Selection of peak detection parameters for isotope-ratio mass spectrometry

Abstract

A method of determining isotope ratios for components of a sample is disclosed. The method comprises separating the sample into components in a chromatography column 204 and coupling the separated sample eluting from the chromatography column into an isotope ratio mass spectrometer 206 . The method further comprises obtaining a chromatogram of the sample 208. The chromatogram comprises a mass-spectrometer signal as a function of time, the mass spectrometer signal comprising a respective detector signal for each of a plurality of isotopes of interest. The method comprises selecting a set of peak detection parameters from a plurality of predetermined sets of peak detection parameters 209a, b, c. The parameters define peak slope and / or peak magnitude and / or peak extent thresholds for the obtained chromatogram and may be position dependent, for example define different respective thresholds for different positions in a chromatogram. The method also comprises automatically segmenting the chromatogram into peaks corresponding to the components using operations based on the selected peak detection parameters 210 and determining one or more isotope ratios for one or more of the components using the mass-spectrometer signal of each respective peak 212. Corresponding computer-readable media, systems and a computer implemented peak classifier method are also disclosed.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to methods for determining isotope ratios for chromatographically separated components of a sample using Isotope Ratio Mass Spectrometry. BACKGROUND

[0002] Chromatographic effluent, which may be liquid or gaseous in form, is routinely subjected to spectroscopic analysis, often mass spectrometry, in order to try and evaluate the eluting components contained in the effluent. Such spectroscopic analysis naturally depends on the accurate processing and separation of the eluting compounds of the chromatographic effluent.

[0003] There remains a need for improved chromatographic processing, e.g., to increase the ability to resolve, identify, and evaluate natural products or samples which are not fully baseline separated during chromatographic analysis. This is particularly pertinent in gas (GC) or liquid (LC) chromatography coupled to an isotope ratio mass spectrometer (IRMS) to determine isotope ratios, because the correct separation of chromatographically separated components influences the determination of isotope ratios. SUMMARY

[0004] Aspects of the disclosure are set out in the independent claims. Particular implementations of these aspects are set out in the dependent claims. Any subject matter contained herein that may not fall within the scope of the appended claims is considered as being useful for understanding the invention.

[0005] The subject matter disclosed herein pertains to improved methods for separating the components of a chromatogram of a sample to be evaluated, in particular for natural product samples or samples, which may not be fully baseline separated, that is in which the chromatography peaks from different components overlap and are not separated by a zero or background signal.

[0006] The methods according to the present invention may be applicable to liquid or gas chromatography and may be suitable to be carried out using any chromatography instrument coupled to an isotope ration mass spectrometer (IRMS). The chromatography instrument may comprise a gas chromatograph, a liquid chromatograph or any other chromatography system.

[0007] In one or more implementations, a method of determining isotope ratios for components of a sample comprises separating the sample into components in a chromatography column and coupling the separated sample eluding from the chromatography column into an isotope ratio mass-spectrometer. The method further comprises obtaining a chromatogram of the sample. The chromatogram comprises a mass-spectrometer signal as a function of time, the mass-spectrometer signal comprising a respective detector signal for each of a plurality of isotopes of interest. The method comprises selecting a set of peak detection parameters from a plurality of predetermined sets of peak detection parameters. The peak detection parameters may comprise peak slope and / or peak magnitude thresholds. Additionally or instead, the peak detection parameters may comprise one or more of: peak extent threshold(s); sample time step; resolution; pulse shape; and background signal magnitude. The peak detection parameters may be position dependent, for example define different respective thresholds for different positions in a chromatogram. The method also comprises automatically segmenting the chromatogram into peaks corresponding to the components using operations based on the selected peak detection parameters and determining one or more isotope ratios for one or more of the components using the mass-spectrometer signal of each respective peak.

[0008] It will be understood that the term “chromatogram” refers to a representation of a complex sample which has been run through a chromatography column in a chromatography instrument to separate out individual analytes or compounds. Separation may be based on their respective elution time, retention time, or time taken to travel through the chromatography instrument. In the present disclosure, separated components are fed into an I RMS, possibly after being broken down into constituents, for example one or more gases, and the chromatogram comprises a mass spectrometer as a function of time, which comprises a detector signal for each isotope of interest, for example several isotopes of CO2.

[0009] The predetermined set of peak detection parameters comprise a selection of parameters which can be used to define the presence and extent of a peak in the chromatogram. For example, parameters defining the thresholds of a peak slope may comprise values indicating a starting gradient of a curve or slope of a peak, a steepest gradient (or ‘top slope’) of a curve or slope of a peak, and / or an end gradient of the slope of a peak. The parameters may also define the number of slope data points used in the evaluation of slope gradients of a chromatogram. The parameters defining the magnitude or extent of the peak may comprise a maximum width of a peak (e.g., in seconds), a minimum height of a peak (e.g., an intensity reading of the sample), an area of a peak, a resolution between adjacent peaks, or any quantitative threshold value for characterising the magnitude or extent of a peak.

[0010] A set of peak detection parameters may also comprise a combination of peak detection parameters for a plurality of segments or time ranges within the chromatogram. For example, for a chromatogram spanning a period of 2000 seconds, a set of peak detection parameters may comprise a first selection of parameters for the first 1000 seconds, and a second selection of parameters for the second 1000 seconds.

