Fluorescence analysis device and fluorescence analysis method

Abstract

The present application relates to a fluorescence analysis device and a fluorescence analysis method. The fluorescence analysis device is a device that analyzes measurement data before operation processing, which is obtained by detecting fluorescence generated when excitation light is respectively irradiated on a plurality of samples each labeled with any one of a plurality of fluorescent pigments, through a plurality of detection channels, and identifies the samples based on the kind of the fluorescent pigment labeled. The fluorescence analysis device includes a first determination section that determines whether the plurality of samples can be separated into two or more groups based on measurement data obtained by one or more of the plurality of detection channels, and a first separation section that separates any one group from among the plurality of samples when the first determination section determines that the plurality of samples can be separated into two or more groups.

Description

Technical Field

[0001] This disclosure relates to fluorescence analysis apparatus and fluorescence analysis methods. Background Technology

[0002] Fluorescence analysis techniques that identify multiple samples labeled with any one of several fluorescent dyes, such as in flow cytometry (FCM), are important. FCM arranges a large number of fluorescently labeled samples (e.g., tiny particles like cells) one by one and allows them to flow in a flow path. The intensity of fluorescence and scattered light produced when each sample flowing in the flow path is illuminated with a laser is measured. By analyzing the measurement data obtained from this detection, the samples can be identified.

[0003] In FCM (Fluorescent Chromatography), multiple fluorescent dyes, each specifically binding to samples belonging to a particular group, are used. Each of these dyes is labeled with one of the multiple fluorescent dyes, and the intensity of fluorescence generated when each sample is irradiated with a laser is measured to obtain measurement data. During this fluorescence detection, a detection device with multiple detection channels that can selectively detect fluorescence in specific wavelengths is used to measure the intensity of fluorescence in each wavelength band. This measurement is called multicolor measurement. By analyzing the measurement data, samples can be identified according to the type of fluorescent dye labeled, that is, samples can be identified by each group. This analysis is called multicolor analysis.

[0004] In multicolor measurements, fluorescence produced by a certain fluorescent dye may be detected not only by a detection channel corresponding to that type of fluorescent dye, but also by one or more other detection channels corresponding to other types of fluorescent dyes. This phenomenon is called fluorescence spillover. If fluorescence spillover occurs, even a sample originally labeled with one type of fluorescent dye will appear as if it were labeled with other types of fluorescent dyes, resulting in errors in the multicolor analysis results using this measurement data. To address this problem, before sample identification, a calculation is performed to subtract the fluorescence intensity of the spillover portion from the fluorescence measurement data obtained from each detection channel. Examples of such calculations include fluorescence compensation as described in Japanese Patent Application Publication No. 2011-232254 and No. 2011-232259, and fluorescence intensity calculation (fluorescence unmixing) as described in Japanese Patent No. 7564408. Summary of the Invention

[0005] To address the issue of fluorescence spillover, computational processing is performed, which can lead to increased measurement data bias and decreased analytical precision. Specifically, in a cytogram (cell distribution map) created by plotting measurement data from two detection channels on a two-dimensional graph, multiple samples labeled with a single fluorescent dye are displayed as a single cluster. When fluorescence spillover occurs, even samples labeled with a single fluorescent dye may appear labeled with other types of fluorescent dyes within the cluster on the cytogram. After computational processing, the cluster on the cytogram will appear labeled with a single fluorescent dye, but the increased bias in the measurement data results in a wider distribution area compared to the unprocessed cluster. When two clusters exist on the cytogram, even if they were separable before computational processing, they become difficult to separate due to their wider distribution area after processing. Consequently, analytical precision deteriorates. This problem becomes more pronounced with the use of more fluorescent dyes.

[0006] This disclosure was made to address the aforementioned problems, and aims to provide an apparatus and method for performing high-precision fluorescence analysis even when using multiple fluorescent dyes.

[0007] The fluorescence analysis device disclosed herein analyzes measurement data before computational processing and identifies samples based on the type of labeled fluorescent pigment. The measurement data before computational processing is obtained by detecting the fluorescence generated when multiple samples labeled with any one of a variety of fluorescent pigments are irradiated with excitation light through multiple detection channels.

[0008] The first aspect of the fluorescence analysis apparatus disclosed herein comprises: (1) a first determination unit that determines, based on measurement data obtained from any one or more detection channels among a plurality of detection channels, whether multiple samples can be separated into two or more groups; and (2) a first separation unit that, if determined by the first determination unit to be able to separate into two or more groups, separates any one of the groups. Furthermore, the processing of the first determination unit and the first separation unit is repeated for multiple samples other than those already identified until the first determination unit determines that they cannot be separated into two or more groups, and the samples contained in the groups determined to be unseparable are identified as samples labeled with any fluorescent dye.

