Analysis method, analysis device, analysis program, and standard shape generation method

Abstract

To provide an analysis method capable of applying an analysis of a high reliability to the data of an insufficient separation.SOLUTION: An analysis method of data generated from at least two parameters includes: a comparison step of comparing a data shape of a target component in a sample to be measured with a previously obtained standard shape; and a step of identifying the target component in the sample to be measured on the basis of the comparison.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an analysis method, an analysis device, an analysis program, and a method for generating a standard shape. [Background technology]

[0002] In the field of chemical analysis, it is known that data containing two or more parameters for a given component in a sample is obtained, and the data is analyzed to identify and quantify the given component. For example, in chromatography, a sample is separated in time, space, and mass, the separated components are detected by a detector, and the signal intensity obtained from the detector is recorded mainly over time. This results in data (e.g., chromatograms) containing mainly two parameters, time and signal intensity. Then, based on the magnitude of the quantitative parameter for the target component to be identified and quantified (e.g., the magnitude of the chromatogram peak), the component can be identified and quantified.

[0003] In the above-mentioned quantification, it is necessary to set a standard for determining the magnitude of a parameter for the target component. For example, in the above-mentioned chromatogram example, it is necessary to set a baseline for determining the height and area of ​​the peak of the target component. However, data is not necessarily independent (separated) for the target component, and it may be difficult to identify the magnitude of the parameter due to the influence of other nearby components. For this reason, a method for quantifying a component based on unseparated data is being considered. For example, Non-Patent Document 1 describes a method for quantifying a component based on unseparated peaks in a chromatogram. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Edited by Hiroshi Nakamura, Japan Society for Analytical Chemistry, Liquid Chromatography Research Forum, "Liquid Chromatography Dog Volume", February 2, 2006, pp. 96-97 Summary of the Invention [Problem to be solved by the invention]

[0005] However, as described in Non-Patent Document 1, there has been no established standard method for quantification based on unseparated data. Therefore, data analysis of the target component cannot be performed appropriately, and it is difficult to ensure the reliability of the obtained quantitative value, and there are cases where a highly reliable analysis cannot be performed.

[0006] In view of the above, one aspect of the present invention provides a data analysis method that enables more reliable analysis. [Means for solving the problem]

[0007] One aspect of the present invention is a method for analyzing data generated from at least two parameters, comprising a comparison step of comparing a data shape of a target component in a sample to be measured with a standard shape obtained in advance, and a step of identifying the target component in the sample to be measured based on the comparison.

[0008] Moreover, one aspect of the present invention is a method for analyzing data generated from at least two parameters, comprising a comparison step of comparing a data shape of a target component in a measured sample with a previously obtained standard shape, and a quantitative value calculation step of determining a quantitative value of the target component in the measured sample based on the comparison, wherein the standard shape includes a group of shapes for the target component obtained from a plurality of standard samples each containing the target component at different quantitative values. Effect of the Invention

[0009] According to one aspect of the present invention, there is provided a data analysis method that enables more reliable analysis. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram illustrating features of a data shape. [Diagram 2] 1 shows an example of a functional configuration of an analysis device according to an embodiment of the present invention. [Diagram 3] 1 illustrates an example of a hardware configuration of an analysis device according to an embodiment of the present invention. [Figure 4] FIG. 4 is a flow diagram of a calibration data generation process in an analysis method according to one embodiment of the present invention. [Diagram 5] FIG. 13 is a diagram showing an example of the relationship between a quantitative value (concentration) and a magnification of a data shape with respect to a representative standard shape. [Figure 6] FIG. 2 is a flow diagram of a quantitative process in an analysis method according to one embodiment of the present invention. [Figure 7] FIG. 4 is a detailed flow diagram of a comparison process in an analysis method according to one embodiment of the present invention. [Figure 8] FIG. 13 is a diagram illustrating partial shape generation. [Figure 9] FIG. 13 is a diagram showing an example of three-dimensional data. [Figure 9a] FIG. 11 is a flow diagram illustrating a modified analysis method according to an embodiment of the present invention. [Figure 10] FIG. 1 shows data from Example 1. [Figure 11] FIG. 1 shows data from Example 2. [Figure 12] FIG. 1 shows data from Example 3. [Figure 13] FIG. 11 is a flow diagram showing details of a calibration data generation process in the embodiment. [Figure 14] FIG. 4 is a flow chart showing details of a quantitative process in an embodiment. [Figure 15A] 13 shows a quality check sheet output regarding the analytical conditions for blood samples in Example 4. [Figure 15B] 13 shows a quality check sheet output regarding the analytical conditions for blood samples in Example 4. [Figure 16A] 1 shows a quality check sheet output for the analysis conditions of urine samples in Example 4. [Figure 16B] 1 shows a quality check sheet output for the analysis conditions of urine samples in Example 4. [Figure 17A] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 17B] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 17C] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 17D] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 17E] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 17F] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 17G] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 17H] 1 shows the results of shape fitting for the blood samples of Example 4. [Figure 18A] 1 shows the results of shape fitting for the urine samples of Example 4. [Figure 18B] 1 shows the results of shape fitting for the urine samples of Example 4. [Figure 18C] 1 shows the results of shape fitting for the urine samples of Example 4. [Figure 18D] 1 shows the results of shape fitting for the urine samples of Example 4. [Figure 18E] 1 shows the results of shape fitting for the urine samples of Example 4. [Figure 18F] 1 shows the results of shape fitting for the urine samples of Example 4. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] One form of the present invention is a method for analyzing data generated from at least two parameters, comprising a comparison step of comparing a data shape of a target component in a sample to be measured with a standard shape obtained in advance, and a step of identifying the target component in the sample to be measured based on the comparison.