[0011] In any case, the sets of peak detection parameters are predetermined. For example, the predetermined sets may have been determined prior to the separating of a sample into components in a chromatography column, for example as factory defaults or as saved preferences of a user. In some examples, the sets of peak detection parameters are already created, prior to running a given sample, by an operator as corresponding to, or being suitable for, various known variations, types, or classes of sample(s). Additionally, the predetermined sets of peak detection parameters may be saved in a database. For example, selecting a set of peak detection parameters may comprise accessing a database comprising the plurality of predetermined sets of peak detection parameters and selecting a suitable set therefrom. The predetermined sets of peak detection parameters which were saved in the database may also be used for training in machine learning methods, as discussed in more detail below.

[0012] Operations based on the set of peak detection parameters may comprise any operation for automatically locating peak segments in a chromatogram using the parameters described above, e.g., by identifying timestamps or regions of the chromatogram bounded by the threshold peak parameters. The chromatogram is segmented into components based on the identified peak regions and isotope ratios of interest are calculated for each peak region, for example by averaging or summing the ratios across the samples in each region. By automatically segmenting the peaks of the chromatogram, the extra time taken for otherwise tedious manual evaluation and re-evaluation of individual chromatograms is avoided and the potential for human error is minimised. Furthermore, by selecting peak detection parameters and automatically segmenting a chromatogram based thereon, less manual labour is needed to achieve reliable and reproducible isotope ratio results, particularly for the evaluation of complex samples which are (in principle) composed of the same components but vary with origin, manufacture, and / or treatment process.

[0013] In some examples, the method comprises presenting to a user a user interface which outputs respective selection options for the predetermined sets of peak detection parameters. In such examples, the method may comprise selecting the set of peak detection parameters based on a user input selecting one of the selection options. For example, the method may comprise displaying a selection panel with a plurality of options corresponding to various sets of peak detection parameters. Each set of peak detection parameters may correspond to a known sample type with known characteristics, such as composition, manufacture, treatment, or origin. An experienced user can therefore efficiently switch between and choose the optimal peak detection parameter set for certain chromatogram from the user interface. In such examples, a fast switch between different peak detection presets is enabled for efficient evaluation from a visual inspection of any or all of the plurality of predetermined sets of peak detection parameters.

[0014] In some alternative examples, the method comprises classifying the chromatogram of a sample as belonging to one of a plurality of classes, each class corresponding to a respective predetermined set of peak detection parameters. In such examples, the selecting of the set of peak detection parameters for the sample to be analysed is done by selecting the predetermined set of peak detection parameters associated with the class to which the chromatogram has been classified as belonging. In this way, classification of the chromatogram facilitates the selection of the set of peak detection parameters, which can be particularly helpful for less experienced users.

[0015] In one or more examples relating to classifying the chromatogram, the method comprises classifying the received chromatogram by determining a similarity score with a predetermined chromatogram pattern associated with a respective predetermined set of peak detection parameters for each of the predetermined sets of peak detection parameters and selecting the set of peak detection parameters based on the similarity scores.

[0016] For example, an archetype chromatogram, i.e., an archetype chromatographic pattern, may have known and / or preassigned peak detection parameters. In such examples, the method may comprise accessing a library of archetype chromatogram patterns with corresponding preassigned sets of peak detection parameters and using the library of patterns with corresponding preassigned sets to evaluate a similarity score or otherwise perform pattern matching of the acquired sample chromatogram with each of the archetype chromatogram patterns. The similarity score can indicate quality of the match with any or each of the archetype chromatogram patterns saved in the library and can be used to select a corresponding predefined set of parameters, for example that for which the similarity score with the corresponding archetype chromatogram was largest compared to the other chromatograms, also referred to as classification herein.

[0017] In some examples, in order to make the similarity score intensity agnostic, the archetype chromatogram patterns are normalised, and the similarity score is based on a similarity with the normalised chromatogram. In one example, the similarity score between two chromatograms is achieved by calculating a dot product or cosine similarity between vectors representing the shape of the two chromatograms. Alternatively, various mathematical methods, such as Frechet distance, or computer implemented shape-matching algorithms are known in the art and can be used to achieve a similarity score or otherwise match pattern to classify the chromatogram and select a predefined set of parameters. The similarity score may be based on the chromatogram with the full MS signal, including the detector signals for each isotope of interest, the detector signal of only one isotope or two or more selected isotopes or on an average of the detector signals for all or a subset of the isotopes of interest.

[0018] In one or more examples relating to a score-based classification, when similarity scores have been obtained between the sample chromatogram and each of the plurality of predetermined chromatogram patterns, selecting the set of peak detection parameters based on the similarity scores comprises selecting the set of peak parameters associated with the highest scoring chromatogram pattern, i.e., the parameters associated with the pattern indicated to have the highest similarity with the obtained sample chromatogram. In one or more examples, if the similarity scores are under a certain predefined threshold, the sample is flagged as ambiguous to alert the user to potential errors in the selection of the set of peak detection parameters.

