Data analysis system and computer program product

Abstract

A data analysis system includes: an analysis data storage unit that associates measured values of a plurality of analysis results obtained by execution using a plurality of analysis conditions with the respective corresponding analysis conditions and stores the measured values as analysis data; a model formula storage unit that stores a model formula that is the basis of a prediction formula that includes the analysis condition as a variable and that is used to calculate a prediction value of the analysis result from the analysis condition; and a data processing unit that performs calculation processing using the analysis data and a model formula, the data processing unit including: a prediction formula generation unit configured to generate a prediction formula relating to the sample by applying measured values of the plurality of analysis results to the model formula together with respective corresponding analysis conditions; and a correlation data generation unit configured to generate correlation data indicating a correlation between a predicted value of the analysis result and a measured value of the analysis result under each of the two or more analysis conditions calculated using the prediction formula.

Description

Technical Field

[0001] The present invention relates to a data analysis system and a computer program product used for searching for the most suitable gradient analysis related to a sample. Background Art

[0002] When developing an analytical method (a collection of analytical conditions such as mobile phase flow rate, column temperature, and gradient conditions) for liquid chromatography analysis of a certain sample, the following operation is generally performed: the user creates multiple analytical methods with different analytical conditions such as gradient conditions, and searches for an analytical method in which the peaks in the chromatogram obtained by analyzing the sample using each analytical method are most separated and the components of each peak are eluted earliest.

[0003] On the other hand, there is also software for assisting in searching for the most suitable analysis conditions for the analysis of the target sample. This software has the following functions: based on the chromatogram obtained by analyzing the target sample under multiple analysis conditions, a prediction formula representing the correlation between the analysis conditions and the analysis results (retention time and peak width) is prepared for each component in the target sample, and based on the prepared prediction formula for each component, the most suitable analysis conditions for the analysis of the target sample are predicted and presented to the user.

[0004] [Prior art literature]

[0005] [Non-patent literature]

[0006] [Non-patent document 1] Weiqiang Hao et al., Computer-aided optimization of multilinear gradient elution in liquid chromatography, Journal of Chromatography A, Volume 1635, 2021, 461754, ISSN 0021-9673)

[0007] [Non-patent document 2] PJ Schoenmakers et al., "Gradient selection in reversed-phase liquid chromatography", Journal of Chromatography, (149 (C) (1978), pp. 519-537)

[0008] [Non-patent document 3] MA Quarry, et al., Prediction of precise isocratic retention data from two or more gradient elution runs. Analysis of some associated errors, Anal. Chem., (58(4) (1986), pp. 907-917)

[0009] [Non-patent document 4] SR Gallant et al., Modeling gradient elution of proteins in ion-exchange chromatography, Journal of the American Institute of Chemical Engineers, (42(9)(1996), pp.2511-2520)

[0010] [Non-patent document 5] JJ Baeza-Baeza, et al., Approaches to model the retention and peak profile in linear gradient reversed-phase liquid chromatography, J. Chromatogr. A, 1284 (2013), pp. 28-35.

[0011] [Non-patent document 6] UD Neue et al., Peak compression in reversed-phase gradient elution, Chromatogr. A, 1111 (2006), pp. 32-39.

[0012] [Non-patent document 7] Weiqiang Hao et al., Influence of the pre-elution of solute in initial mobile phase on retention time and peak compression under linear gradient elution, Journal of Chromatography A, Volume 1618, 2020, 460858, ISSN 0021-9673)

[0013] [Non-patent document 8] H. Poppe, et al., Peak width insolvent-programmed chromatography I. General description of peak broadening insolvent-programmed elution, J. Chromatogr., 204 (1981), pp. 77-84. Summary of the invention

[0014] [Problems to be solved by the invention]

[0015] When using the software to predict the most suitable analysis conditions, according to the analysis conditions and analysis results used in the preparation of the prediction formula, a prediction formula with high reliability is sometimes prepared, and a prediction formula with low reliability is sometimes prepared. If the most suitable analysis conditions for the analysis of the object sample are predicted based on an inaccurate prediction formula with low reliability, then in fact, inappropriate analysis conditions may be predicted as the most suitable analysis conditions, and even if the analysis is performed under the predicted analysis conditions, the possibility of obtaining the chromatogram expected by the user is low. On the contrary, if the most suitable analysis conditions for the analysis of the object sample are predicted based on an accurate prediction formula with high reliability, the predicted analysis conditions are actually most likely to be the most suitable, and if the analysis is performed under the predicted analysis conditions, the possibility of obtaining the chromatogram expected by the user is high.

