Analytical apparatus and analytical method

Abstract

The present invention provides an analytical method and apparatus for detecting fluorescence emitted from multiple types of phosphors at multiple emission points, while distinguishing between them. [Solution] In an analytical method and apparatus for detecting fluorescence from multiple emission points in multiple wavelength bands in order to distinguish the emission of multiple types of phosphors from multiple emission points, spatial crosstalk and spectral crosstalk exist between the signal intensities of the multiple emission points and multiple wavelength bands, which degrades the performance of the above-mentioned identification. Therefore, by inputting all of the detection signals from each of the multiple emission points in multiple wavelength bands into a predetermined calculation formula, spatial crosstalk and spectral crosstalk are eliminated, and the concentration of each phosphor at each emission point is derived.

Description

[Technical Field]

[0001] This disclosure relates to an analytical method and an analytical apparatus for detecting fluorescence emitted from multiple types of phosphors at multiple emission points, while distinguishing between them. [Background technology]

[0002] In recent years, multi-capillary electrophoresis systems have become widely used, in which multiple capillaries are filled with electrophoretic separation media such as electrolyte solutions or electrolyte solutions containing polymer gels or polymers, and electrophoretic analysis is performed in parallel. The analytes range from small molecules to large molecules such as proteins and nucleic acids. There are also many measurement modes, including a mode in which lamp light is shone on the absorption point of each capillary and the absorption of lamp light generated when the analyte passes through the absorption point is detected, or a mode in which laser light is shone on the emission point of each capillary and the fluorescence or scattered light generated when the analyte passes through the emission point is detected.

[0003] For example, in Patent Document 1, all capillaries around A light-emitting points on A capillaries (where A is an integer greater than or equal to 2) are arranged on the same plane, a laser beam is introduced from the side of the arrangement plane to irradiate all the light-emitting points of the capillaries simultaneously, and the fluorescence generated at each light-emitting point is wavelength-dispersed and detected simultaneously from a direction perpendicular to the arrangement plane. In the detection device, the fluorescence emitted from the A light-emitting points is collimated together with one focusing lens, passed through one transmission-type diffraction grating, and the primary diffracted light from each fluorescence is imaged together on one two-dimensional sensor with one imaging lens. Here, by making the arrangement direction of the A light-emitting points and the wavelength dispersion direction by the diffraction grating perpendicular to each other, the wavelength-dispersed images of the emitted fluorescence from each capillary do not overlap on the two-dimensional sensor. By setting B detection regions (where B is an integer greater than or equal to 1) of arbitrary wavelength bands for the wavelength-dispersed image of each capillary, B-color detection becomes possible. When B=1, it is called monochromatic detection, and when B≧2, it is called multi-color detection. The multi-capillary electrophoresis apparatus described in Patent Document 1 allows for, for example, Sanger sequencing of different DNA samples in each capillary. In the Sanger method, four types of phosphors are labeled to DNA fragments contained in the DNA sample, according to the terminal base types A, C, G, and T, and the emission fluorescence of each is identified by multicolor detection.

[0004] In Patent Document 2, all capillaries surrounding A light-emitting points on A capillaries (where A is an integer greater than or equal to 2) are arranged on the same plane, a laser beam is introduced from the side of the arrangement plane to irradiate all the light-emitting points of the capillaries at once, and the fluorescence generated at each light-emitting point is divided according to its wavelength component from a direction perpendicular to the arrangement plane and detected all at once. In the detection device, the fluorescence emitted from A light-emitting points is individually collimated with A focusing lenses to form A beams of light, each beam of light is incident in parallel on a set of dichroic mirror arrays consisting of B dichroic mirrors (where B is an integer greater than or equal to 1), each beam is divided into B beams of light in different wavelength bands, and the resulting total of A × B beams of light are incident in parallel on a single 2D sensor to generate A × B divided images on the image. Here, by making the arrangement direction of the A light-emitting points and the division direction of the B light beams by the dichroic mirror array perpendicular to each other, the A × B divided images do not overlap on the image, and it becomes possible to set A × B detection regions. This makes it possible to detect B color in each capillary. Therefore, in the multi-capillary electrophoresis apparatus of Patent Document 2, for example, DNA sequencing by the Sanger method can be performed on different DNA samples in each capillary, similar to the case of Patent Document 1.

[0005] However, generally speaking, simply performing multicolor detection does not allow for the identification of the fluorescence emission of multiple types of phosphors. This is because the fluorescence spectra of each phosphor overlap, resulting in the fluorescence of multiple types of phosphors being mixed within any given wavelength band (referred to as spectral crosstalk in this disclosure). Furthermore, multiple types of phosphors at different concentrations may emit fluorescence simultaneously. Therefore, the following step (referred to as color conversion in this disclosure) is used to eliminate spectral crosstalk and enable the above identification.

[0006] For each of the A light-emitting points (where A is an integer greater than or equal to 2), the emission fluorescence of C types of phosphors (where C is an integer greater than or equal to 1) is detected at each time point using a detection region of B types of wavelengths (where B is an integer greater than or equal to 1). However, B ≥ C. For each light-emitting point and at each time point, a color transformation is applied to the B-color detection result to obtain the concentration of the C types of phosphors. For each light-emitting point P(a) (a=1,2,…,A), the emission fluorescence of phosphor D(c) (c=1,2,…,C) is detected in a detection region W(b) (b=1,2,…,B) of a different wavelength band. At any given time point, the concentration of phosphor D(c) at light-emitting point P(a) is Z(c), and the signal intensity of the detection region W(b) for light-emitting point P(a) is X(b). Here, let X be a B x 1 matrix with signal intensity X(b) as its element, Z be a C x 1 matrix with density Z(c) as its element, and Y be a B x C matrix with Y(b)(c) as its elements. The following equations hold. Equations (1) to (4) are relationships between b and c, but not relationships between a, and they hold independently for each light emission point P(a). In the case of monochromatic detection with B=1, since B≧C, C=1, and X, Y, and Z are no longer matrices.