[0019] In one or more examples, the method comprises classifying the obtained chromatogram by inputting the obtained chromatogram to a machine learning model and obtaining an output indicating the predetermined set of peak detection parameters to use. The input may comprise the chromatogram with the full MS signal, including the detector signals for each isotope of interest, the detector signal of only one isotope or two or more selected isotopes or on an average of the detector signals for all or a subset of the isotopes of interest. The classifier can take any suitable form, such as a neural network, a support vector machine, a decision tree, naive bayes, random forest, XGBoost, k-Nearest Neighbours (k-NN), or any other algorithm that can be used to implement a multiclass classifier as is well known in the art.

[0020] The machine learning model may take as inputs a numeric representation of the chromatogram, such as a vector of detection signal values with each element of the vector corresponding to a sample time. The numerical representation can either correspond to an obtained chromatogram or to a pre-processed representation of a chromatogram, such as a projection of the chromatogram onto a predefined number of principal components obtained from Principal Component Analysis of a data set of chromatograms in the relevant domain to reduce the dimensionality of the input. The output of the machine learning model may comprise either a probability of belonging to one of the plurality of known classes, for example obtained using a softmax function, or a binary vector indicating which set of peak determination parameters to use. In examples where the machine learning model is trained to output a class probability, the method of selecting a set of peak determination parameters may be the same as described above (i.e., selecting the set of peak determination parameters corresponding to the highest class probability or similarity score). The machine learning model can be trained using a dataset of exemplary chromatograms each paired with a respective set of peak detection parameters of the plurality of predetermined sets of peak detection parameters, wherein the model is trained to correctly predict the predetermined set of peak detection parameters (or an indication of the predetermined set of peak detection parameters) corresponding to the input chromatogram.

[0021] Training the model may comprise providing the exemplary chromatograms (or representations thereof) as inputs to the model, obtaining outputs of the model in response to the input chromatograms, and comparing the outputs of the model with the corresponding predetermined set of peak detection parameters for each input chromatogram. The parameters of the model are then updated and / or tuned to reduce the error between the output values and the known set of the plurality of predetermined sets of peak detection parameters. Various methods of training the model are known in the art and suitable for the methods disclosed herein. For example, the model may be a deep learning model, such as an artificial neural network (ANN). ANN can be trained using gradient descent, stochastic gradient descent or other optimization algorithms like ADAM, as is well known in the field of ANN.

[0022] In one or more examples relating to a user selection of peak detection parameter set, the method includes saving an indication of the selected predetermined set of peak detection parameters in a database in association with a representation of the obtained sample chromatogram and using the database to train a classifier to predict the predetermined set to use for a new sample chromatogram. In this way, selections made by the experienced user can be used to improve future classification of chromatograms.

[0023] Similarly, in some examples relating to machine learning models, if there is a manual or user modification to the selection of the set of peak detection parameters output by the machine learning model, the machine learning model can implement the manually modified set of peak detection parameters along with the respective input chromatogram in training. In other words, the manual user update of the parameters for a newly input chromatogram effectively allows for a further exemplary chromatogram and corresponding predetermined set of peak detection parameters (i.e., a new training pair) to be assigned to the database for model training, and the machine learning model can thus learn from user input.

[0024] In one or more examples described herein, the sample for chromatography can be a liquid sample or a gas sample. In examples relating to liquid samples, for example samples of different honeys, liquid chromatography is used to obtain a chromatogram of the sample. Similarly, gas chromatography is used in implementations relating to a gas sample or an evaporated liquid sample. In examples relating to liquid chromatography coupled with an isotope ratio mass spectrometer, the chromatographically separated liquid components are firstly converted into simple gases (e.g., CO2, SO2, N2, H2, etc.) which can subsequently be introduced into the isotope ratio mass spectrometer for analysis. Conversion to simple gases may also be done in gas chromatography.

[0025] In one or more implementations, one or more computer-readable media comprises coded instructions that, when run on a computing device coupled to a chromatography instrument and an isotope ratio mass spectrometer, implement methods as described above.

[0026] In one or more implementations, a system comprising a gas or liquid chromatography instrument coupled to an isotope ratio mass spectrometer is configured to receive a sample for chromatography and implement methods as described above. The chromatography instrument may be coupled to the isotope ratio mass spectrometer in any suitable way using any suitable interface, or the two instruments may be integrated into one instrument that is both a chromatography instrument and an IRMS, for example integrating a chromatography column with an input to a plasma source of the IRMS. It will be understood that reference to a chromatography instrument and an IRMS instrument in this disclosure does not imply that these two instruments are separate instruments, and they may be the same instrument fulfilling both functions.

[0027] In one or more implementations, a method of training a computer-implemented classifier for identifying a set of peak detection parameters associated with a representation of a chromatogram comprises acquiring a dataset. The dataset associates representations of chromatograms with corresponding predetermined sets of peak detection parameters, wherein the parameters comprise peak slope and / or peak magnitude thresholds, and the model is trained using the dataset. Various methods of training the model are known in the art and are described in more detail above. In some examples, the dataset used for training comprises a saved database from user inputs of chromatogram representations with associated sets of peak detection parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] FIG. 1 is a block diagram of a system for chromatographically evaluating a sample, separating the sample into components, and analysing the separated components in an IRMS.

[0029] FIG. 2 is a flow diagram of a workflow for the processing of a chromatogram of a sample to determine isotope ratios for one or more components of the chromatogram.