[0012] Also, one aspect of the present invention is a method for analyzing data generated from at least two parameters, comprising a comparison step of comparing a data shape of a target component in a measurement target sample with a previously obtained standard shape, and a quantitative value calculation step of calculating a quantitative value of the target component in the measurement target sample based on the comparison, wherein the standard shapes include a group of shapes for the target component obtained from a plurality of standard samples each containing the target component at a different quantitative value.

[0013] Furthermore, one aspect of the present invention is a method for creating standard shapes for use in analyzing data generated from at least two parameters, including a group of shapes of a target component obtained from a plurality of standard samples each containing the target component at different quantitative values.

[0014] In this specification, "data shape" or "shape" (sometimes referred to as "data form" or "shape") refers to the form of data in which the characteristics of a component in a sample are expressed using two or more parameters, and can also be expressed by some function using two or more parameters. "Data shape" is not limited to shapes in physical spaces such as two-dimensional space (plane) and three-dimensional space, but is a concept that can be expanded to shapes in multidimensional space coordinates of four or more dimensions. Here, when determining the quantitative value of a target component, at least one of the parameters is a quantitative parameter for that target component.

[0015] In the following, an embodiment of the present invention will be described mainly based on an example of separation analysis data including two parameters, specifically, data representing time and signal intensity over time (data representing signal intensity over time), but the present invention is not limited to the specific forms described in the specification.

[0016] <About Data Shape> The inventors conducted extensive research into the shape of separation analysis data generated from at least two parameters and discovered that when the data shape of a target component obtained under the same conditions is stretched in one direction (specifically, in the direction of the quantitative parameter axis) by a specified magnification, it has the same shape.

[0017] The above findings will be explained using a chromatogram as an example. A chromatogram is data (diagram, graph) obtained by chromatography that separates a sample over time, and is data generated by two parameters: time and signal intensity detected by a detector. Of the two parameters of a chromatogram, the signal intensity is a parameter that reflects the amount of the target component, i.e., a quantitative parameter, and the time is a parameter that does not reflect the amount of the target component, i.e., a non-quantitative parameter.

[0018] Figure 1(a) shows a part of a chromatogram obtained by subjecting a sample containing D-asparagine to high performance liquid chromatography. Each data (peak) in Figure 1(a) was obtained under the same analytical conditions, but by changing only the concentration of D-asparagine in the sample (the D-asparagine concentrations of each peak are 0.025 pmol / inj, 0.25 pmol / inj, 2.5 pmol / inj, and 5.0 pmol / inj, respectively). In the chromatogram shown in the figure, the horizontal axis represents retention time, and the vertical axis represents signal intensity.

[0019] Figure 1(b) shows the magnification of each peak shown in Figure 1(a) in the vertical axis (intensity axis) direction changed so that the intensity (peak height) at the peak top becomes the same. More specifically, the peaks of 0.025 pmol / inj, 0.25 pmol / inj, and 2.5 pmol / inj were each enlarged in the intensity axis direction so that the intensity at the peak top becomes the same as the intensity at the peak top of the 5 pmol / inj peak. In Figure 1(b), each peak is shifted in the intensity axis direction so that the intensity (vertical axis value) at the peak top position becomes zero.

[0020] As shown in FIG. 1(b), by changing the magnification in the intensity axis direction, all peaks are overlapped (all peaks have almost the same shape). In this specification, "substantially the same" and "almost the same" mean that some deviation due to mechanical, electrical, or artificial errors during measurement is also included. In addition, "same analytical conditions" or "same analytical system" means that, in the case of chromatography, analytical conditions such as column size and shape, packing material, stationary phase, mobile phase, carrier type, flow method, and temperature are substantially the same.

[0021] In this way, it can be seen that the data shape of a given component obtained under the same analytical conditions has characteristics specific to that component.

[0022] In the above chromatogram example, the peak of a given component obtained under the same analytical conditions shows a shape that is specific to the component under those analytical conditions, except for the difference in the size (scale) along the intensity axis due to concentration. In other words, the two peaks p A , p B When comparing the two, the ratio of the increase in the intensity in the direction of the intensity axis in a common arbitrary time range Δt is ΔI A / ΔI Bis approximately constant. Alternatively, when tangents are drawn to both peaks at the same position on the time axis and the slopes of the tangents to both peaks are compared, the ratio between the two is the same or substantially the same at any position on the time axis.

[0023] Therefore, even if data on a target component is inseparable (not independent) and overlaps with data on another adjacent component (impurity), if there is even a partial data shape that is not influenced by the other adjacent component, the target component can be identified (specified) by comparing that partial shape with a data shape (standard shape) of known concentration that has been obtained in advance, and after identifying (specifying) the target component, its concentration can be estimated or quantified.

[0024] For example, in a chromatogram, the shape of a part or the whole of the target component peak of the sample to be measured is superimposed on the shape of a standard peak (standard shape) of known concentration so that the positions of both correspond in the time axis direction, and the magnification / reduction ratio of the target component peak in the signal intensity axis direction relative to the standard peak is calculated (scaling), and the target component can be identified (specified) and its concentration can be predicted from the degree of overlap. When the peak of the target component and the standard peak are superimposed, for example, the positions of the peak tops (the apexes of the shape) in the time axis direction can be made to correspond / correct. In this case, if each point in the intensity axis direction of the shape under the analysis conditions is a probability density function, the intensity can be corrected according to the correction rate of the time axis. For example, when the rate of change in the time axis direction is ΔT (time correction amount) / T (time before correction), the intensity axis direction can also be corrected at the rate of ΔT / T.

[0025] Furthermore, by determining the relationship between the magnitude of the peak on the intensity axis (quantitative parameter axis) and the concentration based on the shapes of multiple standard peaks (calibration standard shapes) and saving it as a regression equation, this can be used as a calibration curve. Then, by using at least one standard peak shape in combination with the regression equation, the target component in a sample can be quantified.