[0016] However, there is no material for judging whether the reliability of the created prediction formula is high or low, and it is difficult for the user to judge whether the analysis conditions predicted to be optimal for the analysis of the target sample are actually optimal.

[0017] The present invention has been made in view of the above-mentioned problems, and an object of the present invention is to provide users with a material for judging the reliability of a prediction formula.

[0018] [Technical means to solve the problem]

[0019] The data analysis system of the present invention comprises:

[0020] An analysis data storage unit that associates actual measurement values ​​of a plurality of analysis results obtained by respectively using a plurality of analysis conditions with the corresponding analysis conditions and stores the values ​​as analysis data;

[0021] a model formula storage unit storing a model formula as a basis of a prediction formula including an analysis condition as a variable and used to calculate a predicted value of an analysis result according to the analysis condition; and

[0022] a data processing unit that performs calculation processing using the analysis data stored in the analysis data storage unit and the model formula stored in the model formula storage unit,

[0023] The data processing unit includes:

[0024] a prediction formula generating unit configured to generate the prediction formula related to the sample by applying the actual measured values ​​of the plurality of analysis results stored in the analysis data storage unit together with the respectively corresponding analysis conditions to the model formula stored in the model formula storage unit; and

[0025] The correlation data generating unit is configured to generate correlation data indicating a correlation between a predicted value of the analysis result under each of two or more analysis conditions calculated using the prediction formula generated by the prediction formula generating unit and an actual measured value of the analysis result.

[0026] The computer program product of the present invention comprises a computer program, and when the computer program is executed by a processor, the following steps are implemented, including:

[0027] An analysis data storage step of associating actual values ​​of a plurality of analysis results obtained by respectively using a plurality of analysis conditions with the corresponding analysis conditions and storing them as analysis data;

[0028] a model formula storage step of storing a model formula as a basis of a prediction formula, the prediction formula including analysis conditions as variables and used to calculate a predicted value of an analysis result based on the analysis conditions; and

[0029] a data processing step, performing calculation processing using the analysis data stored in the analysis data storage step and the model formula stored in the model formula storage step,

[0030] The data processing steps include:

[0031] a prediction formula generating unit step configured to generate the prediction formula related to the sample by applying the measured values ​​of the plurality of analysis results stored in the analysis data storing unit step together with the respectively corresponding analysis conditions to the model formula stored in the model formula storing unit step; and

[0032] The correlation data generating step is configured to generate correlation data indicating a correlation between a predicted value of the analysis result and an actual measured value of the analysis result for each of two or more analysis conditions calculated using the prediction formula generated by the prediction formula generating step.

[0033] [Effects of the Invention]

[0034] According to the data analysis system of the present invention, since the following correlation data is generated, that is, the correlation data indicating the correlation relationship between the predicted value of the analysis result under each analysis condition calculated using the prediction formula and the actual measured value of the analysis result, the generated correlation data can be used as material for judging the reliability of the prediction formula. Thus, the material for judging the reliability of the prediction formula can be provided to the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a block diagram showing an embodiment of the structure of a data analysis system.

[0036] Figure 2 This is a diagram showing an example of a screen displaying reliability information.

[0037] Figure 3 This is a flowchart showing an example of data analysis processing performed in the same embodiment.

[0038] [Explanation of Symbols]

[0039] 1: Data analysis system

[0040] 2: Analysis data storage unit

[0041] 4: Model formula storage unit

[0042] 6: Data Processing Department

[0043] 8: Display DETAILED DESCRIPTION

[0044] Hereinafter, an embodiment of a data analysis system will be described with reference to the drawings.

[0045] The data analysis system 1 is a system constructed by introducing a computer program into a computer device, and includes an analysis data storage unit 2 , a model formula storage unit 4 , a data processing unit 6 , and a display 8 .

[0046] The analysis data storage unit 2 is a storage area for correlating a plurality of analysis results obtained by analyzing a sample using a plurality of different gradient conditions in the analysis device 100 with the gradient conditions and storing them as analysis data. The analysis results include chromatogram information related to the retention time of each component in the sample and the peak width of the peak of each component. The so-called gradient condition refers to the condition of how to make the ratio of the solvent constituting the mobile phase change over time. In addition, in the embodiment, the gradient condition is used as an example to explain the analysis condition, but the present invention is not limited to the gradient condition.

[0047] The model formula storage unit 4 stores a model formula for each component in the sample, which is a basis for a retention time prediction formula representing the relationship between gradient conditions and retention time, and a model formula for a basis for a peak width prediction formula representing the relationship between gradient conditions and peak width.