[0030] FIG. 3A is specific illustration of a honey sample chromatogram separated into components in accordance with the described methods and showing three IRMS detector signals, one for each of a respective selected isotope.

[0031] FIG. 3B illustrates chromatograms of three different honey samples, separated into components in accordance with the described methods.

[0032] FIG. 4 is a block diagram of an example scientific instrument support system in which some or all of the scientific instrument support methods disclosed herein may be performed. DETAILED DESCRIPTION

[0033] The subject matter is described with reference implementations and examples as illustrated in the accompanying drawings. In some cases, as will be recognized by one skilled in the art, the disclosed implementations and examples may be practiced without one or more of the disclosed specific details, or may be practiced with other methods, structures, and materials not specifically disclosed herein. All the implementations and examples described herein and shown in the drawings may be combined without any restrictions to form any number of combinations, unless the context clearly dictates otherwise, such as if the proposed combination involves elements that are incompatible or mutually exclusive. The sequential order of the acts in any process described herein may be rearranged, unless the context clearly dictates otherwise, such as if one act requires the result of another act as input.

[0034] In the interest of conciseness, and for the sake of continuity in the description, same or similar reference characters may be used for same or similar elements in different figures, and description of an element in one figure will be deemed to carry over when the element appears in other figures with the same or similar reference character. In some cases, the term “corresponding to” may be used to describe correspondence between elements of different figures. In an example usage, when an element in a first figure is described as corresponding to another element in a second figure, the element in the first figure is deemed to have the characteristics of the other element in the second figure, and vice versa, unless stated otherwise.

[0035] The word “comprise” and derivatives thereof, such as “comprises” and “comprising”, are to be construed in an open, inclusive sense, that is, as “including, but not limited to”. The singular forms “a”, “an”, “at least one”, and “the” include plural referents, unless the context dictates otherwise. The term “and / or”, when used between the last two elements of a list of elements, means any one or more of the listed elements. The term “or” is generally employed in its broadest sense, that is, as meaning “and / or”, unless the context clearly dictates otherwise. When used to describe a range of dimensions, the phrase "between X and Y" represents a range that includes X and Y. As used herein, an “apparatus” may refer to any individual device, collection of devices, part of a device, or collections of parts of devices.

[0036] FIG. 1 is a block diagram of a scientific instrument support module 1000 for performing support operations for a gas chromatography (GC-) or liquid chromatography (LC-) system 1120 coupled to an isotope ratio mass spectrometer (IRMS) 1140 of a scientific instrument (or scientific instrument system of individual instruments coupled together) 1100. The scientific instrument support module 1000 may be implemented by circuitry (e.g., including electrical and / or optical components), for example including a programmed computing device. The functionality of the scientific instrument support module 1000 may be included in a single computing device or may be distributed across multiple computing devices that are in communication with each other as appropriate. Examples of systems of interconnected computing devices, in which the scientific instrument support module 1000 may be implemented across one or more of the computing devices, is discussed herein with reference to the scientific instrument support system 5000 illustrated in FIG. 4. The scientific instrument support module 1000 implements methods as described below with reference to the workflow of FIG. 2, in order to facilitate the evaluation of chromatograms from the chromatography system 1120 as detected by the IRMS 1140. Generally, chromatography does not always lead to baseline separated peaks, that is peaks may run into each other and may not be separated by zero I background level signal. This is especially true with natural substance mixtures, such as honey, where using an aqueous mobile phase leads to only partially separated signals that are hard to correctly separate into components. When gas or liquid chromatography is coupled to isotope ratio mass spectrometry (GC- or LC-IRMS), this can be a significant problem, as a small variation in the definition of peak start and endpoints and the resulting small change in the eluded component comprised in the isotope ratio calculation may result in a large change of the isotope ratio. The methods described herein and implemented by the support module 1000 may help to address this problem. In particular, the methods described herein and implemented by the support module 1000 facilitate a more reproducible and reliable chromatographical separation of a sample into components based on peak identification. The chromatographically separated components of the detector signals of the IRMS1140 are subsequently used to determine the isotopic composition of each component.

[0037] FIG. 2 illustrates an example process flow for separating peaks corresponding to components of a chromatogram for IRMS analysis. At step 202, a complex sample, such as a honey, is input into a chromatography instrument for analyte component separation. At step 204 the sample is separated into components according to elution / retention time by the chromatography instrument. The chromatography instrument may be any suitable chromatography instrument, such as high-performance liquid chromatograph (HPLC) which typically utilizes a liquid mobile phase at a high-pressure flow to carry an injected sample around the instrument and into a separation column. The various analytes of the input sample interact differently and at different rates with a stationary phase comprised within the separation column. The sample therefore begins to separate into distinct analytes as it flows through and elutes from the column.

[0038] At step 206 the components eluting from the chromatography column are coupled into the I RMS to obtain detector signals for respective selected isotopes from the I RMS. To couple the components in into the I RMS, the components eluding from the chromatography column may be converted to simple gases (e.g., CO2, SO2, N2, H2, ...) and transferred to an ion source, for example a plasma source, electron ionisation source, chemical ionisation source, or the like, of the I RMS, where the gases are ionised. The gases are then accelerated towards the IRMS detector through an electric and / or magnetic field resulting in the separation of different particles according to charge I mass ratio and hence isotopes. Detector signals are identified for each of a set of isotopes of interest and the magnitude of each identified detector indicates the abundance, for example concentration, of the corresponding isotope. The operation of the chromatography and IRMS systems to obtain isotope IRMS detector signals is well known to the person skilled in the art and is typically automated at least to an extent in commercially available systems. One example of an interface between Liquid Chromatography and IRMS is an LC IsoLink™ which is often used in contract labs for measurement of isotopes in natural product samples (e.g., honey).