[0026] In this manner, in this embodiment, the shape of the object to be measured is superimposed on a previously obtained standard shape and compared (shape fitting, shape scaling) to enable identification and quantification. Therefore, for example, in a chromatographic analysis, it is not necessary to set a baseline on a chromatogram to obtain peak height or peak area. Therefore, compared to conventional methods in which errors, etc., were artificially generated by setting an arbitrary baseline, the reliability of the quantitative values ​​obtained can be improved.

[0027] Generally, if data on the target component that is completely separated can be obtained (for example, the peaks can be completely separated), the target component can be easily identified and quantified. However, there are cases where the target component and other components cannot be chemically separated, or where it takes a huge amount of time physically, making it unrealistic to obtain specific, completely separated data on the target component. Therefore, one embodiment of the present invention has the effect of shortening the analysis time. In addition, even in the case of data in which the target component cannot be separated and identification and quantitative analysis are difficult using conventional methods, it is possible to analyze the data using this embodiment as long as the data shape is partially present.

[0028] <Example of functional configuration of analysis device> An example of the functional configuration of a waveform data analysis device for carrying out an analysis method according to one embodiment of the present invention is shown in Fig. 2. As shown in Fig. 2, the analysis device 10 can be configured to have an input means 31, an output means 32, a storage means 33, a data acquisition means 11, a smoothing means 12, a target component data detection means 13, a comparison means 14, a quantitative value calculation means 15, a regression equation generation means 16, and a control means 17.

[0029] As shown in FIG. 2, the analysis device 10 is connected to an analysis device 40. The analysis device 40 may be any chemical analysis device capable of analyzing a sample and outputting quantitative data. Examples of the analysis device 40 include chromatography using a gas, liquid, supercritical fluid, or the like as a mobile phase, mass spectrometry such as quadrupole type, double focusing type (magnetic field type), time-of-flight type, ion trap type, and ion cyclotron resonance type, spectroscopic analysis using infrared light, visible light, ultraviolet light, X-rays, fluorescence, and the like, and methods and devices that combine these. The analysis device 40 may include a detector that can quantitatively detect components in a sample. The detector can be capable of detecting components and outputting optical or mass values ​​for the components.

[0030] The analytical device 10, for example, compares the data shape (e.g., peak shape) of a target component in a measured sample with a standard shape for the target component that has been acquired in advance, and based on the comparison, identifies the target component (identification process) and determines a quantitative value of the target component, the amount of which is unknown (quantitation process).

[0031] The analysis device 10 can also generate calibration data such as a calibration curve for the above-mentioned quantitative processing (calibration data generation processing). In this case, as described later, data shapes of multiple standard product data are obtained as standard shapes, and a regression equation showing the relationship between the quantitative value and the size of the standard shape is obtained. At least one of the standard shapes and the regression equation can be used as calibration data (calibration curve data). In this case, the data shapes of the multiple standard product data do not necessarily need to be separated.

[0032] The data acquisition means 11 acquires data from the analysis device 40. The data may be data including two or more parameters obtained for one or more samples. At least one of the two or more parameters is a quantitative parameter related to the amount of the target component. For example, the data may be waveform data indicating signal intensity with respect to time (time-series signal intensity) obtained by separation analysis. The signal intensity may be a signal intensity representing an optical or mass quantity that may be generated depending on the type of detector. The data acquisition means 11 can also acquire different signal intensities (quantitative parameters) from multiple analysis devices or multiple detectors. The data may be acquired from the analysis device 40 as image data, or the data acquisition means 11 can generate image data.

[0033] The data acquisition means 11 can acquire data on a sample of unknown concentration (sample to be measured) and can also acquire data for creating calibration data (calibrator). In addition to the above data, the data acquisition means 11 can acquire data on the sample and target component (retention time of the target component, etc.). In the calibration data creation process, data on the concentration of the target component can be acquired. Furthermore, when both quality control and quality evaluation are performed using the analysis method according to this embodiment, the data acquisition means 11 also acquires data on a quality control sample.

[0034] The smoothing means 12 performs a smoothing process to remove noise from data acquired as an image, etc. As the smoothing process, smoothing by polynomial fitting such as the Savitzky-Golay method is preferable, but filtering such as simple average, median, maximum / minimum value, opening / closing, edge-preserving smoothing, K-nearest neighbor method, selective averaging method, weight method, etc. can also be used.

[0035] The target component data detection means 13 detects and extracts data of the target component from the data acquired by the data acquisition means 11, which may include data other than the target component, based on the position information of the target component.

[0036] The comparison means 14 compares the data shapes of the target component in the acquired data. In the identification process or quantification process, the data shape of the target component of the sample to be measured (also called the measurement target shape) is compared with a standard shape acquired in advance. In this case, the measurement target shape is superimposed on the standard shape to compare the magnitudes of the quantitative parameters. The comparison means 14 can also compare multiple standard shapes with each other in the calibration data generation process. In this case, the data shapes of the multiple standard data do not necessarily need to be separated.

[0037] As shown in FIG. 2, the comparison means 14 may include a partial shape generation means 14a, a superimposition means 14b, and a magnification acquisition means 14c.

[0038] The partial shape generating means 14a divides the measurement target shape in a non-quantitative parameter axis direction (time axis direction, etc.) to generate a partial shape. Therefore, the partial shape is a part of the measurement target shape (data shape). Moreover, this partial shape can be a shape that is a part of the data shape and includes the vertices of the data shape. The division process is included in the process (described later) for improving the accuracy of comparison in the quantitative process.

[0039] The superimposing means 14b can superimpose the entire shape to be measured or the above-mentioned partial shape on the standard shape in the quantitative processing. Also, in the calibration data generation processing, the superimposing means 14b can superimpose standard shapes on each other. In the superimposing, previously acquired position information of the target component in the non-quantitative parameter axis direction (time axis direction) is used.