[0048] The analysis data storage unit 2 and the model formula storage unit 4 are implemented by a part of the area of ​​an information storage device such as a hard disk drive. The analysis device 100 is, for example, a liquid chromatograph.

[0049] The data processing unit 6 includes a prediction formula generating unit 10 and a related data generating unit 12. The data processing unit 6 is implemented by a computer circuit including a central processing unit (CPU), and the prediction formula generating unit 10 and the related data generating unit 12 are functions obtained by executing a computer program in the data processing unit 6.

[0050] The prediction formula generation unit 10 is configured to generate a retention time prediction formula and a peak width prediction formula related to each component in the sample by performing statistical analysis (fitting) on ​​the model formula read from the model formula storage unit 4 by applying the chromatogram information of the sample read from the analytical data storage unit 2. As an algorithm for statistical analysis, Bayesian inference or the like can be used.

[0051] Here, the model formula is explained.

[0052] First, an example of a model formula serving as a basis for the retention coefficient prediction formula will be described.

[0053] The migration speed of the compound can be expressed by the inverse of the retention coefficient, the retention time (t R ) and the holding coefficient k(t) at each moment can be expressed by the following [Formula 1] (see Non-Patent Document 1).

[0054] [Formula 1]

[0055]

[0056] Here, t p is the dwell time (data acquisition time), t0 is the hold-up time (the time from the introduction of the sample to the peak of the component not retained by the column). k(0) represents the retention factor in the mobile phase composition before time 0. Here, it is assumed that the mobile phase composition is fixed before the analysis starts and the retention factor does not change.

[0057] It is known that the relationship between the organic solvent concentration and the retention factor in reverse phase analysis in a liquid chromatograph can be expressed by the following [Formula 2] (see Non-Patent Documents 1 to 5).

[0058] [Formula 2]

[0059] lnk(t)=lnk0-S1φ(t)+S2{φ(t)} 2

[0060] Here, k(t) is the retention coefficient at each time, φ(t) is the organic solvent ratio of the mobile phase at each time, and k0 is the retention coefficient when the organic solvent ratio is 0. S1 and S2 are coefficients, which are values ​​that vary depending on the solute, stationary phase, and mobile phase.

[0061] The prediction formula generation unit 10 applies the chromatogram information and gradient information of the sample to the model formulas [Formula 1] and [Formula 2] and performs fitting to obtain k0, S1, and S2 for each component in the sample, thereby generating a retention time prediction formula for each component in the sample.

[0062] Next, an example of a model formula serving as a basis for the peak width prediction formula will be described.

[0063] In reversed-phase gradient analysis, the peak width tends to decrease relative to the retention time, a phenomenon known as peak compression. Peak compression can be expressed by the following [Formula 3] using a compression factor G (see Non-Patent Document 6).

[0064] [Formula 3]

[0065]

[0066] Here, W is the peak width, t0 is the retention time, k(t R ) is the retention factor during dissolution, and N is the theoretical number of stages of the column.

[0067] The retention factor k(t) can be obtained using the same algorithm as that used to generate the retention factor prediction formula. However, the peak width has a large deviation compared to the retention time, and is sometimes too fit to the noise. Therefore, the prediction formula generation unit 10 applies the chromatogram information and gradient information to the following [Formula 4] ignoring the quadratic term of [Formula 2] and performs fitting, thereby obtaining k0 and S1, and obtaining the peak width prediction formula for each component in the sample.

[0068] [Formula 4]

[0069] lnk(t)=lnk0-S1φ(t)

[0070] In addition, the compression coefficient G can be calculated using the following formula [5] using the retention coefficient (see Non-Patent Document 7-Non-Patent Document 8).

[0071] [Formula 5]

[0072]

[0073] The related data generating unit 12 is configured to use the retention time prediction formula and the peak width prediction formula generated based on the model formula to calculate the predicted value of the retention time and the predicted value of the peak width for each gradient condition related to each component in the sample, and generate related data representing the correlation between each calculated predicted value and the actual measured value under the corresponding gradient condition (the retention time value and the peak width value in the analysis data).

[0074] The data processing unit 6 is connected to a display 8 , and the correlation data generated by the correlation data generating unit 12 is displayed on the display 8 , for example, when the user wishes.

[0075] Figure 2 An example of related data displayed on the display 8 is shown.

[0076] Figure 2 The correlation data shown in is data related to a component (compound) specified by the user. The correlation data includes retention time (RT) prediction formula information and peak width prediction formula information for each component in the sample.