[0009] The second aspect of the fluorescence analysis apparatus disclosed herein may further include, based on the first aspect,: (3) a second judgment unit, which, for multiple samples contained in the separated cluster, determines whether they can be further separated into two or more clusters based on measurement data obtained from any one or more detection channels among multiple detection channels; and (4) a second separation unit, which, if determined by the second judgment unit to be further separated into two or more clusters, separates any one of them. Furthermore, it is also possible to repeatedly process multiple samples other than the already identified samples using the second judgment unit and the second separation unit until the second judgment unit determines that they cannot be separated into two or more clusters, and then identify the samples contained in the clusters determined to be unseparable as samples labeled with any fluorescent dye.

[0010] The third embodiment of the fluorescence analysis apparatus disclosed herein may further include, based on the first or second embodiment, an identification unit that identifies a sample as being labeled with any fluorescent dye based on a fluorescence spectrum obtained from measurement data obtained from multiple detection channels.

[0011] In the fourth aspect of the fluorescence analysis apparatus disclosed herein, it is also possible that, based on any of the first to third aspects, the first judgment unit makes a judgment based on measurement data, wherein the measurement data is obtained by detecting the fluorescence generated when multiple samples labeled with any one of a variety of fluorescent dyes are irradiated with excitation light of multiple wavelengths by multiple detection channels.

[0012] In the fifth embodiment of the fluorescence analysis apparatus disclosed herein, it is also possible that, based on any of the first to fourth embodiments, the first judgment unit makes a judgment after processing the measurement data into bins.

[0013] The fluorescence analysis method disclosed herein is a method of analyzing measurement data before computational processing and identifying samples based on the type of fluorescent pigment labeled. The measurement data before computational processing is obtained by detecting the fluorescence generated when multiple samples labeled with any one of a variety of fluorescent pigments are irradiated with excitation light through multiple detection channels.

[0014] The first aspect of the fluorescence analysis method disclosed herein includes: (1) a first judgment step, which, based on measurement data obtained from any one or more detection channels among a plurality of detection channels, determines whether multiple samples can be separated into two or more groups; and (2) a first separation step, which, if it is determined in the first judgment step that they can be separated into two or more groups, separates any one of the groups. Furthermore, the first judgment step and the first separation step are repeatedly performed on multiple samples other than the already identified samples until it is determined in the first judgment step that they cannot be separated into two or more groups, and the samples contained in the groups that are determined to be unseparable are identified as samples labeled with any fluorescent dye.

[0015] The second aspect of the fluorescence analysis method disclosed herein may further include, based on the first aspect, the following: (3) a second judgment step, for multiple samples contained in the separated cluster, determining whether they can be further separated into two or more clusters based on measurement data obtained from any one or more detection channels; and (4) a second separation step, in which, if it is determined in the second judgment step that they can be further separated into two or more clusters, any one of them is separated. Furthermore, it is also possible to repeatedly perform the second judgment step and the second separation step on multiple samples other than the already identified samples until it is determined in the second judgment step that they cannot be separated into two or more clusters, and then identify the samples contained in the clusters determined to be unseparable as samples labeled with any fluorescent dye.

[0016] The third aspect of the fluorescence analysis method disclosed herein may further include, based on the first or second aspect, an identification step, wherein for a sample identified as being labeled with any fluorescent dye, identification is performed based on a fluorescence spectrum obtained from measurement data obtained from multiple detection channels.

[0017] In the fourth aspect of the fluorescence analysis method disclosed herein, it is also possible that, based on any of the first to third aspects, in the first judgment step, the judgment is made based on measurement data. The measurement data is obtained by detecting the fluorescence generated when each of the multiple samples labeled with any one of the multiple fluorescent dyes is irradiated with excitation light of multiple wavelengths through multiple detection channels.

[0018] In the fifth aspect of the fluorescence analysis method disclosed herein, it is also possible to make a judgment in the first judgment step after the measurement data has been binned, based on any of the first to fourth aspects.

[0019] According to this disclosure, high-precision fluorescence analysis can be performed even when using multiple fluorescent dyes. Attached Figure Description

[0020] Figure 1 This is a diagram showing the structure of the fluorescence analysis device 10.

[0021] Figure 2 This is an example of a flowchart for a fluorescence analysis method.

[0022] Figure 3 This is another example of a flowchart for a fluorescence analysis method.

[0023] Figure 4 This is a diagram showing an example of the fluorescence spectra of fluorescent pigments A through C.

[0024] Figure 5 This is a diagram that schematically represents four types of samples.

[0025] Figure 6 This is a graph showing the cell distribution of the measurement data in the first cycle.

[0026] Figure 7 It is a graph showing the cell distribution data of the group X (group A) separated in the first cycle.