[0040] The magnification acquisition means 14c can acquire the magnification in the quantitative parameter axis direction (magnification in the signal intensity axis direction) of two data shapes at any position in the non-quantitative parameter axis direction (time axis direction). That is, in the quantitative processing, the magnification of the measurement target shape of the target component relative to a standard shape can be acquired. Also, in the calibration data generation processing, the magnification of one standard shape relative to the other standard shape can be acquired. Then, a representative magnification value can be obtained based on each magnification acquired at each position in the time axis direction.

[0041] In the quantitative processing, the quantitative value calculation means 15 substitutes the magnification value of the measurement target shape obtained by the magnification acquisition means 14c in the comparison means 14 into the regression equation of the calibration data obtained in advance. This makes it possible to calculate the quantitative value.

[0042] In the calibration data generation process, the regression equation generating means 16 can generate a regression equation indicating the relationship between the size in the quantitative parameter axis direction of the standard shape and the quantitative value from the standard shapes of a plurality of standard product data.

[0043] Input means 31 accepts inputs such as start / end instructions and settings relating to data analysis from a user or the like who uses analysis device 10. In addition, the user or the like can use input means 31 to perform the above-mentioned inputs while viewing an image displayed by output means 32 described below.

[0044] The output means 32 outputs the contents input by the input means 31, the contents executed based on the input contents, etc. The output means 32 can also output (display), for example, the data acquired by the data acquisition means 11 and its image. When the quality control of the analysis conditions is also performed by the method according to this embodiment, the output means 32 also outputs the results of the quality check.

[0045] The storage means 33 stores various information required in this embodiment. Specifically, it stores various programs for executing the analysis process in this embodiment, various setting information, etc. The storage means 33 can store not only all data obtained from the data acquisition means 11, but also data that can be generated within the analysis device 10, such as data on each partial shape, and the calculated magnification of the quantitative parameter axis direction of the standard shape. It can also store calibration data (standard shape, regression equation, etc.) generated by the calibration data generation process.

[0046] The control means 17 controls the above-mentioned means 11 to 16 and 31 to 33 of the analysis device 10.

[0047] <Example of hardware configuration of analysis device> In the above-mentioned analysis device 10, an execution program (evaluation program) capable of executing each function on a computer is generated, and the analysis process can be realized by installing the execution program (analysis program) on, for example, a general-purpose PC or a server.

[0048] An example of the hardware configuration of a computer capable of implementing the analysis process in this embodiment is shown in Fig. 3. The analysis device 10 is configured to have an input device 21, an output device 22, a drive device 23, an auxiliary storage device 24, a memory device 25, a CPU (Central Processing Unit) 26 that performs various controls, and a network connection device 27, which are interconnected as a system.

[0049] The input device 21 may be a touch panel, a keyboard, a mouse, or other pointing device operated by a user, etc. The input device 21 may also be a voice input device such as a microphone that allows input by voice, etc.

[0050] The output device 22 may be a monitor, a display, a speaker, etc. The output device 22 may also be a printing device such as a printer.

[0051] The input device 21 and the output device 22 correspond to the above-mentioned input means 11 and output means 12. In addition, when the analysis device 10 is, for example, a smartphone or a tablet terminal, the input device 21 and the output device 22 may be configured as an integrated input / output type such as a touch panel.

[0052] Here, in this embodiment, the execution program installed in analysis device 10 is provided by a portable recording medium 28 such as a USB (Universal Serial Bus) memory or a CD-ROM. Recording medium 28 can be set in drive device 23, and the execution program included in recording medium 28 is installed from recording medium 28 to auxiliary storage device 24 via drive device 23.

[0053] The auxiliary storage device 24 is a storage means such as a hard disk, which stores the execution program of this embodiment and the control program provided in the computer, and can perform input and output as necessary.

[0054] The memory device 25 stores the execution program read from the auxiliary storage device 24 by the CPU 26. The memory device 25 is a ROM (Read Only Memory), a RAM (Random Access Memory), etc. The auxiliary storage device 24 and the memory device 25 may be integrated into one storage device.

[0055] The CPU 26 realizes the analysis process in this embodiment by controlling the overall processing of the computer, such as various calculations and input / output of data to / from each hardware component, based on a control program such as an OS (Operating System) and an execution program stored in the memory device 25. Note that various information required during program execution may be obtained from the auxiliary storage device 24, and the execution results may be stored.

[0056] The network connection device 27 acquires, for example, execution programs and various data from other devices connected to the communication network by connecting to a communication network such as the Internet or a LAN. The network connection device 27 can also provide the execution results obtained by executing a program to other devices.

[0057] <Data analysis method> Next, a data analysis method according to one embodiment of the present invention will be described. This analysis method includes comparing a measurement target shape for a target component with a standard shape for the same target component that has been acquired in advance. Here, the standard shape acquired in advance can be calibration data used to calculate a quantitative value. In addition to the standard shape, a regression equation that expresses the relationship between the quantitative value and the magnitude of the shape in the quantitative parameter axis (intensity axis) direction can be obtained in advance, so that the target component can be quantified using the standard shape and the regression equation as calibration data.

[0058] In the following, first, the generation of the calibration data will be described.

[0059] (Calibration data generation process) FIG. 4 shows a flowchart of a process for generating calibration data (calibration data generation process). In the following example, the data is represented by time, which is a non-quantitative parameter, and signal intensity, which is a quantitative parameter. Such data can be obtained, for example, as a chart showing a time axis and a signal intensity axis. In such a chart, a component can be specified (identified) by matching the position in the time axis direction with the shape.

[0060] The data acquisition means 11 of the analysis device 10 acquires a calibration data generation file (standard data) (S11). The standard data is data of a plurality of samples containing a target component in different known amounts (e.g., known concentrations). This standard data may also include other qualitative information of the target component, such as the position along the time axis where the data of the target component appears (retention time). The standard data can be acquired directly from the analysis device 40 connected to the analysis device 10. Note that the standard data prepared for the quantification of a specific target component has different quantitative values ​​(concentrations, etc.) and was obtained under the same analysis conditions of the analysis device 40.