[0007]

number

number

number

number

[0008] Here, the elements Y(b)(c) of matrix Y, row B and column C, represent the signal intensity ratios in which the emission fluorescence of phosphor D(c) is detected in different wavelength bands in detection region W(b) due to spectral crosstalk. By causing any one type of phosphor D(c0) to emit fluorescence alone, one column Y(b)(c0) (b=1,2,…,B) of matrix Y can be determined. Here, since it is generally difficult to control the concentration of phosphor D(c0), it is convenient to normalize one column Y(b)(c0). For example, it is good to set the largest element among the B elements to 1 and express the other elements as ratios to the maximum value. Alternatively, it is good to determine the ratio of each element so that the sum of the B elements is 1. That is,

number

[0009]

number

[0010] (Equation 1) is a system of simultaneous equations showing the relationship between the concentrations of C types of unknown phosphors and the known B - color fluorescence intensity, and (Equation 6) corresponds to obtaining its solution. Therefore, generally, as described above, the condition B≥C is necessary. If B<C, the solution cannot be uniquely obtained (that is, since multiple solutions may exist), color conversion cannot be performed as in (Equation 6).

[0011] As an example, the case of C = 4 in the Sanger method and B = 4 in four - color detection will be described in detail. When producing copies of DNA fragments of various lengths with respect to template DNA by the Sanger reaction, they are labeled with four types of phosphors D(1), D(2), D(3), and D(4) according to the terminal base types A, C, G, and T, and while separating them by length through electrophoresis in order, they are irradiated with a laser beam to emit fluorescence. Fluorescence emitted is detected in four - color in four detection regions W(1), W(2), W(3), and W(4) corresponding to the respective maximum emission wavelengths of the phosphors D(1), D(2), D(3), and D(4), and time - series data of their signal intensities X(1), X(2), X(3), and X(4) (in the present disclosure, referred to as raw data. However, it is not limited to the case of B = 4, C = 4) are acquired. Let the concentrations of the phosphors D(1), D(2), D(3), and D(4) at each time be Z(1), Z(2), Z(3), and Z(4), then (Equation 1) becomes the following formula.

[0012]

Equation

[0013] Here, the elements Y(b)(c) of the 4x4 matrix Y represent the intensity ratios detected by spectral crosstalk in the wavelength band W(b) (b is 1, 2, 3, or 4) for the emission fluorescence of phosphor D(c) (c is 1, 2, 3, or 4). The elements Y(b)(c) of matrix Y can be determined by electrophoretic analysis of samples in which phosphor D(c) (c is 1, 2, 3, or 4) fluoresces individually in each capillary. For example, the four-color fluorescence intensities X(1), X(2), X(3), and X(4) when phosphor D(1) fluoresces individually give elements Y(1)(1), Y(2)(1), Y(3)(1), and Y(4)(1), respectively. Furthermore, when phosphor D(2) is fluorescing on its own, the four-color fluorescence intensities X(1), X(2), X(3), and X(4) give elements Y(1)(2), Y(2)(2), Y(3)(2), and Y(4)(2), respectively. The same applies to phosphors D(3) and D(4). Y(b)(c) are fixed values ​​determined solely by the properties of phosphor D(c) and wavelength band W(b), and do not change during electrophoresis. Therefore, for each capillary, the concentrations of phosphors D(1), D(2), D(3), and D(4) at each time step can be determined from the four-color fluorescence intensities X(1), X(2), X(3), and X(4) at each time step by the following equation, which is an embodiment of (Equation 6).

[0014]

number

[0015] Thus, the inverse matrix Y -1 By multiplying this by the four-color fluorescence intensity, spectral crosstalk is eliminated, and time-series data of the concentrations of the four types of phosphors, i.e., the concentrations of DNA fragments whose ends are composed of four types of bases (referred to as color-converted data in this disclosure; however, this is not limited to the case where B=4 and C=4), is obtained.

[0016] As already mentioned, the color conversion process described above is performed independently for each of the A capillaries. This is based on the premise that the crosstalk between capillaries (referred to as spatial crosstalk in this disclosure) is sufficiently small. In other words, it means that the ratio of signal intensity derived from fluorescence emitted from other capillaries to the signal intensity obtained by detecting fluorescence emitted from any one capillary is sufficiently small. Therefore, the process of determining matrix Y by sequentially performing the process described above, which involves making one type of phosphor D(c0) fluoresce on its own, for each of the C types of phosphor D(c) is performed substantially simultaneously for the A capillaries, and the matrix Y for each of the A capillaries is obtained in parallel. This is a necessary step in order to derive matrix Y quickly and without much effort.

[0017] While spatial crosstalk is primarily reduced through optical system design, attempts have also been made to reduce it through computational processing. Patent Document 3 describes a method where fluorescence emission from multiple randomly arranged light-emitting points on a plane, rather than multiple capillaries, is imaged onto a single two-dimensional sensor using a single focusing lens. The fluorescence emission from each light-emitting point is obtained as the signal intensity of a detection region located at each imaging position on the two-dimensional sensor. The ratio of spatial crosstalk that the signal intensity of fluorescence emission from any one light-emitting point imposes on the signal intensity of fluorescence emission from other light-emitting points can be expressed as a function of the distance between the two light-emitting points, or the distance between the corresponding two detection regions, and it has been found that this ratio decreases with increasing distance. By pre-determining this function, the spatial crosstalk between any two detection regions is calculated from the distance and this function in the fluorescence image of multiple randomly arranged light-emitting points on a plane, and the calculated spatial crosstalk is subtracted from the original fluorescence image to reduce spatial crosstalk. [Prior art documents] [Patent Documents]

[0018] [Patent Document 1] Patent No. 3897277 [Patent Document 2] Patent No. 6456983 [Patent Document 3] Special Publication No. 2018-529947 [Overview of the project] [Problems that the invention aims to solve]

[0019] In both Patent Document 1 and Patent Document 2, the fluorescence emitted from A emission points on A capillaries is imaged separately on a 2D sensor, so spatial crosstalk is inherently kept low. Possible causes of spatial crosstalk include (1) lens aberrations, (2) multiple reflections between fluorescence elements occurring within the optical system composed of elements such as multi-capillaries, lenses, filters, dichroic mirrors, and 2D sensors, and (3) blooming between pixels of the 2D sensor. To reduce spatial crosstalk, it is necessary to construct the optical system so that the effects of the above factors (1) to (3) are minimized. For example, to suppress the effect of factor (2), a dereflective coating with lower reflectivity can be selected for the lens. However, it is not possible to completely eliminate spatial crosstalk. Spatial crosstalk effectively raises the detection limit in the detection of fluorescence emitted from any single capillary, thus potentially reducing detection sensitivity and narrowing the detection dynamic range. Therefore, reducing spatial crosstalk, even slightly, is extremely important in mitigating or resolving the above-mentioned problems.