[0027] Figure 8 (a) is a graph showing the fluorescence spectra of the four groups A to D. Figure 8 (b) is a graph showing the fluorescence spectrum of only the isolated group A. Figure 8 (c) is a graph showing the fluorescence spectra of other groups B to D.

[0028] Figure 9 (a) is a diagram showing the distribution of cells in ch2 and ch7 when the multiple groups containing group X in the cell distribution diagram C1 are divided into four groups A to D. Figure 9 (b) is a graph showing the distribution of only group X (group A) in cell distribution map C1 for ch2 and ch7. Figure 9 (c) is a diagram showing the distribution of cells in ch2 and ch7 when the other groups in the cell distribution diagram C1 are divided into groups B to D and represented separately.

[0029] Figure 10 This is a graph showing the cell distribution of the measurement data in the second cycle (except for the measurement data of the separated group A).

[0030] Figure 11 This is a graph showing the cell distribution data of the separated group Y (group D) in the second cycle.

[0031] Figure 12 (a) is a graph showing the fluorescence spectra of the three groups B to D. Figure 12 (b) is a graph showing the fluorescence spectrum of only the isolated group D. Figure 12 (c) is a graph showing the fluorescence spectra of other groups B and C.

[0032] Figure 13 (a) is a diagram showing the distribution of cells in ch6 and ch7 when the multiple groups containing group Y in the cell distribution diagram C2 are divided into three groups B to D, each representing a different group. Figure 13 (b) is a graph showing the distribution of only group Y (group D) in cell distribution map C2 for ch6 and ch7. Figure 13 (c) is a diagram showing the distribution of cells in ch6 and ch7 when the other groups in the cell distribution diagram C2 are divided into groups B and C respectively.

[0033] Figure 14 It is a graph representing the cell distribution data of the measurement data in the third cycle (but excluding the measurement data of the separated groups A and D).

[0034] Figure 15 This is a graph showing the cell distribution data of a single group (group C) separated in the third cycle.

[0035] Figure 16 (a) is a graph showing the fluorescence spectra of the two groups B and C respectively. Figure 16 (b) is a graph showing only the fluorescence spectrum of the isolated group C. Figure 16 (c) is a graph showing the fluorescence spectrum of other group B.

[0036] Figure 17 (a) is a diagram showing the distribution of cells in ch5 and ch6 when the two groups in C3 are divided into groups B and C respectively. Figure 17 (b) is a diagram showing the distribution of only one group (group C) in cell distribution map C3 for ch5 and ch6. Figure 17 (c) is a diagram showing the distribution of other groups (group B) in cell distribution map C3 for ch5 and ch6.

[0037] Figure 18 This is a graph showing the cell distribution of measurement data from the fourth cycle (excluding the measurement data of the separated groups A, C, and D).

[0038] Figure 19 This is a graph representing the cell distribution of the measurement data from the second cycle (excluding the measurement data from the separated group X (group A)).

[0039] Figure 20 It is a graph representing the cell distribution data of the groups separated in the second cycle (including one group Z on the surface of groups B and C).

[0040] Figure 21 (a) is a diagram showing the distribution of cells in ch2 and ch7 when multiple groups in C4 are divided into three groups B to D. Figure 21 (b) is a diagram showing the distribution of cells in ch2 and ch7, in C4, which contains groups B and C and is ostensibly a single group Z.

[0041] Figure 22 Figures (a) to (f) show the fluorescence spectra of the six isolated groups.

[0042] Figure 23 (a) through (d) illustrate how fitting processes are used to identify and display [the desired feature]. Figure 22 (a) A graph showing the fluorescence spectra of the group of samples contained in the method and results.

[0043] Figure 24 (a) through (d) illustrate how fitting processes are used to identify and display [the desired feature]. Figure 22 Figures showing the results of the fluorescence spectra of the samples contained in the group, (b) to (e). Detailed Implementation

[0044] Hereinafter, with reference to the accompanying drawings, the methods for carrying out this disclosure will be described in detail. Furthermore, in the description of the drawings, the same reference numerals are used to denote the same elements, and repeated descriptions are omitted. This disclosure is not limited to these illustrations, but is represented by the claims, and includes all modifications within the meaning and scope equivalent to the claims.

[0045] Figure 1 This is a diagram showing the structure of the fluorescence analysis device 10. The fluorescence analysis device 10 and the detection device 20 together constitute a flow cytometer.

[0046] The detection device 20 arranges a large number of fluorescently labeled samples (e.g., tiny particles such as cells) one by one and flows them in a flow path. It detects the intensity of fluorescence and scattered light generated when the samples flowing in the flow path are irradiated with a laser (excitation light), and acquires measurement data obtained through this detection. The detection device 20 has multiple detection channels that can selectively detect fluorescence in specific wavelength bands. These multiple detection channels for fluorescence detection, as a whole, can detect continuous fluorescence spectra or fluorescence spectra other than a certain range of wavelengths. The detection device 20 also has detection channels for detecting side scatter (SSC) and forward scatter (FSC) light generated when irradiated with a laser. The laser can also be of multiple wavelengths.