[0039] At step 208, as components elute from the chromatography column into the IRMS, a chromatogram of the detector signals for each of the set of isotopes as a function of time is obtained and recorded. The stored chromatogram can then be separated into peaks corresponding to the eluted components to determine respective isotope ratios for the set of isotopes.

[0040] Specifically, steps 209a, 209b, and 209c illustrate three example process options for selecting a set of peak detection parameters with which to evaluate the isotope ratios. These options may be alternatives included separately in different versions of the support module 1000 or may be provided as options in the support module 1000, for example any two as two options or all three as three options provided by the support module 1000 in respective operating modes.

[0041] At step 209a, a user interface outputs options for selection of respective predetermined sets of peak detection parameters by a user of the support module 1000, for example by displaying the options on a display screen or causing the sending of a display screen to a remote display. In some implementations, the options are presented as a dropdown menu, for example as part of a display option selection panel. The predefined sets have been previously defined for example by expert users as relevant to a type or types of sample that may be encountered and have been saved in a database for subsequent selection and use. The predetermined sets of peak detection parameters may be standard sets of peak detection parameters for various known samples, for example, in the case of honey, the predetermined sets of peak detection parameters saved in a database may comprise predetermined sets relating to known honey variants, such as parameter sets for rape honey or acacia honey. From the output selection of predetermined sets of peak detection parameters, a user may choose and switch between peak detection parameter sets and choose a preferred peak detection parameter set for the obtained chromatogram. Fast switching and selecting of different &easily accessible predetermined sets of peak detection parameters for a single obtained chromatogram of a sample beneficially facilitates the efficient evaluation of a sample which cannot be otherwise evaluated by application of fixed peak-finding parameters, e.g., if the sample is not chromatographically fully baseline separated. The provision of readily selectable pre-sets avoids the need for a user to tweak a potentially large number of parameters over and over again for each sample.

[0042] In another option, at step 209b, the obtained chromatogram of the sample is classified by applying a pattern matching or similarity algorithm which scores standard chromatogram pattern from a library of standard chromatogram patterns corresponding to the predefined sets of parameters. The pattern matching algorithm is done by matching the obtained chromatogram to normalised chromatograms from the library, to make the match agnostic to intensities of the chromatogram measurements, which may vary from sample to sample. The pattern matching may be achieved by any known method, such as a cosine similarity of vectors representing the obtained chromatogram and chromatograms from a library, respectively. The pattern matching returns a score for the quality of a match between the obtained chromatogram and the normalised chromatograms of the library. The selection of the set of peak detection parameters may, for example, be achieved by finding the parameter set which corresponds to the standard normalised chromatogram pattern with the highest score, i.e., the closest matching chromatogram.

[0043] In yet another option, at step 209c, the library of step 209b is, in effect, replaced by a trained neural network. Specifically, at step 209c, a machine learning model is used that has been trained to take an input of the obtained chromatogram (or of a representation thereof) and to output an indication of which predetermined set of peak detection parameters to select. In particular, the input of the machine learning model can be the obtained chromatogram itself, or it may be a representation of the line measurement of the chromatogram, e.g., a representation of the chromatogram with dimensional reduction. The output of the machine learning model can be a score or a likelihood value that the input chromatogram is a certain type of chromatogram, i.e., a chromatogram of a known classification and / or with a known corresponding predetermined set of peak detection parameters. The model has been trained using pairs of chromatogram patterns or representations and corresponding standard sets of peak detection parameters, e.g., such as the data stored in the library of step 206. The model has been trained by obtaining an output indicating which of the standard sets of peak detection parameters is a corresponding standard set for an input chromatogram pattern or representation, comparing the output with the known corresponding standard set of peak detection parameters for the input, and adjusting parameters of the model to reduce a discrepancy between the predicted set between the predicted standard parameter set and the actual standard parameter set. Various types and training methods of a suitable model for the workflow in FIG. 2 are known in the art.

[0044] In either steps 209b or 209c, a scoring system may be optionally implemented to help the user identify chromatograms that need to be checked after the respective algorithms select a set of peak detection parameters. For example, the process could output a flag to the user to indicate that a selected predetermined set of parameters was selected on the basis of relatively low similarity score, for example a similarity score below a pre-set threshold. This may beneficially indicate to the user that visual inspection of the peak separation should be carried out to make sure the peaks were separated appropriately and the next step.

[0045] Once the set of peak detection parameters is selected through any of steps 209a, 209b or 209c, the method proceeds to step 210. At step 210, the chromatogram is automatically segmented into regions corresponding to peaks based on the selected set of peak detection parameters, for example by automatically (under software control) setting peak boundaries based on the selected set of peak detection parameters. For example, the selected set of peak detection parameters may comprise start and end gradient thresholds for a slope. In such an example, operations are carried out to determine the regions of the obtained chromatogram which are bounded by the threshold gradient values. Operations to analyse a line based on threshold values are well known in the art and may comprise any such known mathematical techniques. For example, in the above case of threshold gradient values, the operation may simply comprise computing slope gradients between points at a defined resolution along the chromatogram line and comparing the slope gradients of the chromatogram with the threshold gradient values.