[0020] The present inventors attempted to reduce spatial crosstalk in the optical systems of Patent Document 1 or Patent Document 2 using the method of Patent Document 3, but it did not work well. First, the case of monochromatic detection with B=1 will be explained. The spatial crosstalk that the signal intensity X(α)(β) of the detection region β for any one light-emitting point α of A light-emitting points, X(α')(β), imparts to the signal intensity X(α')(β) of the detection region β for any other light-emitting point α' (α≠α') tended to decrease with the distance between the two detection regions, but it became clear that it could not be expressed as a single function of the distance. For example, even when the distance between the two detection regions was constant, the ratio of spatial crosstalk differed depending on whether the two detection regions were located near the central axis of the optical system or far from the central axis of the optical system. Also, the ratio of spatial crosstalk from one of the two detection regions to the other differed from the ratio of spatial crosstalk from the other to the first. In the case of multicolor detection with B≧2, the method of Patent Document 3 became even less effective. The spatial crosstalk between the signal intensity X(α)(β) of detection region β, which is any one detection region of B detection regions, for any one light-emitting point α (α≠α') other than the above, and the signal intensity X(α')(β') of detection region β' (β=β' or β≠β', where β=β' means detection regions in the same wavelength band and β≠β' means detection regions in different wavelength bands) of any one detection-emitting point α' (α≠α') of B detection regions, could not be expressed as a function of the distance between the two detection regions. In particular, when β≠β', spatial crosstalk and spectral crosstalk are mixed, so the crosstalk between the two detection regions could not be expressed as a function of the distance between them. This problem will be explained in detail in [Modes for Carrying Out the Invention].

[0021] Therefore, this disclosure proposes a method for identifying and independently detecting light emission from multiple light emission points by reducing the spatial crosstalk between multiple light emission points that occurs in any optical system that detects light emission from multiple light emission points, through computational processing. Furthermore, this disclosure proposes a method for identifying and independently detecting fluorescence emission from multiple types of phosphors from multiple light emission points by reducing the spatial crosstalk between multiple light emission points and the spectral crosstalk between multiple types of phosphors for each light emission point, both of which occur in any optical system that detects the fluorescence emission from multiple types of phosphors emitted from multiple light emission points, through computational processing. Alternatively, this disclosure proposes a method for identifying and independently detecting absorption at multiple absorption points by reducing the spatial crosstalk between multiple absorption points that occurs in any optical system that detects absorption at multiple absorption points, through computational processing. Furthermore, we propose a method for identifying and independently detecting the absorption of multiple types of absorbers at multiple absorption points by computationally reducing the spatial crosstalk between multiple absorption points and the spectral crosstalk between multiple types of absorbers at each absorption point, which occur in any optical system used to detect the absorption of multiple types of absorbers at multiple absorption points. [Means for solving the problem]

[0022] This section describes an optical system described in Patent Document 1 or Patent Document 2, where, for each of A light-emitting points (where A is an integer of 2 or more), the emission fluorescence of C types of phosphors (C being an integer of 1 or more) is detected in B detection regions (B being an integer of 1 or more) with different wavelength bands. Here, each detection region with a different wavelength band detects a different wavelength component of fluorescence. Furthermore, each type of phosphor emits fluorescence having a different fluorescence spectrum. For each light-emitting point P(a) (a=1,2,…,A), the emission fluorescence of phosphor D(a,c) (c=1,2,…,C) present at light-emitting point P(a) is detected in different wavelength band detection regions W(a,b) (b=1,2,…,B). Let Z(a,c) be the concentration of phosphor D(a,c) at light-emitting point P(a) at an arbitrary time, and X(a',b) be the signal intensity of the detection region W(a',b) for light-emitting point P(a'). Here, the inventors have discovered for the first time that the following equation holds, where X is an A×B x 1 matrix with elements X(a',b), Z is an A×C x 1 matrix with elements Z(a,c), and Y is an (A×B) x (A×C) matrix with elements Y(a',b)(a,c).

[0023]

number

number

number

number

[0024] Equation (9) is the same as equation (1), but a comparison of (1)-(4) and (9)-(12) reveals that they are completely different. Equations (9)-(12) are not only relationships between b and c, but also relationships between a, indicating that different light emission points P(a) are interconnected. In other words, (9)-(12) allows for the consideration of spatial and spectral crosstalk between different light emission points P(a), in addition to spectral crosstalk for the same light emission point P(a), and is fundamentally different from (1)-(4).