[0047] The fluorescence analysis device 10 inputs measurement data obtained by the detection device 20, analyzes and processes the measurement data before processing, and identifies the sample based on the type of fluorescent dye labeled. The fluorescence analysis device 10 can be a computer. The fluorescence analysis device 10 includes a first judgment unit 11, a first separation unit 12, a second judgment unit 13, a second separation unit 14, an identification unit 15, an input unit 16, a display unit 17, and a storage unit 18.

[0048] The input unit 16 inputs measurement data obtained from the detection device 20, as well as analysis conditions, etc. The display unit 17 displays the status of the analysis process and the analysis results on the screen. The display mode of the display unit 17 is arbitrary; for example, it can be a cell distribution map or a histogram. The storage unit 18 stores measurement data input from the detection device 20 before processing, as well as data during the analysis process and data at the end of the analysis. Furthermore, the storage unit 18 stores programs for executing the processing of the first judgment unit 11, the first separation unit 12, the second judgment unit 13, the second separation unit 14, and the identification unit 15. The processing of the first judgment unit 11, the first separation unit 12, the second judgment unit 13, the second separation unit 14, and the identification unit 15 will be described later.

[0049] Figure 2 This is an example of a flowchart for a fluorescence analysis method. In the flowchart shown, the fluorescence analysis method includes a first determination step S1, a first separation step S2, and an identification step S5. The first determination step S1 is performed by a first determination unit 11. The first separation step S2 is performed by a first separation unit 12. The identification step S5 is performed by an identification unit 15. Alternatively, in this fluorescence analysis method, the fluorescence analysis apparatus 10 may not include a second determination unit 13 and a second separation unit 14.

[0050] In the first determination step S1, the first determination unit 11 determines, based on the measurement data from one or more of the multiple detection channels of the detection device 20, whether it is possible to separate the multiple samples flowing in the flow path into two or more groups. In the first separation step S2, if the first determination unit 11 determines that it is possible to separate them into two or more groups, the first separation unit 12 separates one of the groups.

[0051] For multiple samples other than the already identified samples, the first judgment step S1 and the first separation step S2 are repeatedly processed until it is determined in the first judgment step S1 that they cannot be separated into more than two groups. The samples contained in the groups that are determined to be inseparable are then identified as samples labeled with any fluorescent dye.

[0052] In the identification step S5, the identification unit 15 identifies samples labeled with any fluorescent dye based on the fluorescence spectrum obtained from the measurement data obtained from multiple detection channels.

[0053] Figure 3 This is another example of a flowchart for a fluorescence analysis method. In the flowchart shown, the fluorescence analysis method includes a first judgment step S1, a first separation step S2, and an identification step S5, as well as a second judgment step S3 and a second separation step S4. The second judgment step S3 is performed by the second judgment unit 13. The second separation step S4 is performed by the second separation unit 14.

[0054] In the second determination step S3, the second determination unit 13, based on measurement data obtained from any one or more detection channels among the multiple samples contained in the group separated by the first separation unit 12 or the second separation unit 14, determines whether they can be further separated into two or more groups. In the second separation step S4, if the second determination unit 13 determines that they can be further separated into two or more groups, the second separation unit 14 separates any one of the groups.

[0055] For multiple samples other than those already identified, the second judgment step S3 and the second separation step S4 are processed repeatedly until it is determined in the second judgment step S3 that they cannot be separated into more than two groups. If it is determined in the second judgment step S3 that they cannot be separated into more than two groups, identification is performed in the identification step S5. That is, the samples contained in the groups determined to be inseparable are identified as samples labeled with any fluorescent dye.

[0056] Figure 2 and Figure 3 The flowchart of the fluorescence analysis method shown is merely illustrative, and the flowchart of the fluorescence analysis method can also be in other forms. The identification step S5 can also identify the samples contained in a group whenever a group is determined to be inseparable through the first judgment step S1 or the second judgment step S3. Alternatively, if it is determined in the second judgment step S3 that the group cannot be separated into more than two groups, identification can be omitted in the identification step S5, and the process can return to the first judgment step S1 to reselect different groups. In the first judgment step S1 or the second judgment step S3, the judgment can be made after binning the measurement data. When separating a group from other groups in the display of the cell distribution map, an arbitrary threshold can be set. For example, when the threshold is set to 0 counts, a region can be defined that surrounds the area with a 0 count and contains no other groups, and one group within that region can be separated.