[0061] The smoothing means 12 smoothes each standard sample data obtained as image data (S12). As mentioned above, the smoothing method is not particularly limited, but it is preferable to use the Savitzky-Golay method. In this case, the number of data points (N value) used for smoothing is preferably 10 or more, although the accuracy improves as the number of data points (N value) used is larger.

[0062] The target component data detection means 13 detects the target component data from each of the smoothed standard product data, and sets the shape of the target component data as a standard shape. The storage means 33 stores the standard shape (S13). In detecting the target component data, position information in the time axis direction, etc., previously obtained, can be used.

[0063] Also, one of the standard shapes can be selected as a representative standard shape, and the storage means 33 stores this representative standard shape (S14). The representative standard shape may be any one of the standard shapes, but it can be the standard shape of the standard product data containing the target component at the highest concentration. In this case, an internal standard substance can be added to each standard sample, and the standard shape of the standard product data containing the target component at the highest concentration can be selected after correcting the concentration with the internal standard substance.

[0064] The comparison means 14 compares the above-mentioned multiple standard shapes with the representative standard shape (S15). Here, the superimposition means 14b in the comparison means 14 superimposes each of the above-mentioned multiple standard shapes on the representative standard shape. At this time, the positions in the time axis direction of each standard shape and the representative standard shape may be matched based on the time axis direction position information of the target component data acquired in advance.

[0065] The magnification acquisition means 14c acquires the magnification of each standard shape relative to the representative standard shape, that is, the magnification of the magnitude (intensity in the quantitative parameter axis direction) (S15). As described above, if the same target component is analyzed by the same analysis system, the magnification will be approximately the same at each position in the time axis direction, but the magnification has an error depending on the position. Therefore, the intensity axis direction magnification (sometimes simply referred to as magnification) relative to the representative standard shape is acquired at each of a plurality of positions in the time axis direction (for example, positions every 0.001 to 0.5 of the width of the shape in the time axis direction). Then, the most frequent value can be regarded as the intensity axis direction magnification representative value (sometimes simply referred to as magnification representative value) of each standard shape.

[0066] The regression equation generating means 16 generates (S16) a regression equation that represents the relationship between the quantitative value (concentration) of the target component and the representative magnification value obtained in the above step (S15) for each standard shape. As an example, FIG. 5 shows the relationship between the concentration of D-asparagine and the intensity axial magnification in the chromatogram, which was obtained for each of the chromatogram peaks of D-asparagine at different concentrations shown in FIG. 1(a).

[0067] The relationship between concentration and magnification can be approximated linearly. However, as shown in FIG. 5, the concentration C and magnification R show a particularly good correlation on a log-log graph. That is, the power regression equation logR=alogC+b (a and b are coefficients) can be approximated as

[0068] The memory means 33 stores the regression equation together with the representative standard shape. At this time, the representative standard shape can also be corrected. For example, the representative standard shape can be corrected based on a shape obtained by resizing each standard shape in the quantitative parameter axis direction by the magnification representative value of the standard shape.

[0069] In this way, the calibration data generating process can generate calibration data including a standard shape and a regression equation. Based on this calibration data, data on components whose quantitative values ​​are unknown can be analyzed and quantified.

[0070] The calibration data generation process is preferably performed each time the analysis conditions are changed. For example, if the analysis device 40 is a chromatograph, the calibration data generation process is preferably performed each time the column or mobile phase is changed. Furthermore, the calibration data generation process is more preferably performed each time a measurement target sample is analyzed.

[0071] (Quantitative processing) Next, a process for performing quantification on a sample in which the quantitative value of a target component is unknown (quantification process) will be described. Fig. 6 shows a flowchart of the quantification process in this embodiment.

[0072] The data acquisition means 11 acquires the calibration data (including the standard shape and the regression equation) obtained by the above-mentioned calibration data generation process (S21). If the calibration data has been stored in the storage means 33, it can also be read out from the storage means 33.

[0073] The data acquisition means 11 acquires data of the measurement target sample containing a target component with an unknown quantitative value from the analysis device 40 (S22). The format of this data is the same as that of the data for generating the calibration data described above. That is, if the standard shape obtained from the calibration data is data represented by time, which is a non-quantitative parameter, and signal intensity, which is a quantitative parameter (for example, a chart represented by a time axis and an intensity axis), the data acquired in this step (S22) is also data represented by time and signal intensity. In addition, the data of the measurement target sample is acquired under the same analysis conditions as the data acquired in generating the calibration data.

[0074] The smoothing means 12 performs a smoothing process (smoothing) on ​​the data (S23). The smoothing process is similar to the smoothing process (S12) described in the calibration data generation process.

[0075] The target component data detection means 13 detects the target component data of the measurement target sample from the smoothed data, and detects the data shape of the target component (measurement target shape) (S24). At this time, the data position information (time axis direction position information) of the target component acquired in advance can be used. The storage means 33 stores the measurement target shape.

[0076] Next, the comparison means 14 compares the measurement object shape with the representative standard shape previously obtained in the calibration data generation process (S25). Then, the magnification of the size (intensity in the quantitative parameter axial direction) of the measurement object shape relative to the representative standard shape is obtained by the comparison (S25). However, when the target component contained in the measurement object sample is in a trace amount, the accuracy of the comparison may be reduced due to errors caused by mechanical and electrical noise of the device. Therefore, as shown below, in order to improve the accuracy of the comparison, the part of the measurement object shape that is most suitable for comparison can be extracted and compared with the standard shape. In addition, the S / N ratio of the measurement object shape containing a trace amount of the target component can be improved by adding a standard sample in advance, thereby preventing a decrease in accuracy.

[0077] FIG. 7 shows a flow diagram illustrating the comparison step (S25) in more detail.