[0025] Here, the element Y(a',b)(a,c) of the (A×B) x (A×C) matrix Y represents (1) the signal intensity ratio in which the fluorescence emission of phosphor D(a',c) at P(a') is detected in any detection region W(a',b) due to spectral crosstalk between the same emission point (a'=a), and (2) the signal intensity ratio in which the fluorescence emission of phosphor D(a,c) at emission point P(a) is detected in any detection region W(a',b) at P(a') due to spatial crosstalk and spectral crosstalk between different emission points (a'≠a). By causing one type of phosphor D(a0,c0) to fluoresce alone at any one emission point P(a0), one column Y(a,b)(a0,c0) (a=1,2,…,A, and b=1,2,…,B) of matrix Y can be determined. Here, since it is generally difficult to control the concentration of the phosphor D(a0,c0), it is convenient to normalize the first row Y(a,b)(a0,c0). For example, among A × B elements, it is good to set the largest element to 1 and express the other elements as ratios to the maximum value. Alternatively, it is good to determine the ratio of each element so that the sum of A × B elements equals 1. In other words,

number

[0026]

Equation

[0027] (Equation 14) is the same formula as (Equation 6), but comparing (Equation 2) to (Equation 4) and (Equation 6) with (Equation 10) to (Equation 12) and (Equation 14), it can be seen that they are completely different. According to (Equation 14), for each time, by multiplying the Moore-Penrose pseudoinverse Y - of the previously obtained matrix Y and the A×B signal intensities X(a,b), both spectral crosstalk and spatial crosstalk can be eliminated, and the concentrations Z(a,c) of A×C phosphors can be obtained. Y - is a matrix of (A×C) rows and (A×B) columns. In the present disclosure, corresponding to eliminating spectral crosstalk being called color conversion, eliminating spatial crosstalk is called spatial correction. That is, among the (A×C)×(A×B) elements of Y - , both a part for performing color conversion and a part for performing spatial correction are included. And the time-series data of Z(a,c) obtained by executing (Equation 14) for each time is called color conversion + spatial correction data in the present disclosure.

[0028] Each element of matrix X may be a value after subtracting background light beforehand, or a value after applying appropriate noise reduction processing. Similarly, each element of matrix Y may be a value after subtracting background light beforehand, or may be subjected to appropriate changes.

[0029] In the case of monochromatic detection with B=1 and C=1, the above is simplified as follows: For each emission point P(a) (a=1,2,...,A), the emission fluorescence of one type of phosphor is detected in one detection region. Let Z(a) be the concentration of the phosphor at emission point P(a) at any given time, and X(a') be the signal intensity for emission point P(a'). Here, if X is an A x 1 matrix with X(a') as its elements, Z is an A x 1 matrix with Z(a) as its elements, and Y is an A x A matrix with Y(a')(a) as its elements, then (Equations 10) to (Equations 12) are simplified as follows. (Equations 9) and (Equation 14) remain true.

[0030]

number

number

number

[0031] Equations (9) and (14) to (17) are formally the same as the general formulas used in Patent Document 3, for example, [Equations 7] to [Equation 9] in Patent Document 3, but their content is significantly different (hereafter, formulas used in Patent Document 3 will be indicated in [ ]). The element αij of matrix A in [Equation 7] is a function of the distance dij between the detection regions of each emission point φ(meas)i and φ(meas)j, and specifically, as shown in [Equation 10], it is expressed as a sum of exponential functions of dij, and αij decreases as dij increases. In Patent Document 3, different fluorescence images are obtained for different samples, and the positions of the detection regions of multiple emission points change randomly in each fluorescence image. Therefore, in Patent Document 3, the above function is determined in advance, and for each fluorescence image, the distance between the detection regions of any two emission points among the multiple emission points is determined for all combinations of two emission points, and αij is derived by substituting this into the above function. In other words, αij, i.e., matrix A, changes for each sample or for each fluorescence image. In contrast, as described above, or as described later in [Modes for Carrying Out the Invention], this disclosure has found that the elements Y(a')(a) of matrix Y (Equation 16) cannot be expressed as a function of the distance between the detection regions of each emission point P(a) and P(a'). Furthermore, it has been found that it is impossible to calculate the elements Y(a')(a) from the optical system configuration, etc. Therefore, in this disclosure, under the condition that the position of the detection regions of multiple emission points does not change even for different samples, the elements Y(a')(a) are determined by actual measurement under that condition. Specifically, as already stated, one column Y(a')(a0) (a'=1,2,…,A) of matrix Y is determined by causing fluorescence emission at only one emission point P(a0) alone. Then, all columns of matrix Y are determined by performing the above process sequentially for all A emission points P(a). Matrix Y is determined solely by the characteristics of the emission points P(a) and does not change during the analysis. Furthermore, as long as the optical system, the light emission point P(a) and its detection region are fixed, the matrix Y remains constant even when analyzing different samples. It is impossible to perform the above steps with the optical system or apparatus configuration described in Patent Document 3.This is because the positions of the detection areas of multiple light-emitting points cannot be fixed, and even if they could be fixed, it is not possible to make each light-emitting point emit light individually and sequentially. Therefore, the method of this disclosure cannot be conceived from Patent Document 3.

[0032] The above can be reinterpreted as follows: For each of the A emission points (where A is an integer greater than or equal to 2), the emission of C types of emitters (C being an integer greater than or equal to 1) is detected in B detection regions (B being an integer greater than or equal to 1) of arbitrary wavelength bands. For each emission point P(a) (a=1,2,…,A), the emission of emitters D(a,c) (c=1,2,…,C) present at emission point P(a) is detected in different detection regions W(a,b) (b=1,2,…,B) of different wavelength bands. Let Z(a,c) be the concentration of emitters D(a,c) at emission point P(a) at any given time, and X(a',b) be the emission intensity of W(a',b) for P(a'). In this case as well, (Equations 9) to (Equations 17) hold, and similarly, the concentration Z(a,c) of A × C emitters can be determined by eliminating both spectral crosstalk and spatial crosstalk. Here, "luminescence" refers to phenomena such as fluorescence, phosphorescence, and scattered light.

[0033] Furthermore, the above can be reinterpreted as follows: For each of A absorbance points (where A is an integer greater than or equal to 2), the absorbance of C types of absorbers (where C is an integer greater than or equal to 1) is detected in B detection regions (where B is an integer greater than or equal to 1) of arbitrary wavelength bands. For each absorbance point P(a) (a=1,2,…,A), the absorbance of the absorber D(a,c) (c=1,2,…,C) present at absorbance point P(a) is detected in different detection regions W(a,b) (b=1,2,…,B) of the same wavelength band. Let Z(a,c) be the concentration of the absorber D(a,c) at absorbance point P(a) at any given time, and X(a',b) be the absorbance of W(a',b) at absorbance point P(a'). In this case as well, equations (9) to (17) hold, and similarly, we can eliminate both spectral crosstalk and spatial crosstalk to determine the concentration Z(a,c) of A × C absorbers.