[0046] Segmentation into peaks may be done individually for each isotope trace and the resulting peak delimiting positions averaged. Alternatively, the isotope traces may be summed or averaged, and the summed or averaged trace may be used for segmentation into peaks. Alternatively, segmentation may be done based on a subset of one or more of the isotope traces and the identified peak limits may be applied to the remaining isotope traces. For example, a single isotope trace may be used for peak segmentation and the segmentation used with that and the remaining traces.

[0047] The regions corresponding with the threshold values are segmented and the method proceeds to step 212. At step 212, isotope ratios for each segmented peak / components of the sample are evaluated. In some implementations, the ratio of the respective peak values of the traces for two isotopes is computed to find the isotope ratio for the corresponding peak I component. In other implementations, the area under each respective isotope trace is integrated across the peak and the isotope ratio is calculated as a ratio of the area. Any other suitable way of computing the isotope ratio of interest for each peak may be used in other implementations. The computed ratio(s) are stored and / or output on a display.

[0048] FIG. 3A shows a chromatogram of a honey sample after conversion to CO2 for 13C / 12C determination in an isotope ratio mass spectrometry (IRMS) system as a plot of IRMS detector intensities in V against time for each of three CO2 species with different carbon and / or oxygen isotopes, specifically with atomics masses of 44, 45 and 46, corresponding to 12C, 13C and 14C / 18O. As described above, to find the isotope ratio of 12C to 13C, say, for each peak I component, the ratio of the respective peak heights for the 12C and 13C traces may be evaluated, for example. The first three fractions (Trisaccharides, Disaccharides, and Glucose) overlap and are not perfectly baseline separated, either due to the limitations of chromatography, or because the conversion process of the fractions into CO2 introduces additional peak broadening. The disaccharides “peak” consists of a class of substances, such as sucrose and maltose, which are evaluated as a bulk peak.

[0049] Natural samples, such as honeys, mostly contain the same or similar substances but with different contents based on their e.g., origin, manufacturing process, treatment process, or possible manipulation (e.g., fraud with syrups). Due to this variability of (natural) samples composition it is often difficult to detect the start and end of the individual peaks using fixed peak finding parameters. The incomplete separation is resulting in untypical peak shapes, which are therefore not detectable with a single peak detection parameter set.

[0050] Fig. 3B shows chromatograms of three different honey samples after conversion to CO2 as a plot of IRMS detector intensities in V against time for a CO2 species with an atomic mass of 44 (12C). As can be seen, the three honey samples have natural variations which cause differences in the shape of the respective fractions. Such variety in peaks from natural samples may mean that the same peak detection parameters are not suitable for different samples, which often requires manual integration of each chromatogram which can be time consuming and inefficient. As discussed herein, by having predetermined sets of peak identification parameters (e.g., parameters for identifying the presence and extent for each peak or fraction of a sample) and wherein each of the predetermined sets of parameters may correspond to or be suitable for a different variation of a complex sample, the timeconsuming and repeated manual integration of each chromatogram can be avoided. In this way, evaluating a sample by selecting a suitable parameter set from a pre-existing parameter set improves efficiency and allows for more samples to be evaluated over the course of a day.

[0051] The methods described herein are particularly advantageous for use with samples (e.g., particular classes of samples) having chromatograms where the peaks have some overlap and indistinctness, such that the correct separations (i.e., correct peak detection parameters) are not immediately obvious. In particular, by having suitable sets of predetermined peak detection parameters, e.g., parameters which may be known or may correspond to known classes of sample, peak detection parameters may be more efficiently selected, by human and / or machine, for peak identification and / or more efficiently identified as suitable peak detection parameters for use. As such, by selecting a set of peak detection parameters from sets of predetermined peak detection parameters, the requirement for determining new peak detection parameters afresh for each input chromatogram can be avoided, and the efficiency in determining sample isotope ratios, particularly of multiple components, can be increased.

[0052] The presently described methods, such as the workflow described in FIG. 2, circumvent the high amount of manual re-evaluation of the measured chromatogram otherwise required for such variable sample compositions. In particular, the methods described herein allow for fast selection of peak detection parameters and the efficient classification of complex natural compounds, such as honey, which have no baseline-separated peaks. It will be appreciated that these methods allow for reproducible results whilst reducing the amount of time needed for chromatogram evaluation in a lab, and thus increase productivity. Additionally, whilst the illustration of FIGs. 3A and 3B relate to honey chromatograms, it will be understood that the methods described herein and the corresponding advantages that come therewith can be implemented with any liquid or gas sample which is suitable for chromatographical analyte component separation.

[0053] One or more computing devices implementing any of the scientific instrument support modules or methods disclosed herein may be part of a scientific instrument support system. FIG. 4 is a block diagram of an example scientific instrument support system 5000 in which some or all of the scientific instrument support methods disclosed herein may be performed, in accordance with various embodiments. The scientific instrument support modules and methods disclosed herein (e.g., the LC-IRMS or GC-IRMS support module 1000 of FIG. 1 and the methods of FIG. 2) may be implemented by one or more of the scientific instrument 5010, the user local computing device 5020, the service local computing device 5030, or the remote computing device 5040 of the scientific instrument support system 5000.