[0057] The first judgment step S1, the first separation step S2, the second judgment step S3, and the second separation step S4 can each be performed based on the cell distribution map, histogram, etc., displayed on the display unit 17, or they can be performed automatically through numerical calculations without being displayed on the display unit 17. When the cell distribution map, histogram, etc., are displayed on the display unit 17, judgment and separation can be performed automatically or manually by the operator. When judgment and separation are performed automatically through numerical calculations, unsupervised learning machine learning methods can be used to perform judgment and separation on the measurement data without prior knowledge.

[0058] Next, use Figures 4-18 Further explanation Figure 2 The fluorescence analysis method is shown. Here, the number of detection channels used for fluorescence detection is set to 12, and the 12 detection channels are sequentially designated as ch1~ch12 from the shortest wavelength side. Three fluorescent dyes, A to C, are used. Figure 4 This is a diagram showing an example of the fluorescence spectra of fluorescent dyes A through C. Suppose there are four types of samples: a sample labeled with fluorescent dye A, a sample labeled with fluorescent dye B, a sample labeled with fluorescent dye C, and a sample not labeled with any fluorescent dye. Figure 5 This diagram schematically represents four types of samples. Hereinafter, the group of samples labeled with one fluorescent dye A will be referred to as Group A, the group of samples labeled with one fluorescent dye B as Group B, the group of samples labeled with one fluorescent dye C as Group C, and the group of samples not labeled with any fluorescent dye as Group D. Furthermore, in practice, if identification is not performed in identification step S5, the relationship between each group and the fluorescent dye cannot be determined; however, for the sake of simplicity, the markings in the cell distribution diagram will be described and explained in association with groups A through D.

[0059] Figures 6-9 This diagram illustrates the processing of the first loop in the repeated processing of the first judgment step S1 and the first separation step S2.

[0060] Figure 6 This is a graph representing the cell distribution data of the measurement data before computational processing. Furthermore, in the actual cell distribution graph, as shown in this figure, the overlap of the four clusters is not explicitly shown; only a density map is displayed. However, similarly to the above, for the sake of simplicity, the four clusters A through D are explicitly shown below. This graph represents the cell distribution for all combinations of two detection channels selected from 12 detection channels. As shown in this figure, from the display of multiple cell distribution graphs, any two or more of the four clusters (here, clusters A through D) overlap, and the four clusters cannot be separated. However, in a certain cell distribution graph (for example, in...) Figure 6In the cell distribution map (C1) obtained by plotting the measurement data of ch2 and ch7 on a two-dimensional graph, one group X (group A in this case) does not overlap with other groups (groups B to D in this case) and can be separated from each other. Figure 9 Therefore, in the first judgment step S1, it is determined that the cells can be separated into group X and other groups on the display of the cell distribution maps of ch2 and ch7. Then, in the first separation step S2, group X is separated from the four groups.

[0061] Figure 7 This is a graph showing the cell distribution data of the separated cluster X. The graph also shows the cell distribution for all combinations of two detection channels selected from the 12 detection channels. From the display of these cell distribution graphs, since no multiple clusters were found, it can be confirmed that cluster X is only one cluster (cluster A in this case).

[0062] Figure 8 (a) is a graph showing the fluorescence spectra of the four groups A to D. Figure 8 (b) is a graph showing the fluorescence spectrum of only the isolated group A. Figure 8 (c) is a graph showing the fluorescence spectra of other groups B to D. Figure 9 (a) is a diagram showing the distribution of cells in ch2 and ch7 when multiple groups containing group X are divided into four groups A to D. Figure 9 (b) is a graph showing the distribution of only group X (group A) in cell distribution map C1 for ch2 and ch7. Figure 9 (c) is a diagram showing the distribution of cells in ch2 and ch7 when the other groups in the cell distribution diagram C1 are divided into groups B to D and represented separately.

[0063] Figures 10-13 This diagram illustrates the processing of the second loop in the repeated processing of the first judgment step S1 and the first separation step S2.

[0064] Figure 10 This is a graph representing the cell distribution data of the measurement data before computational processing (excluding the measurement data of the separated group X (group A)). This graph also shows the cell distribution for all combinations of two detection channels selected from 12 detection channels. As shown in this graph, from the display of multiple cell distribution graphs, any two or more of the three groups (in this case, groups B to D) overlap, making it impossible to separate the three groups separately. However, in a certain cell distribution graph (e.g., in...) Figure 10In the cell distribution map (C2) obtained by plotting the measurement data of ch6 and ch7 on a two-dimensional graph, one group Y (group D in this case) does not overlap with other groups (groups B and C in this case) and can be separated from each other. Figure 13 Therefore, in the first judgment step S1, based on the cell distribution maps of ch6 and ch7, it is determined that they can be separated into group Y and other groups. Then, in the first separation step S2, group Y is separated from the three groups.