[0034] Alternatively, the above can be reinterpreted as multipoint detection other than optical measurement. That is, for each of A signal generation points (where A is an integer greater than or equal to 2), the signals of C types of signal generators (where C is an integer greater than or equal to 1) are detected in B detection regions of arbitrary frequency bands (where B is an integer greater than or equal to 1). For each signal generation point P(a) (a=1,2,...,A), the signals of signal generators D(a,c) (c=1,2,...,C) present at signal generation point P(a) are detected in detection regions W(a,b) (b=1,2,...,B) of different frequency bands. Let Z(a,c) be the concentration of signal generators D(a,c) at signal generation point P(a) at any given time, and X(a',b) be the signal intensity of W(a',b) for signal generation point P(a'). In this case as well, equations (9) to (17) hold, and similarly, both spectral crosstalk and spatial crosstalk can be eliminated to determine the concentration Z(a,c) of A × C signal generators.

[0035] The above explanation has primarily used mathematical formulas, which is intended to facilitate understanding of the contents of this disclosure. When implementing the technology of this disclosure, it is sufficient to use methods based on the contents of this disclosure, and it is not necessary to follow the formulas exactly. The formulas may be modified, or not used at all. Furthermore, in this disclosure, the term "concentration of the light-emitting material" can be read as "luminescence intensity of the light-emitting material," the term "concentration of the phosphor" can be read as "fluorescence intensity of the phosphor," and the term "concentration of the light-absorbing material" can be read as "absorbance of the light-absorbing material."

[0036] Further features relating to this disclosure will become apparent from the description herein and the accompanying drawings. Furthermore, aspects of this disclosure are achieved and realized by elements and various combinations of elements and the modes of the claims described herein in detail thereafter. The descriptions herein are typical examples only and do not limit in any way the claims or applications of this disclosure.

[0037] Further features relating to this disclosure will become apparent from the description herein and the accompanying drawings. Furthermore, aspects of this disclosure are achieved and realized by elements and various combinations of elements and the modes of the claims described herein in detail thereafter. The descriptions herein are typical examples only and do not limit in any way the claims or applications of this disclosure. [Effects of the Invention]

[0038] According to this disclosure, by resolving or reducing the spatial crosstalk between multiple light-emitting points that occurs in any optical system that detects light emission from multiple light-emitting points, it becomes possible to identify and independently detect light emission from multiple light-emitting points. Furthermore, by resolving or reducing the spatial crosstalk between multiple light-emitting points and the spectral crosstalk between multiple types of phosphors for each light-emitting point, which occur in any optical system that detects the fluorescence of multiple types of phosphors emitted from multiple light-emitting points, it becomes possible to identify and independently detect the fluorescence emission of multiple types of phosphors from multiple light-emitting points.

[0039] Furthermore, by eliminating or reducing spatial crosstalk and spectral crosstalk, it becomes possible to avoid the decrease in detection sensitivity or detection dynamic range that occurs as a result of spatial and spectral crosstalk. Other issues, configurations, and effects not mentioned above will be clarified by the following description of the embodiments. [Brief explanation of the drawing]

[0040] [Figure 1] Schematic diagram of a simple optical system [Figure 2] Light emission image acquired with a 2D sensor using a simple optical system [Figure 3] Signal intensity distribution of the emission image in the emission image [Figure 4]Relationship between the absolute signal intensity at the center of the emission image and the absolute signal intensity at a position to the left of the center of the emission image. [Figure 5] Relationship between the absolute signal intensity at the center of the emission image and the absolute signal intensity at a position to the right of the center of the emission image. [Figure 6] Relationship between the absolute signal intensity at the center of the emission image and the absolute signal intensity at a position to the left of the center of the emission image. [Figure 7] Relationship between the absolute signal intensity at the center of the emission image and the absolute signal intensity at a position to the right of the center of the emission image. [Figure 8] Schematic diagram of the model experimental system [Figure 9] Schematic diagram of spectral crosstalk and spatial crosstalk in a model experimental system. [Figure 10] Schematic diagram of the model experimental system [Figure 11] Raw data obtained from electrophoretic analysis of known samples [Figure 12] Color-converted data obtained by applying color conversion to the raw data in Figure 11. [Figure 13] Color-converted and spatially corrected data obtained by applying color conversion and spatial correction to the raw data in Figure 11. [Figure 14] Raw data obtained from electrophoretic analysis of unknown samples [Figure 15] Color-converted data obtained by applying color conversion to the raw data in Figure 14. [Figure 16] Color-converted and spatially corrected data obtained by applying color conversion and spatial correction to the raw data in Figure 14. [Figure 17] Raw data obtained from electrophoretic analysis of samples containing an unknown phosphor. [Figure 18] Color-converted and spatially corrected data obtained by applying color conversion and spatial correction to the raw data in Figure 17. [Figure 19] Raw data obtained from electrophoretic analysis of samples with gradually increasing concentrations of known phosphors. [Figure 20] Color-converted data obtained by applying color conversion to the raw data in Figure 19. [Figure 21] Color-converted and spatially corrected data obtained by applying color conversion and spatial correction to the raw data in Figure 19. [Figure 22] Schematic diagram of a method for injecting samples into multiple capillaries at staggered timings. [Figure 23] Schematic diagram of a method for obtaining fluorescence emission by staggering the timing of multiple capillaries. [Figure 24] Schematic diagram of a multi-capillary electrophoresis apparatus [Figure 25] Wavelength dispersion image of Raman scattered light from four capillaries using the optical system of Patent Document 1. [Figure 26] Figure 25 shows the settings for the detection regions of 20 wavelength bands for each capillary. [Figure 27] A schematic diagram of the optical system described in Patent Document 2, which divides the emission fluorescence from four capillaries into four individual wavelength bands and forms an image. [Figure 28] An optical system and schematic diagram that simultaneously divides the fluorescence emission from five light-emitting points arranged on a plane into four wavelength bands for imaging. [Figure 29] Flowchart of the conventional method [Figure 30] Flowchart of this law [Figure 31] Flowchart for determining the matrix used for color conversion and spatial correction in this method. [Figure 32] Flowchart for repeating analysis sessions multiple times [Figure 33] Computer configuration diagram [Modes for carrying out the invention]

