor comprising instructions which, when carried out by the processor, causes the processor to carry out the method according to any one of claims 1 to 19.

Images