[0078] The partial shape generating means 14a divides the measurement target shape in the time axis direction to form a plurality of divided parts (S251). Then, a partial shape including one or more consecutive divided parts is obtained. In this case, it is preferable to obtain a plurality of partial shapes including a vertex of the measurement target shape (a position in the shape where the magnitude or absolute value of the quantitative parameter is the largest) and including one or more consecutive divided parts. It is also preferable to obtain partial shapes for all combinations of the divided parts.

[0079] 8 is a schematic diagram for explaining the partial shape generation procedure (S251). When dividing the shape to be measured, for example, a range h that is 90% of the height h of the shape is divided. 0.9 (FIG. 8(a)). The number m of divided parts obtained by division is not particularly limited depending on the size of the shape, but it is preferable that the number m is about 5 or more relative to the width of the shape in the time axis direction.

[0080] Furthermore, when the division range is set to a range in which the position of the shape vertex is the center, it is preferable to set m to an odd number so that the shape vertex is included in the central division part. Once the division parts are formed, a shape (partial shape) is obtained that is a range that includes the shape vertex (peak top) and one or more consecutive division parts. An example of a partial shape is shown by the thick line in Figure 8(b). Multiple such partial shapes are generated, and from these partial shapes, a part that is optimal for comparison with the standard shape is found.

[0081] The comparison means 14 compares each of the obtained partial shapes with the representative standard shape (S252). Here, the superimposition means 14b in the comparison means 14 superimposes each partial shape with the representative standard shape. At this time, the positions (vertex positions) in the time axis direction of each partial shape and the representative standard shape are matched based on the time axis direction position information of the target component data acquired in advance.

[0082] When quality control (described later) is also performed in the method according to this embodiment, the comparison means 14 can compare the standard shape with a data shape obtained from data of a quality control sample.

[0083] The magnification obtaining means 14c obtains the intensity axis direction magnification for the representative standard shape at each of a plurality of positions (points) in the time axis direction of each partial shape (S252). At this time, the positions in the time axis direction from which the magnification is obtained may be, for example, 0.01 to 0.5 of the width in the time axis direction. Furthermore, the most frequent value of the obtained magnification is set as the intensity axis direction magnification of each partial shape.

[0084] Then, each partial shape is resized (enlarged or reduced) by the intensity axis direction magnification of each partial shape (S253). At this time, the overlapping means 14b overlaps the partial shape after the resizing with the representative standard shape. Then, the magnification acquisition means 14c obtains the magnification at each position in the time axis direction, obtains the error of this magnification with respect to the intensity axis direction magnification of each partial shape, and calculates the average error per point (per position where the magnification is obtained) for each partial shape (S254).

[0085] In order to further improve accuracy, when comparing each partial shape with the representative standard shape and superimposing them, the magnification at each position can be obtained by shifting the positions of the vertices of the partial shape and the vertices of the representative standard shape in the time axis direction. The average error can be calculated including the error when the vertices are shifted in this way.

[0086] Of the multiple partial shapes, the partial shape with the smallest average error is determined as the shape most suitable for comparison, and the intensity axis direction magnification of that partial shape is determined as the representative value of the intensity axis direction magnification of the shape to be measured (S255).

[0087] Returning to FIG. 6, the quantitative value calculation means 15 substitutes the representative value of the intensity axial magnification into the regression equation previously obtained in the calibration data generation process, and can obtain the quantitative value of the target component in the measurement sample (S26).

[0088] In the above explanation, we have mainly described two-dimensional data represented by two parameters, time and signal intensity, but one embodiment of the present invention can be used to analyze data represented by n parameters, i.e., n-dimensional data.

[0089] As an example, three-dimensional data is shown in FIG. 9. The data in FIG. 9 is a graph obtained by the HPLC-SPE-NMR method for a commercially available formulation of Ginkgo biloba. The data shown in FIG. 9 includes information (parameters) of retention time and signal intensity from a detector obtained by HPLC (high performance liquid chromatography), as well as 1 This is an addition of information (parameters) obtained from H-NMR (nuclear magnetic resonance), and the characteristics of the components in the sample are expressed by a total of three parameters.

[0090] Furthermore, in the analysis method according to this embodiment, quality control (QC) of the analysis conditions can also be performed. That is, the analysis method according to this embodiment may include a quality control step for performing quality control of the analysis conditions. Here, the above analysis conditions include the accuracy of the calibration curve, blank noise / carryover, quantitative accuracy, fractionation accuracy, etc. In addition, prepared quality control data can be used for quality control. This quality control data is data of a validation sample prepared using a standard.

[0091] In one analytical method, in addition to calibration data (standard data) and data on the sample to be measured (actual sample data), quality control data is input, and based on this data, the calibration curve can be created, quality control can be performed, and the target components in the sample to be measured can be identified and quantified.

[0092] For example, a flow chart showing an example of the procedure of an analysis method including a quality control step is shown in FIG. 9a. The series of procedures shown in the figure can be performed using dedicated software. As shown in FIG. 9a, in this method, data for creating a calibration curve, data for quality control, and data of the sample to be measured (actual sample data) are input. Based on the input data, a calibration curve can be created and the lower limit of quantification can be calculated. In addition, a standard shape can be obtained from the data for creating the calibration curve. The standard shape or representative standard shape can be obtained by the method described above for the calibration data generation process.

[0093] Thereafter, the standard shape obtained from the calibration curve creation data is compared with the data shape of the quality control data, i.e., shape fitting can be performed on the quality control data, and the standard shape is compared with the data shape of the real sample data, i.e., shape fitting can be performed on the real sample data.

[0094] The quality and shape fitting results can be output as a quality check sheet using the software's automatic analysis and reporting function, which can display results such as calibration curve precision, blank noise / carryover, fractionation precision, and quantitative precision (see Figures 15A and 15B below).