[0041] [Example 1] To investigate the characteristics of spatial crosstalk in detail, a simple optical system shown in Figure 1 was constructed. The optical system in Figure 1 comprises a pinhole plate 1-1, an aperture plate on the light-emitting point side 1-2, a focusing lens 1-3, a sensor-side aperture plate 1-4, a colored glass filter 1-5, a two-dimensional sensor 1-6, and a halogen lamp (light source) that emits halogen lamp light 1-7. Specifically, the optical system in Figure 1 was configured as follows: A light-emitting point 1-8 with a diameter of φ0.05 mm was formed by irradiating a pinhole plate 1-1 with a φ0.05 mm pinhole from below with halogen lamp light 1-7. An aperture plate 1-2 on the light-emitting point side, with a diameter of φ0.2 mm, was placed 0.2 mm above the light-emitting point 1-8, a focusing lens 1-3 with a focal length of f=1.4 mm was placed 1.54 mm above the light-emitting point 1-8, and a sensor-side aperture plate 1-4 with a diameter of φ0.7 mm was placed directly above the focusing lens 1-3. Furthermore, the 2D sensor 1-6 was positioned 15 mm above the condensing lens 1-3, and the colored glass filter 1-5 was placed directly below the 2D sensor 1-6. With these arrangements, the pinhole plate 1-1, the light-emitting point side aperture plate 1-2, the sensor side aperture plate 1-4, the colored glass filter 1-5, and the 2D sensor 1-6 were arranged parallel to each other. Light 1-9 emitted from the light-emitting point 1-8 passed through the φ0.2 mm aperture, was focused by the condensing lens 1-3, passed through the φ0.7 mm aperture, passed through the colored glass filter 1-5, and formed a φ0.5 mm light-emitting image 1-10 on the 2D sensor 1-6. Here, the light-emitting point 1-8 was focused on the 2D sensor 1-6 and imaged at a magnification of 10x.

[0042] Figure 2 shows the emission image including images 1-10 obtained by 2D sensors 1-6 in the simple optical system of Figure 1. The sensor size of 2D sensors 1-6 is 13 × 13 mm, and the signal range of each pixel is 0 to 65536. Figures 2(a) and 2(b) are the same emission image, but the signal display scale (grayscale) in Figure 2(a) is set to 0 to 50000, while the signal display scale in Figure 2(b) is set to 0 to 500. The maximum signal value of the emission image is approximately 50000, and in Figure 2(a), the signals of all pixels are displayed without saturation, whereas in Figure 2(b), the emission image is displayed as saturated. According to Figure 2(a), an emission image of φ0.5 mm is obtained as expected, and the area outside the φ0.5 mm emission image appears to have zero signal. However, according to Figure 2(b), it can be confirmed that a low-intensity tail extends outside the φ0.5 mm emission image.

[0043] In Figures 3(a) to 3(c), the dotted line α represents the horizontal signal intensity distribution passing through the center of the emission image in Figure 2. Figures 3(a), 3(b), and 3(c) use the same data, but the vertical axis is changed. The horizontal axis is common to all figures and shows the distance from the center of the emission image. The positive and negative values ​​on the horizontal axis indicate the right and left sides of the emission image, respectively. In Figures 3(a) and 3(b), the vertical axis shows absolute signal intensity, with a vertical axis scale of 0 to 60000 in Figure 3(a) and 0 to 100 in Figure 3(b). On the other hand, in Figure 3(c), the vertical axis is on a logarithmic scale and represents relative signal intensity with the maximum signal intensity set to 100%, with a vertical axis scale of 0.001% to 100%. Figures 3(a) to 3(c) show the signal intensity distribution for β (solid line) and γ (dashed line) under the same conditions as the acquisition of the emission image in Figure 2, but with the output intensity of halogen lamp light 1-7 gradually reduced. As can be seen from Figure 3(a), the maximum signal intensities of the emission images α, β, and γ are approximately 50,000, 25,000, and 10,000, respectively. On the other hand, Figure 3(b) shows that the signal intensity tends to decrease as you move away from the center of the emission image, but it can be seen that the tail is much larger compared to the size of the emission image seen in Figure 3(a) (width of about 0.5 mm). It can also be seen that as the maximum signal intensity decreases, the intensity of the tail also decreases. Looking at Figure 3(c), it can be seen that α, β, and γ overlap. This is a new discovery and provides several important insights. For example, at a position ±1 mm away from the center of the emission image, that is, outside the emission image seen in Figure 2(a), a signal intensity of approximately 0.1% of the maximum signal intensity is observed. If detection areas for adjacent emission images exist at a distance of ±1 mm, this means that there is approximately 0.1% spatial crosstalk. As shown in Figure 1, it is surprising that such a significant amount of spatial crosstalk exists even when the emission point is in perfect focus using an extremely simplified optical system. Even more surprising is that, since α, β, and γ overlap, it was found that, as long as the optical system and the position of the emission point are fixed, the relative signal intensity distribution of the imaging point is constant, regardless of the emission intensity of the emission point or the signal intensity of the imaging point.