[0077] The retention time prediction formula information and the peak width prediction formula information each include: a two-dimensional graph in which one of two mutually orthogonal axes is used as the axis of the measured value under each gradient condition and the other axis is used as the axis of the predicted value under each gradient condition; a table showing the predicted values ​​and the measured values ​​under each gradient condition; and a determination coefficient associated with each prediction formula.

[0078] Each curve (plot) on the two-dimensional chart represents the intersection of the predicted value and the measured value under each gradient condition. When the prediction formula is suitable under all gradient conditions (the predicted value takes a value close to the measured value = appropriate), the curve of the measured value is located on the straight line x=y with a slope of 1 in the two-dimensional chart. On the other hand, in the case of a gradient condition where the prediction formula is not suitable (inappropriate), there is a curve of the measured value that deviates from the straight line x=y in the two-dimensional chart. In this way, if you observe the two-dimensional chart, it is clear at a glance under which gradient condition the prediction formula has high reliability and under which gradient condition the reliability is low, and the user can easily identify the reliability of the prediction formula. Furthermore, by showing a table of predicted values ​​and measured values ​​under each gradient condition at the same time, the user can easily confirm to what extent the predicted value deviates from the measured value.

[0079] In addition, the "coefficient of determination" shown in each of the so-called retention time prediction formula information and the peak width prediction formula information is a statistic used to evaluate whether the prediction formula is appropriate. As such a statistic, the Rhat statistic is known. If the inference of the prediction formula is appropriate, the Rhat statistic is approximately 1. If the Rhat statistic exceeds 1.1, although the inference is made, it is generally evaluated as not being an appropriate inference. Rhat is a value obtained when using Bayesian inference as a statistical analysis algorithm for fitting used in the generation of the prediction formula. If the variation of the measured value can be fully explained by the predicted value obtained by the prediction formula, the statistic becomes 1, and the more variations that cannot be explained (the worse the prediction performance of the prediction formula), the statistic becomes a value smaller than 1. By referring to such numerical values, users can also confirm the reliability of each prediction formula.

[0080] Next, use Figure 3 The flowchart of FIG. 1 illustrates the data analysis processing performed by the data processing unit 6.

[0081] The prediction formula generation unit 10 of the data processing unit 6 reads the analysis data of the sample to be analyzed from the analysis data storage unit 2 (step 101), and further reads the model formula used in the generation of the prediction formula from the model formula storage unit 4 (step 102). Thereafter, the data processing unit 6 applies the analysis data to the model formula, performs fitting using a predetermined statistical analysis algorithm, thereby determining each coefficient of the model formula, and generates a retention time prediction formula and a peak width prediction formula (step 103).

[0082] When the retention time prediction formula and the peak width prediction formula are generated, the predicted values ​​of retention time and peak width under each gradient condition are calculated using the generated prediction formulas, and the calculated predicted values ​​and the measured values ​​are used to generate Figure 2 The related data generated here are displayed on the display 8 when the user wishes, for example.

[0083] In addition, the embodiment described above is only an example of an embodiment of the present invention. Regarding the analysis conditions, the relevant data under each gradient condition are described, but the present invention is not limited to this and can also be applied to other analysis conditions (column oven temperature or flow rate, etc.). The embodiment of the data analysis system and computer program of the present invention is as follows.

[0084] In one embodiment of the data analysis system of the present invention, the system comprises: an analysis data storage unit for associating actual measured values ​​of a plurality of analysis results obtained by respectively using a plurality of analysis conditions with the corresponding analysis conditions and storing the values ​​as analysis data;

[0085] a model formula storage unit storing a model formula as a basis of a prediction formula including an analysis condition as a variable and used to calculate a predicted value of an analysis result according to the analysis condition; and

[0086] a data processing unit that performs calculation processing using the analysis data stored in the analysis data storage unit and the model formula stored in the model formula storage unit,

[0087] The data processing unit includes:

[0088] a prediction formula generating unit configured to generate the prediction formula related to the sample by applying the actual measured values ​​of the plurality of analysis results stored in the analysis data storage unit together with the respectively corresponding analysis conditions to the model formula stored in the model formula storage unit; and

[0089] The correlation data generating unit is configured to generate correlation data indicating a correlation between a predicted value of the analysis result under each of two or more analysis conditions calculated using the prediction formula generated by the prediction formula generating unit and an actual measured value of the analysis result.

[0090] In an embodiment [1] of the one embodiment, the correlation data includes a two-dimensional graph in which one of two intersecting axes is an axis of actual values ​​of the analysis results under each of the analysis conditions, and the other is an axis of predicted values ​​of the analysis results under each of the analysis conditions.