[0095] In this manner, the analysis method according to this embodiment makes it possible to simultaneously perform an accuracy / quality evaluation of the analysis system and a quantitative analysis, thereby enabling a more accurate analysis to be performed. EXAMPLES

[0096] Example 1 A separation analysis was performed by high performance liquid chromatography-fluorescence detection method using standard human plasma sample A. From the obtained chromatogram, it was confirmed that sample A did not contain D-glutamic acid (arrow in Figure 10(a)). Note that the time axis position (retention time) at which the peak of D-glutamic acid appeared under the same analytical conditions was confirmed in advance.

[0097] Next, 19.75 fmol / inj of D-glutamic acid was added to sample A, and separation analysis was performed under the same analytical conditions to obtain a chromatogram (data) (Figure 10(b)). Based on the peak of D-glutamic acid in the obtained chromatogram, the amount of D-glutamic acid was quantified by the analytical method described above with reference to Figures 4, 6, and 7.

[0098] In this embodiment, in the calibration data generation process, when obtaining the magnification of the standard shape (standard peak) to the representative shape (standard representative peak) (S15), the magnification was obtained every 0.2 seconds along the time axis, and the most frequent value was used as the representative magnification.

[0099] In the quantitative processing, in the generation of partial shapes (partial peaks) (S251), the range of heights up to the shape apex (peak top) was divided into 21, and all patterns of partial shapes (partial peaks) that included one or more of these 21 divided parts and that included consecutive peak tops were regarded as partial shapes. In addition, in the acquisition of the magnification of the partial shape (partial peak) to the representative standard shape (standard representative peak) (S252), the magnification was acquired every 0.2 seconds in the time axis direction, and the most frequent value was obtained.

[0100] The calibration data generation process and the quantification process in this embodiment are shown in detail in the flow charts of FIG. 13 and FIG. 14, respectively.

[0101] Using the above analytical method, the concentration was calculated to be 18.86 fmol / inj. When compared with the amount actually added to sample A, 19.75 fmol / inj, the error (18.86 / 19.75×100) was approximately 4.5%.

[0102] Example 2 Using the human standard plasma sample B, separation and analysis by high performance liquid chromatography was carried out in the same manner as in Example 1. From the obtained chromatogram (FIG. 11(a)), the target component D-serine was quantified by the same analytical method as in Example 1, and the concentration was quantified to be 14.93 fmol / inj.

[0103] Next, 16.46 fmol / inj of D-serine was added to sample B, and separation analysis was performed under the same analytical conditions to obtain a chromatogram (FIG. 11(b)). Based on the obtained chromatogram, D-serine was quantified again using the same analytical method as in Example 1, and the concentration was quantified to be 30.33 fmol / inj.

[0104] The difference in D-serine concentration before and after addition, calculated using the analytical method of the present invention, was 30.33-14.93=13.87 fmol / inj. Comparing this to the increased D-serine concentration of 16.46 fmol / inj due to the addition, the error (13.87 / 14.93×100) can be said to be approximately 7%.

[0105] Example 3 Using human standard plasma sample C, separation and analysis were performed by high performance liquid chromatography in the same manner as in Example 1 to obtain a chromatogram. From the time axis position (retention time) of the target component D-asparagine under the same analytical conditions previously confirmed, a slight peak of D-asparagine was confirmed, which suggested its presence, but quantification was difficult ( FIG. 12(a)).

[0106] After adding 4.12 fmol / inj of D-asparagine to sample C, separation analysis was performed under the same analytical conditions to obtain a chromatogram (FIG. 12(b)). Based on the obtained chromatogram, D-asparagine was quantified again using the same analytical method as in Example 1, and the concentration was quantified to be 4.22 fmol / inj. From this result, the content of D-asparagine contained in sample C can be estimated to be 4.22-4.12=0.1 fmol / inj.

[0107] As described above, there is no established method or standard for drawing a baseline for a peak of a target component, and the baseline can be drawn arbitrarily, so there is a limit to the separation from surrounding impurity components, and the error may exceed 100%. In contrast, an analysis method according to one embodiment of the present invention uses a shape separation means, so that it is possible to quantify a target component of a peak that overlaps with a peak of another adjacent component with an error of 20% or less, about 10%, or even less.

[0108] According to the embodiment of the present invention, it is possible to analyze data on trace components that are difficult to analyze using data obtained by a normal separation and analysis method. Therefore, the embodiment of the present invention can be used to quantify trace components of 1 fmol / inj or less, preferably 100 amol / inj or less. For example, the embodiment of the present invention can be suitably used for the analysis of D-amino acids, peptides, drug metabolites, etc. in vivo, which are contained in a sample in trace amounts.

[0109] Example 4 Example 4 shows an example of quantitative analysis and quality evaluation of the analytical system. In this example, D-serine and L-serine were separated and analyzed by two-dimensional high performance liquid chromatography fluorescence detection using samples for creating standard shapes and calibration curves, samples for quality control (QC), and human blood (actual serum sample) and human urine samples (actual samples) to be analyzed.

[0110] After the above separation analysis, calibration curves were created for each of the human blood and human urine samples from the data of the samples for creating the standard shapes and calibration curves, and the lower limit of quantification was calculated. In addition, standard shapes were also obtained. The obtained standard shapes and calibration curves were applied to the chromatograms of the QC samples, as well as the human blood and human urine samples. This allowed us to check the quality of the analytical conditions (calibration curve accuracy, blank noise / carryover, quantitative accuracy, and first-dimension fractionation accuracy) and calculate the quantitative values ​​of the D-serine and L-serine to be analyzed.

[0111] For quality evaluation, the automatic analysis and reporting function of the dedicated software was used, and a quality check sheet was output. The quality check sheet for blood samples is shown in Figures 15A and 15B, and the quality check sheet for urine samples is shown in Figures 16A and 16B. As shown in Figures 15A and 15B, the quality check sheet evaluates (a) the accuracy of the calibration curve, (b) blank noise / carryover, (c) quantitative accuracy, and (d) fractionation accuracy.