[0044] Figures 4(a) and 5(a) are the results derived from Figure 3. Figure 4(a) is a graph with absolute signal intensity at positions -1, -2, -3, -4, -5, and -6 mm away from the center of the image point on the vertical axis, and Figure 5(a) is a graph with absolute signal intensity at positions +1, +2, +3, +4, +5, and +6 mm away from the center of the image point on the horizontal axis in both cases, and absolute signal intensity at the center of the emitted image on the horizontal axis. The absolute signal intensity above is the average value of the absolute signal intensity within a ±0.1 mm width centered on each position in Figure 3. As a result, it was found that all three plots at each position (corresponding to α, β, and γ in Figure 3) lie on an approximate straight line passing through the origin. The slope of each approximate straight line indicates the spatial crosstalk ratio at the corresponding position. In other words, it can be seen that the slope of the approximate straight line decreases as you move away from the center of the image point, and the spatial crosstalk ratio decreases. For comparison, the straight lines for spatial crosstalk of 0.1% and 0.01% are superimposed as dotted lines. For example, the spatial crosstalk ratio at a position -1 mm away from the center of the image point is slightly over 0.1%, while the spatial crosstalk ratio at a position -3 mm away is approximately 0.01%. These results indicate that the absolute signal intensity of spatial crosstalk at any position away from the center of the image point is linear with respect to the absolute signal intensity at the center of the image point, and that the spatial crosstalk ratio is constant. In other words, it has been newly discovered that it is possible to eliminate or reduce spatial crosstalk at any position by subtracting the value obtained by multiplying the absolute signal intensity at the center of the image point by the spatial crosstalk ratio from the absolute signal intensity at any position away from the center of the image point. This corresponds to subtracting the value of the corresponding approximation line at the same position from the absolute signal intensity of each plot in Figures 4(a) and 5(a).

[0045] Figures 4(b) and 5(b) show the results of the subtraction operation described above applied to the results in Figures 4(a) and 5(a), respectively. The type of plots representing each position and the horizontal axis are common to Figures 4(a) and 4(b), and to Figures 5(a) and 5(b). The vertical axis in Figures 4(b) and 5(b) shows the absolute signal intensity after the subtraction operation described above, and the scale is enlarged compared to Figures 4(a) and 5(a). For comparison, a dotted line representing a spatial crosstalk of ±0.01% is superimposed. As a result, it was found that the absolute signal intensity after the subtraction operation at each position becomes almost zero, regardless of the absolute signal intensity at the image center, and the spatial crosstalk at each position becomes smaller than ±0.01%. When comparing positions at a distance of -1 mm from the center of the image point, this method was able to reduce spatial crosstalk by at least one order of magnitude. The reason why the absolute signal intensity after subtraction at each location is not exactly zero is because there is a discrepancy or error between each plot in Figure 4(a) and Figure 5(a) and the corresponding approximation line. The coefficient of determination (R) of the approximation line passing through the origin at each location is... 2 The higher the value of (closer to 1), that is, the higher the linearity of the absolute signal intensity at each position relative to the absolute signal intensity at the center of the image point, the smaller the above error becomes, and the absolute signal intensity after the subtraction calculation approaches zero. Naturally, the absolute signal intensity after the subtraction calculation can also be negative. However, as long as the above linearity exists, the magnitude of the absolute signal intensity at each position will be at least reduced by the subtraction calculation. Also, the smaller the slope of the approximation line, that is, the smaller the spatial crosstalk before the subtraction calculation, the smaller the absolute value of the difference between the absolute signal intensity at each position and the approximation line, and therefore the smaller the above error becomes.

[0046] Figures 6 and 7 show experimental results similar to those in Figures 4 and 5. In the optical system of Figure 1, the pinhole plate 1-1 was removed and then reattached, and the same results as in Figures 2 and 3 were obtained. The results in Figures 6 and 7 were then derived using the same method. The mounting position of the pinhole plate 1-1 is almost the same as the original position, but not exactly the same. The results in Figures 6 and 7 are equivalent to the results in Figures 4 and 5, demonstrating the high reproducibility of the method of eliminating or reducing spatial crosstalk at arbitrary positions by subtraction calculation.

[0047] On the other hand, a comparison of Figures 4 to 7 yielded another important finding, as shown below. We compared the approximation lines at -1 mm in Figure 4(a), +1 mm in Figure 5(a), -1 mm in Figure 6(a), and +1 mm in Figure 7(a). The slopes of these four approximation lines all represent the spatial crosstalk ratio at a position 1 mm away from the center of the image point, but it was found that they differ from one another. Therefore, for example, if we use the approximation line at -1 mm in Figure 4(a) to perform a subtraction operation on the plots at +1 mm in Figure 5(a), -1 mm in Figure 6(a), or +1 mm in Figure 7(a), the above error becomes large, and it became clear that the reduction of spatial crosstalk is insufficient. This phenomenon occurs similarly at other positions, and in some cases, the subtraction operation can have the opposite effect, that is, increase spatial crosstalk. The factors causing this phenomenon include (1) even in a simple optical system like the one in Figure 1, the point symmetry of the spatial crosstalk caused by imaging is not necessarily high (for example, in Figure 2, the spatial crosstalk ratio differs between the left and right directions of the imaging point), and (2) even slight changes in the optical system, such as a slight shift in the position of the light-emitting point, can significantly alter the spatial crosstalk caused by imaging. Therefore, it has been found that it is impossible to express the spatial crosstalk ratio occurring between the center of the imaging point and an arbitrary position as a function of the distance between them. This means that the method shown in Patent Document 3 does not function effectively. This disclosure differs from Patent Document 3 in that the conditions of the optical system and multiple imaging points, i.e., multiple detection regions, are fixed, and the spatial crosstalk ratios between them are derived experimentally, regardless of their distance from each other, as shown in Figures 4 to 7. If the conditions are changed, it is necessary to reacquire the spatial crosstalk ratio each time.

[0048] Extending and generalizing the findings obtained from the above experiments—namely, the fact that the absolute signal intensity of spatial crosstalk at any position far from the center of the image point is linear with respect to the absolute signal intensity at the center of the image point, and that the spatial crosstalk ratio is constant—we can derive equations (9) to (17).

[0049] [Example 2] A model experiment was conducted on the simplest case of an optical system that detects the fluorescence of multiple types of phosphors emitted from multiple emission points, each in a detection region of multiple wavelength bands. In this example, and in other examples using the same model experimental system, as an example, in (Equation 9) to (Equation 12), the case was chosen in which A = 2 emission points, B = 2 wavelength bands, and C = 2 types of phosphor emission fluorescence are detected for each of them. It goes without saying that the same effect can be obtained by the same method even when A, B, and C are other values.