[0065] Figure 11 This is a graph showing the cell distribution of the measurement data for the separated group Y. The graph also shows the cell distribution for all combinations of two detection channels selected from the 12 detection channels. From these cell distribution graphs, since no multiple groups were found, it can be confirmed that group Y is only one group (group D in this case).

[0066] Figure 12 (a) is a graph showing the fluorescence spectra of the three groups B to D. Figure 12 (b) is a graph showing the fluorescence spectrum of only the isolated group D. However, since the sample of group D was not fluorescently labeled, the fluorescence from the sample of group D is not produced by the fluorescent dye being targeted, but rather by the autofluorescence of the sample. In this embodiment, for the sake of simplicity, autofluorescence is not considered. Figure 12 (c) is a graph showing the fluorescence spectra of other groups B and C. Figure 13 (a) is a diagram showing the distribution of cells in ch6 and ch7 when the multiple groups containing group Y in the cell distribution diagram C2 are divided into three groups B to D, each representing a different group. Figure 13 (b) is a graph showing the distribution of only group Y (group D) in cell distribution map C2 for ch6 and ch7. Figure 13 (c) is a diagram showing the distribution of cells in ch6 and ch7 when the other groups in the cell distribution diagram C2 are divided into groups B and C respectively.

[0067] Figures 14-17 This diagram illustrates the processing of the third loop in the repeated processing of the first judgment step S1 and the first separation step S2.

[0068] Figure 14 This is a graph representing the cell distribution data before computational processing (excluding the measurement data of separated groups X and Y (groups A and D)). The graph also represents the cell distribution for all combinations of two detection channels selected from 12 detection channels. As shown in the graph, in a certain cell distribution (e.g., in...), Figure 14In the cell distribution map (C3) obtained by plotting the measurement data of ch5 and ch6 on a two-dimensional graph, the two groups (group B and group C in this case) do not overlap and can be separated from each other. Figure 17 Therefore, in the first judgment step S1, based on the display of the cell distribution maps of ch5 and ch6, it is determined that the two groups can be separated into different groups. Then, in the first separation step S2, one group (here, group C) is separated from the two groups.

[0069] Figure 15 This is a graph showing the cell distribution data of the isolated group C. The graph also shows the cell distribution for all combinations of two detection channels selected from the 12 detection channels. From these cell distribution graphs, since no multiple groups were found, it can be confirmed that there is only one group (group C in this case).

[0070] Figure 16 (a) is a graph showing the fluorescence spectra of the two groups B and C respectively. Figure 16 (b) is a graph showing the fluorescence spectrum of only the isolated group C. Figure 16 (c) is a graph showing the fluorescence spectrum of other group B. Figure 17 (a) is a diagram showing the distribution of cells in ch5 and ch6 when the two groups in C3 are divided into groups B and C. Figure 17 (b) is a diagram showing the distribution of only one group (group C) in cell distribution map C3 for ch5 and ch6. Figure 17 (c) is a diagram showing the distribution of other groups (group B) in cell distribution map C3 for ch5 and ch6.

[0071] Figure 18 This diagram shows the cell distribution data from the fourth cycle (excluding the measurement data of the separated clusters (clusters A, C, and D in this case). The diagram also shows the cell distribution for all combinations of two detection channels selected from the 12 detection channels. As shown in the diagram, from the display of these cell distribution diagrams, since no multiple clusters were found, it can be confirmed that there is only one cluster (cluster B in this case). In this fourth cycle, in the first judgment step S1, it is determined that the cluster cannot be further separated.

[0072] Thus, through repeated processing of the first judgment step S1 and the first separation step S2, the four groups (groups A to D) can be separated respectively. Samples contained in each group that is determined to be unable to be further separated are identified as samples labeled with any fluorescent dye. Then, in the identification step S5, the samples identified as labeled with any fluorescent dye are identified based on the fluorescence spectrum obtained from measurement data from multiple detection channels. That is, in effect, as a result of the identification step S5, it is possible to determine which of groups A to D the separated group belongs to.

[0073] Next, use Figures 19-21 Further explanation Figure 3 The fluorescence analysis method is shown. Here, the case where the same treatment as the first cycle described above is followed by a treatment different from the second cycle described above will be explained.

[0074] Figure 19 This is a graph representing the cell distribution of measurement data before computational processing (excluding measurement data from separated group X (group A)). This graph is related to... Figure 10 The cell distribution maps shown are the same, but here we focus on cell distribution maps C4 for ch2 and ch7. In the display of cell distribution maps C4 for ch2 and ch7, groups B and C do not overlap with group D and can be separated from each other. Figure 21 However, in the display of cell distribution map C4, group B and group C overlap, so they appear as a single group Z. Therefore, in the first judgment step S1, in the display of cell distribution maps C4 for ch2 and ch7, it is determined that they can be separated into group Z (groups B and C) and group D. Then, in the first separation step S2, group Z, which contains groups B and C and appears as a single group, is separated from the two groups.