[0054] The scientific instrument 5010, the user local computing device 5020, the service local computing device 5030, or the remote computing device 5040 may each include a processing device 5002, a storage device 5004, and an interface device 5006. The processing device 5002 may take any suitable form, and the processing devices 5002 included in different ones of the scientific instrument 5010, the user local computing device 5020, the service local computing device 5030, or the remote computing device 5040 may take the same form or different forms. The storage device 5004 may take any suitable form, and the storage devices 5004 included in different ones of the scientific instrument 5010, the user local computing device 5020, the service local computing device 5030, or the remote computing device 5040 may take the same form or different forms. The interface device 5006 may take any suitable form and the interface devices 5006 included in different ones of the scientific instrument 5010, the user local computing device 5020, the service local computing device 5030, or the remote computing device 5040 may take the same form or different forms.

[0055] The scientific instrument 5010, the user local computing device 5020, the service local computing device 5030, and the remote computing device 5040 may be in communication with other elements of the scientific instrument support system 5000 via communication pathways 5008. The communication pathways 5008 may communicatively couple the interface devices 5006 of different ones of the elements of the scientific instrument support system 5000, as shown, and may be wired or wireless communication pathways. The particular scientific instrument support system 5000 depicted in FIG. 4 includes communication pathways between each pair of the scientific instrument 5010, the user local computing device 5020, the service local computing device 5030, and the remote computing device 5040, but this “fully connected” implementation is simply illustrative, and in various implementations, various ones of the communication pathways 5008 may be absent. For example, in some implementations, a service local computing device 5030 may not have a direct communication pathway 5008 between its interface device 5006 and the interface device 5006 of the scientific instrument 5010, but may instead communicate with the scientific instrument 5010 via the communication pathway 5008 between the service local computing device 5030 and the user local computing device 5020 and the communication pathway 5008 between the user local computing device 5020 and the scientific instrument 5010.

[0056] The scientific instrument 5010 may include any appropriate scientific instrument, such as the scientific instrument 1100 of FIG. 1, which includes a liquid or gas chromatography system 1120 coupled to an isotope ratio mass spectrometer 1140.

[0057] The user local computing device 5020 may be a computing device that is local to a user of the scientific instrument 5010. In some implementations, the user local computing device 5020 may also be local to the scientific instrument 5010, but this need not be the case; for example, a user local computing device 5020 that is in a user’s home or office may be remote from, but in communication with, the scientific instrument 5010 so that the user may use the user local computing device 5020 to control and / or access data from the scientific instrument 5010. In some implementations, the user local computing device 5020 may be a laptop, smartphone, or tablet device. In some implementations the user local computing device 5020 may be a portable computing device.

[0058] The service local computing device 5030 may be a computing device that is local to an entity that services the scientific instrument 5010. For example, the service local computing device 5030 may be local to a manufacturer of the scientific instrument 5010 or to a third-party service company. In some implementations, the service local computing device 5030 may communicate with the scientific instrument 5010, the user local computing device 5020, and / or the remote computing device 5040 (e.g., via a direct communication pathway 5008 or via multiple “indirect” communication pathways 5008, as discussed above) to receive data regarding the operation of the scientific instrument 5010, the user local computing device 5020, and / or the remote computing device 5040 (e.g., the results of self-tests of the scientific instrument 5010, calibration coefficients used by the scientific instrument 5010, the measurements of sensors associated with the scientific instrument 5010, etc.). In some implementations, the service local computing device 5030 may communicate with the scientific instrument 5010, the user local computing device 5020, and / or the remote computing device 5040 (e.g., via a direct communication pathway 5008 or via multiple “indirect” communication pathways 5008, as discussed above) to transmit data to the scientific instrument 5010, the user local computing device 5020, and / or the remote computing device 5040 (e.g., to update programmed instructions, such as firmware, in the scientific instrument 5010, to initiate the performance of test or calibration sequences in the scientific instrument 5010, to update programmed instructions, such as software, in the user local computing device 5020 or the remote computing device 5040, etc.). A user of the scientific instrument 5010 may utilize the scientific instrument 5010 or the user local computing device 5020 to communicate with the service local computing device 5030 to report a problem with the scientific instrument 5010 or the user local computing device 5020, to request a visit from a technician to improve the operation of the scientific instrument 5010, to order consumables or replacement parts associated with the scientific instrument 5010, or for other purposes.

[0059] The remote computing device 5040 may be a computing device that is remote from the scientific instrument 5010 and / or from the user local computing device 5020. In some implementations, the remote computing device 5040 may be included in a datacentre or other large-scale server environment. In some implementations, the remote computing device 5040 may include network-attached storage (e.g., as part of the storage device 5004). The remote computing device 5040 may store data generated by the scientific instrument 5010, perform analyses of the data generated by the scientific instrument 5010 (e.g., in accordance with programmed instructions), facilitate communication between the user local computing device 5020 and the scientific instrument 5010, and / or facilitate communication between the service local computing device 5030 and the scientific instrument 5010.