[0112] The separation conditions for each dimension of the two-dimensional high performance liquid chromatography were as follows. <First dimension> Column: KSAARP (3 μm) 1.0 mm inner diameter x 50 mm, 40°C Mobile phase: 2% MeCN 0.1% Formic acid in H2O, 400μL / min <Second Dimension> Column: KSAACSP-001S (5 μm) 1.5 mm inner diameter x 75 mm, 40°C Mobile phase: 0.075% Formic acid in MeOH / MeCN (90 / 10, v / v), 1000μL / min

[0113] The calibration error ranges for the analytical conditions of blood samples were within 3.65% for D-serine and within 3.84% for L-serine, the lower limits of quantification were 0.2 nmol / mL for D-serine and 10 nmol / mL for L-serine, the blank noise was within 4.22% for D-serine and within 0.438% for L-serine, and the errors for QC samples of known concentrations were within 7.35% for D-serine and within 3.35% for L-serine.

[0114] In addition, quantitative analysis of blood samples was performed on 48 specimens. The results are shown in Figures 17A to 17H. As shown in Figures 17A to 17H, a standard shape (dotted line) was superimposed on the chromatogram (fitted part: thick dashed line), and the quantitative values ​​of D-serine and L-serine were calculated from the size of the standard shape (dotted line). The respective quantitative values ​​(nmol / mL) are shown at the top of each chromatogram.

[0115] The calibration error ranges for the analytical conditions for urine samples were within 2.23% for D-serine and within 4.95% for L-serine, the lower limits of quantification were 2 nmol / mL for D-serine and 2 nmol / mL for L-serine, the blank noise was within 2.58% for D-serine and within 3.3% for L-serine, and the errors for QC samples of known concentrations were within 7.6% for D-serine and within 7.07% for L-serine.

[0116] In addition, quantitative analysis of urine samples was performed on 32 specimens. The results are shown in Figures 18A to 18F. As shown in Figures 18A to 18F, a standard shape (dotted line) was superimposed on the chromatogram (fitted part: thick dashed line), and the quantitative values ​​of D-serine and L-serine were calculated from the size of the standard shape (dotted line). The respective quantitative values ​​(nmol / mL) are shown at the top of each chromatogram.

[0117] In this way, the use of shape fitting makes it possible to simultaneously evaluate the accuracy and quality of an analytical system and perform quantitative analysis with a higher throughput than ever before. [Explanation of symbols]

[0118] 10 Analysis device 11 Data Acquisition Methods 12 Smoothing means 13. Means for detecting target component data 14 Means of comparison 14a Partial Shape Generation Method 14b Superposition means 14c Magnification acquisition method 15 Quantitative value calculation method 16 Regression equation generation method 17 Control Means 21 Input Devices 22 Output Devices 23 Drive device 24 Auxiliary storage 25 Memory Device 26 CPU 27 Network connection devices 28 Recording media 31 Input Method 32 Output Method 33 Memory means 40 Analysis device

Claims

1. A method for analyzing data generated from two parameters, comprising the steps of: A comparison step of comparing a data shape of a chromatogram, which is generated from two parameters, time and signal intensity, of a target component in a measurement sample, without a baseline set, with a standard shape generated from the two parameters previously obtained; An analytical method comprising the step of identifying and quantifying a target component in the measurement sample based on the comparison.

2. A method for analyzing data generated from two parameters, comprising the steps of: A comparison step of comparing a data shape of a chromatogram, which is generated from two parameters, time and signal intensity, of a target component in a measurement sample, without a baseline set, with a standard shape generated from the two parameters previously obtained; and calculating a quantitative value of the target component in the measurement sample based on the comparison. An analytical method, wherein the standard shapes include a group of shapes for the target component obtained from a plurality of standard samples each containing the target component at different quantitative values.

3. The analysis method according to claim 1 or 2, wherein in the comparison step, the entire data shape or a partial data shape which is a part of the data shape is obtained, and the partial data shape is compared with the standard shape.

4. The analysis method of claim 3 , wherein the partial data shape is a portion of the data shape and includes a vertex of the data shape.

5. In the quantitative value calculation step, a quantitative value of the target component is calculated based on a regression equation obtained in advance, The analysis method according to claim 2 , wherein the regression equation indicates a relationship between a quantitative parameter axial magnification of the data shape with respect to a representative standard shape selected from the standard shapes and a quantitative value of the target component.

6. In the comparison step, a quantitative parameter axial magnification of the data shape to the representative standard shape is calculated; The analysis method according to claim 5 , wherein in the quantitative value calculation step, the quantitative value of the target component is calculated by substituting the quantitative parameter axial direction magnification into the regression equation.

7. An apparatus for analyzing data generated from two parameters, comprising: A comparison means for comparing a data shape of a chromatogram, which is generated from two parameters, namely, time and signal intensity, of a target component in a measurement sample, without a baseline set, with a standard shape generated from the two parameters previously obtained; and means for identifying and quantifying a target component in the measurement sample based on the comparison.

8. An apparatus for analyzing data generated from two parameters, comprising: a comparison means for comparing a baseline-free data shape of a chromatogram generated from two parameters, time and signal intensity, of a target component in a measurement sample with a standard shape generated from the two parameters previously obtained, the standard shape being a group of shapes for the target component obtained from a plurality of standard samples each containing the target component at different quantitative values; and a quantitative value calculation means for determining a quantitative value of the target component in the measurement sample based on the comparison.

9. An analysis program for causing a computer to function as each of the means included in the analysis device according to claim 7 or 8.

10. A method for creating standard shapes to be used in analyzing data generated from two parameters, including a group of chromatograms generated from two parameters, time and signal intensity, of a target component obtained from multiple standard samples, each containing the target component at different quantitative values, with no baseline set.

Images