[0091] In implementation [2] of the one embodiment, the relevant data includes reliability information of the prediction formula, and the reliability information of the prediction formula also takes into account the degree of deviation between the measured value of the analysis result related to the plurality of analysis conditions and the predicted value of the analysis result. The implementation [2] may be combined with the implementation [1].

[0092] In embodiment [3] of the one embodiment, the analysis result includes a retention time of each component contained in the sample. The embodiment [3] can be combined with the embodiment [1] and / or the embodiment [2].

[0093] In embodiment [4] of the one embodiment, the analysis result includes the peak width of each component contained in the sample. Embodiment [4] can be combined with embodiment [1], embodiment [2] and / or embodiment [3].

[0094] In the embodiment [5] of the embodiment, a display (8) is further included, and the data processing unit (6) is configured to display the relevant data generated by the relevant data generating unit (12) on the display (8). The embodiment [5] can be combined with the embodiment [1], the embodiment [2], the embodiment [3] and / or the embodiment [4].

[0095] In embodiment [6] of the one embodiment, the two or more analysis conditions may include analysis conditions other than the plurality of analysis conditions. This embodiment makes it possible to confirm whether the prediction of the analysis result under the analysis conditions other than the analysis conditions used in generating the prediction formula is appropriate. Embodiment [6] may be combined with embodiment [1], embodiment [2], embodiment [3], embodiment [4] and / or embodiment [5].

[0096] In one embodiment of the computer program of the present invention, the data analysis system is constructed by being introduced into a computer.

Claims

1. A data analysis system, comprising: An analysis data storage unit that associates actual measurement values ​​of a plurality of analysis results obtained by respectively using a plurality of analysis conditions with the corresponding analysis conditions and stores the values ​​as analysis data; a model formula storage unit storing a model formula serving as a basis for a prediction formula, the prediction formula including an analysis condition as a variable and used to calculate a predicted value of an analysis result based on the analysis condition; as well as a data processing unit that performs calculation processing using the analysis data stored in the analysis data storage unit and the model formula stored in the model formula storage unit, The data processing unit includes: a prediction formula generating unit configured to generate the prediction formula related to the sample by applying the actual measured values ​​of the plurality of analysis results stored in the analysis data storage unit together with the respectively corresponding analysis conditions to the model formula stored in the model formula storage unit; as well as The correlation data generating unit is configured to generate correlation data indicating a correlation between a predicted value of the analysis result and an actually measured value of the analysis result for each of two or more analysis conditions calculated using the prediction formula generated by the prediction formula generating unit.

2. A data analysis system according to claim 1, wherein the related data includes a two-dimensional graph in which one of two intersecting axes is an axis of actual values ​​of the analysis results under each of the analysis conditions, and the other is an axis of predicted values ​​of the analysis results under each of the analysis conditions.

3. The data analysis system according to claim 1, wherein the relevant data includes reliability information of the prediction formula, and the reliability information of the prediction formula takes into account the degree of deviation between the actual measured value of the analysis result related to the multiple analysis conditions and the predicted value of the analysis result. 4 . The data analysis system according to claim 1 , wherein the analysis result includes a retention time of each component contained in the sample. 5 . The data analysis system according to claim 1 , wherein the analysis result includes a peak width of each component contained in the sample.

6. The data analysis system according to claim 1, further comprising a display, The data processing unit is configured to display the correlation data generated by the correlation data generating unit on the display. 7 . The data analysis system according to claim 1 , wherein the two or more analysis conditions include analysis conditions other than the plurality of analysis conditions.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the following steps are implemented, including: An analysis data storage step of associating actual values ​​of a plurality of analysis results obtained by respectively using a plurality of analysis conditions with the corresponding analysis conditions and storing the values ​​as analysis data; a model formula storage step of storing a model formula as a basis of a prediction formula, the prediction formula including analysis conditions as variables and used to calculate a predicted value of an analysis result based on the analysis conditions; and a data processing step, performing calculation processing using the analysis data stored in the analysis data storage step and the model formula stored in the model formula storage step, The data processing steps include: a prediction formula generating unit step configured to generate the prediction formula related to the sample by applying the measured values ​​of the plurality of analysis results stored in the analysis data storing unit step together with the respectively corresponding analysis conditions to the model formula stored in the model formula storing unit step; and The correlation data generating step is configured to generate correlation data indicating a correlation between a predicted value of the analysis result for each of two or more analysis conditions calculated using the prediction formula generated by the prediction formula generating step and an actual measured value of the analysis result.

Images