[0075] Figure 20 This is a cell distribution map representing the measurement data of the separated group Z (here, a group containing groups B and C that appears as one unit). Observing the cell distribution maps C51 for ch2 and ch6, C52 for ch3 and ch6, C53 for ch4 and ch6, and C54 for ch5 and ch6, it can be seen that, in their display, group Z, separated in the first separation step S2, can actually be separated into two groups. Therefore, in the second judgment step S3, it is determined that group Z separated in the first separation step S2 can be further separated into multiple groups. Then, in the second separation step S4, any one of these multiple groups is separated. This second judgment step S3 and the second separation step S4 are repeated.

[0076] Figure 21(a) is a schematic diagram of the distribution of cells in ch2 and ch7 when multiple groups in C4 are divided into three groups B to D. Figure 21 (b) is a diagram showing the distribution of cells in ch2 and ch7, including the distribution of group Z, which appears to be one cell in groups B and C in C4.

[0077] Next, use Figures 22-24 Further explanation of the identification step S5: In identification step S5, samples contained in groups determined to be unable to be further separated in the first determination step S1 or the second determination step S3 are identified based on fluorescence spectra obtained from measurement data from multiple detection channels used for fluorescence detection. Each sample within a group is similarly labeled based on fluorescent dyes. Samples may be unlabeled, labeled with only one of several fluorescent dyes, or labeled with two or more fluorescent dyes. Furthermore, when a sample is labeled with two or more fluorescent dyes, the number of dyes may vary depending on the type of fluorescent dye. The fluorescence spectra measured for each sample within a group correspond to the type and number of fluorescent dyes used to label each sample.

[0078] In identification step S5, the type and number of fluorescent pigments labeled in each sample are identified. As identification methods, two cases are considered: one is directly determined by comparing the fluorescence spectra of various fluorescent pigments, and the other is determined through computational processing. In the former case, since no computational processing is performed, high-speed processing is possible. On the other hand, as... Figures 22-24 As will be described later, there are many situations where the relationship between a sample and a fluorescent dye can take various forms, making it difficult to determine without computational processing. Various computational processing methods exist, such as fluorescence compensation and fluorescence intensity calculation. More specifically, there are methods that pre-obtain the fluorescence spectra of various fluorescent dyes and use them as a reference for fitting, and methods that do not require a reference (see Patent Document 3). Below, as an example, we will describe an identification method based on fitting using a reference. Here, we also use three fluorescent dyes A to C, and their fluorescence spectra (reference) are as follows: Figure 4 As shown. Additionally, in the first separation step S2 or the second separation step S4, the cells are separated into 6 groups.

[0079] Figure 22 Figures (a) through (f) show the fluorescence spectra of the six isolated groups. Among them, Figure 22 The fluorescence spectra of (a) to (e) and Figure 4 The different fluorescence spectra shown indicate fluorescence produced in samples labeled with two or more fluorescent dyes. Figure 22The fluorescence spectrum of (f) indicates autofluorescence of a sample not labeled with any fluorescent dye, since no fluorescence was confirmed from the fluorescent dye.

[0080] Figure 23 (a) means Figure 22 The fluorescence spectrum of (a) and Figure 4 The fluorescence spectrum of fluorescent pigment A in the sample (reference) is shown in the figure. Figure 23 (b) means Figure 22 The fluorescence spectrum of (a) and Figure 4 The fluorescence spectrum of fluorescent pigment B in the sample (reference) is shown in the figure. Figure 23 (c) represents Figure 22 The fluorescence spectrum of (a) and Figure 4 The fluorescence spectrum of fluorescent pigment C in the sample (reference) is shown in the figure. Figure 22 The fluorescence spectrum of (a) is a composite spectrum expressed as a linear sum of the fluorescence spectra (reference) of each of the fluorescent dyes A to C. Furthermore, although it is actually a composite spectrum expressed as a linear sum of the fluorescence spectra (reference) of samples with added autofluorescence, in this embodiment, for the sake of simplicity, as mentioned above, the autofluorescence of the samples is assumed to have no effect (not considered). When performing fitting processing using this setting, it can be seen that the samples included in the group are labeled with 1 fluorescent dye A, 2 fluorescent dyes B, and 1 fluorescent dye C. Figure 23 (d)

[0081] By performing the same fitting process, it can be seen that the display... Figure 22 The samples included in the group of fluorescence spectra in (b) were labeled with one fluorescent dye A, one fluorescent dye B, and one fluorescent dye C. Figure 24 (a)). It can be seen that the display shows Figure 22 The group containing the fluorescence spectrum of (c) consists of samples labeled with one fluorescent dye B and one fluorescent dye C. Figure 24 (b)). It can be seen that the display Figure 22 The group containing the (d) fluorescence spectrum is labeled with two fluorescent dyes B and one fluorescent dye C. Figure 24 (c)). It can be seen that the display Figure 22 The group containing the fluorescence spectrum of (e) consists of samples labeled with one fluorescent dye A and one fluorescent dye B. Figure 24 (d)

[0082] As described above, in this embodiment, the groups are separated as much as possible using the measurement data before computational processing, and then identified based on the fluorescence spectra of the samples from each separated group. Therefore, deviations in the measurement data caused by computational processing can be suppressed, improving analytical accuracy. Furthermore, limitations on fluorescence intensity and fluorescence wavelength can be mitigated, increasing the number of usable fluorescent dyes.