[0060] In some implementations, one or more of the elements of the scientific instrument support system 5000 illustrated in FIG. 4 may not be present Further, in some implementations, multiple ones of various ones of the elements of the scientific instrument support system 5000 of FIG. 4 may be present. For example, a scientific instrument support system 5000 may include multiple user local computing devices 5020 (e.g., different user local computing devices 5020 associated with different users or in different locations). In another example, a scientific instrument support system 5000 may include multiple scientific instruments 5010, all in communication with service local computing device 5030 and / or a remote computing device 5040; in such an embodiment, the service local computing device 5030 may monitor these multiple scientific instruments 5010, and the service local computing device 5030 may cause updates or other information may be “broadcast” to multiple scientific instruments 5010 at the same time. Different ones of the scientific instruments 5010 in a scientific instrument support system 5000 may be located close to one another (e.g., in the same room) or farther from one another (e.g., on different floors of a building, in different buildings, in different cities, etc.). In some implementations, a scientific instrument 5010 may be connected to an Internet-of-Things (loT) stack that allows for command and control of the scientific instrument 5010 through a web-based application, a virtual or augmented reality application, a mobile application, and / or a desktop application. Any of these applications may be accessed by a user operating the user local computing device 5020 in communication with the scientific instrument 5010 by the intervening remote computing device 5040. In some implementations, a scientific instrument 5010 may be sold by the manufacturer along with one or more associated user local computing devices 5020 as part of a local scientific instrument computing unit 5012.

[0061] In some implementations, different ones of the scientific instruments 5010 included in a scientific instrument support system 5000 may be different types of scientific instruments 5010; for example, one scientific instrument 5010 may be a gas and / or liquid chromatography system, while another scientific instrument 5010 may be an isotope ratio mass spectrometer. In some such implementations, the remote computing device 5040 and / or the user local computing device 5020 may combine data from different types of scientific instruments 5010 included in a scientific instrument support system 5000.

Claims

1. A method of determining isotope ratios for components of a sample, the method comprising:separating the sample into components in a chromatography column;coupling the separated sample eluting from the chromatography column into an isotope ratio mass-spectrometer;obtaining a chromatogram of the sample comprising a mass-spectrometer signal as a function of time, the mass-spectrometer signal comprising a respective detector signal for each of a plurality of isotopes of interest;selecting a set of peak detection parameters for the chromatogram from a plurality of predetermined sets of peak detection parameters;automatically segmenting the chromatogram into peaks corresponding to the components using operations based on the peak detection parameters of the selected set; anddetermining one or more isotope ratios for one or more of the components using the mass-spectrometer signal of each respective peak.

2. The method of claim 1, wherein the plurality of predetermined sets of peak detection parameters are determined prior to the separating the sample into components in a chromatography column.

3. The method of any preceding claim, further comprising accessing a database comprising the plurality of predetermined sets of peak detection parameters, and selecting a set of peak detection parameters for the chromatogram from the plurality of predetermined sets of peak detection parameters in the database.

4. The method of any preceding claim, wherein the predetermined sets previously have been saved in a database for subsequent selection and use.

5. The method of any of claims 1 to 4, comprising presenting a user interface for selecting one of the plurality of predetermined sets of peak detection parameters to a user, receiving a user selection of one of the predetermined sets and selecting the set of peak detection parameters based on the user selection.

6. The method of any of claims 1 to 4, comprising classifying the chromatogram of a sample as belonging to one of a plurality of classes corresponding to the plurality of predetermined sets of peak detection parameters and selecting the set of peak detection parameters corresponding to the one of the plurality of classes.

7. The method of claim 6, comprising classifying the chromatogram by determining a similarity score with a predetermined chromatogram pattern associated with a respective predetermined set of peak detection parameters for each of the predetermined sets of peak detection parameters and selecting the set of peak detection parameters based on the similarity scores.

8. The method of any preceding claim, comprising classifying the chromatogram by inputting the chromatogram to a machine learning model and obtaining an output indicating the predetermined set of peak detection parameters to use.

9. The method of claim 5 comprising:saving an indication of the selected predetermined set of peak detection parameters in a database in association with a representation of the chromatogram and using the database to train a classifier to predict the predetermined set to use for a new chromatogram.

10. The method of any preceding claim, wherein separating the sample is by liquid chromatography.

11. The method of any preceding claim, wherein the sample comprises honey.

12. One or more computer-readable media comprising coded instructions that, when run on a computing device coupled to a chromatography instrument and an isotope ratio mass spectrometer, implement a method as claimed in any preceding claim.

13. A scientific instrument system comprising a gas or liquid chromatography system coupled to an isotope ratio mass spectrometer, wherein the scientific instrument system is configured to receive a sample for chromatography and implement a method as claimed in any preceding claim.

14. A method of training a computer-implemented classifier for identifying a set of peak detection parameters associated with a representation of a sample chromatogram, the method comprising:acquiring a dataset, wherein the dataset associates representations of chromatograms with corresponding predetermined sets of peak detection parameters, the parameters comprising peak slope and / or peak magnitude and / or peak extent thresholds; andtraining the model using the dataset.

15. The method of claim 14, wherein the dataset comprises a database containing saved instances of user inputs of a selection of peak detection parameters input for corresponding chromatogram representations.21

Images