Claims

1. A fluorescence analysis device, wherein, The fluorescence analysis device analyzes the measurement data before computational processing and identifies samples based on the type of fluorescent dye labeled. The measurement data before computational processing is obtained by detecting the fluorescence generated when multiple samples labeled with any one of a variety of fluorescent dyes are irradiated with excitation light through multiple detection channels. The fluorescence analysis device includes: The first judgment unit determines, based on measurement data obtained from any one or more of the plurality of detection channels, whether the plurality of samples can be separated into two or more groups. and The first separation unit separates any one of the groups if the first determination unit determines that the group can be separated into two or more groups. For multiple samples other than the already identified samples, the first judgment unit and the first separation unit process them repeatedly until the first judgment unit determines that they cannot be separated into more than two groups. The samples contained in the groups that are determined to be inseparable are then identified as samples labeled with any fluorescent dye.

2. The fluorescence analysis device according to claim 1, wherein, It also has: The second judgment unit determines, based on measurement data obtained from any one or more of the multiple detection channels, whether the samples contained in the separated group can be further separated into two or more groups. and The second separation unit separates any one of the groups if the second determination unit determines that the group can be further separated into two or more groups. For multiple samples other than the already identified samples, the second judgment unit and the second separation unit process them repeatedly until the second judgment unit determines that they cannot be separated into more than two groups. The samples contained in the groups that are determined to be inseparable are then identified as samples labeled with any fluorescent dye.

3. The fluorescence analysis device according to claim 1, wherein, It also has: The identification unit identifies samples labeled with any fluorescent dye based on fluorescence spectra obtained from measurement data obtained from the plurality of detection channels.

4. The fluorescence analysis device according to claim 1, wherein, The first judgment unit makes a judgment based on measurement data, which is obtained by detecting the fluorescence generated when multiple samples labeled with any one of the multiple fluorescent dyes are irradiated with excitation light of multiple wavelengths through multiple detection channels.

5. The fluorescence analysis device according to claim 1, wherein, The first judgment unit makes a judgment after processing the measurement data into bins.

6. A fluorescence analysis method, wherein, This fluorescence analysis method analyzes measurement data before computational processing to identify samples based on the type of fluorescent dye used for labeling. The measurement data before processing is obtained by detecting the fluorescence produced when multiple samples labeled with any one of various fluorescent dyes are irradiated with excitation light through multiple detection channels. The fluorescence analysis method includes: The first judgment step involves determining, based on measurement data obtained from any one or more of the plurality of detection channels, whether the plurality of samples can be separated into two or more groups; and In the first separation step, if it is determined in the first judgment step that the group can be separated into two or more groups, then any one of those groups is separated. For multiple samples other than the already identified samples, the first judgment step and the first separation step are repeated until it is determined in the first judgment step that they cannot be separated into more than two groups. The samples contained in the groups that are determined to be inseparable are identified as samples labeled with any fluorescent dye.

7. The fluorescence analysis method according to claim 6, wherein, Also includes: The second judgment step involves determining, based on measurement data obtained from any one or more of the multiple detection channels, whether a sample contained within a separated cluster can be further separated into two or more clusters; and In the second separation step, if it is determined in the second judgment step that the group can be further separated into two or more groups, then one of the groups is separated. For multiple samples other than the already identified samples, the second judgment step and the second separation step are repeated until it is determined in the second judgment step that they cannot be separated into more than two groups. The samples contained in the groups that are determined to be inseparable are identified as samples labeled with any fluorescent dye.

8. The fluorescence analysis method according to claim 6, wherein, Also includes: The identification step involves identifying samples that are identified as being labeled with any fluorescent dye based on fluorescence spectra obtained from measurement data obtained from the plurality of detection channels.

9. The fluorescence analysis method according to claim 6, wherein, In the first judgment step, the judgment is based on measurement data, which is obtained by detecting the fluorescence generated when multiple samples labeled with any one of the multiple fluorescent dyes are irradiated with excitation light of multiple wavelengths through multiple detection channels.

10. The fluorescence analysis method according to claim 6, wherein, In the first judgment step, the judgment is made after the measurement data is binned.

Images