[0050] Figure 8 shows a model experimental system. The model experimental system in Figure 8 comprises two capillaries, Cap(1) and Cap(2), and a light source (not shown) that irradiates the two capillaries, Cap(1) and Cap(2), with a laser beam LB in the direction of their alignment. Emission points P(1) and P(2) are provided on the two capillaries, Cap(1) and Cap(2), at the positions where the laser beam LB is irradiated. Substances labeled with two types of phosphors D(1,1) and D(1,2) are subjected to electrophoresis on capillary Cap(1), and substances labeled with two types of phosphors D(2,1) and D(2,2) are subjected to electrophoresis on capillary Cap(2). Here, D(1,1) and D(2,1) are of the same type of phosphor, and D(1,2) and D(2,2) are of the same type of phosphor. Being of the same type of phosphor means that the fluorescence spectra of the emitted fluorescence are equal. Phosphors D(1,1) and D(1,2) emit fluorescence at emission point P(1) upon irradiation with laser beam LB and are detected by detection regions W(1,1) and W(1,2) provided on sensor S. The detection wavelength bands of detection region W(1,1) and detection region W(1,2) are designed to primarily detect the fluorescence emitted by phosphor D(1,1) and phosphor D(1,2) respectively, but significant spectral crosstalk exists between them. Similarly, phosphors D(2,1) and D(2,2) emit fluorescence at emission point P(2) upon irradiation with laser beam LB and are detected by detection regions W(2,1) and W(2,2) provided on sensor S. The detection wavelength bands are designed so that detection region W(2,1) primarily detects the emission fluorescence of phosphor D(2,1), and detection region W(2,2) primarily detects the emission fluorescence of phosphor D(2,2), but significant spectral crosstalk exists between them. Furthermore, the fluorescence emission of phosphors D(1,1) and D(1,2) at emission point P(1) is also detected in detection regions W(2,1) and W(2,2), and similarly, the fluorescence emission of phosphors D(2,1) and D(2,2) at emission point P(2) is also detected in detection regions W(1,1) and W(1,2), indicating significant spatial crosstalk.If the signal intensities (fluorescence intensities) in the detection regions W(1,1), W(1,2), W(2,1), and W(2,2) are denoted as X(1,1), X(1,2), X(2,1), and X(2,2), respectively, then, as shown in the lower part of Figure 8, time-series data of signal intensities X(1,1) and X(1,2), X(2,1) and X(2,2) are obtained for each emission point P(1) and emission point P(2) in conjunction with electrophoresis.

[0051] In Figure 8, the detection regions W(1,1), W(1,2), W(2,1), and W(2,2) are depicted as being located on a single sensor S, but this is not necessarily required. The detection regions W(1,1) and W(1,2) may be located on one sensor, and the detection regions W(2,1) and W(2,2) may be located on another sensor, or the detection regions W(1,1), W(1,2), W(2,1), and W(2,2) may be located on four different sensors. Furthermore, some kind of imaging and spectral means are required between the light-emitting points P(1) and P(2) and the sensor S, but in this embodiment, the type of imaging and spectral means is not specified, so the imaging and spectral means are omitted in Figure 8.

[0052] Figure 9 is a schematic diagram showing the relationship between spectral crosstalk and spatial crosstalk in Figure 8. In Figure 9(a), the fluorescence emission of phosphor D(1,1) at emission point P(1) is (i) mainly detected in detection region W(1,1) and (ii) secondarily detected in detection region W(1,2). Furthermore, the above emission fluorescence is detected in (iii) detection region W(2,1) at a much lower intensity than in (i) and (ii), and further, in (iv) detection region W(2,2) at an even lower intensity than in (iii). (i) and (ii) show spectral crosstalk, and (iii) and (iv) show spatial crosstalk and spectral crosstalk. In other words, conventional methods only considered (i) and (ii), but this disclosure considers all of (i) to (iv). In Figure 9, the relationship between the strength of the detection intensities and the relationship between crosstalk are commonly shown by the thickness of the arrows, the solid or dotted line, and the numbers (i) to (iv). Similar to Figure 9(a), Figure 9(b) depicts the fluorescence emission of phosphor D(1,2) at emission point P(1), Figure 9(c) depicts the fluorescence emission of phosphor D(2,1) at emission point P(2), and Figure 9(d) depicts the fluorescence emission of phosphor D(2,2) at emission point P(2). In actual analysis, Figures 9(a) to (d) occur simultaneously, and their fluorescence intensities differ.

[0053] Figure 10 shows a model experimental setup similar to that in Figure 8. However, in the model experimental setup in Figure 10, lamp light LL is irradiated from a light source perpendicular to the arrangement direction of the two capillaries Cap(1) and Cap(2). Therefore, instead of detecting emission fluorescence associated with irradiation by the laser beam LB, absorbance or absorbance associated with the transmission of lamp light LL is detected. Absorbance points P(1) and P(2) are provided on the two capillaries Cap(1) and Cap(2) at the positions where the lamp light LL is irradiated. Two types of absorbers D(1,1) and D(1,2) are subjected to electrophoresis in capillary Cap(1), and two types of absorbers D(2,1) and D(2,2) are subjected to electrophoresis in capillary Cap(2). Here, D(1,1) and D(2,1) are the same type of absorber, and D(1,2) and D(2,2) are the same type of absorber. The term "same type of absorber" means that the light absorption spectra are identical. Absorbers D(1,1) and D(1,2) absorb light at the absorption point P(1) upon irradiation with lamp light LL, and the unabsorbed light is detected in detection regions W(1,1) and W(1,2) provided on the sensor S. The detection wavelength bands of detection region W(1,1) and detection region W(1,2) are designed to primarily detect the absorption of absorber D(1,2), respectively, but significant spectral crosstalk exists between them. Similarly, absorbers D(2,1) and D(2,2) absorb light at the absorption point P(2) upon irradiation with lamp light LL, and the unabsorbed light is detected in detection regions W(2,1) and W(2,2) provided on the sensor S. The detection wavelength bands are designed such that detection region W(2,1) primarily detects the absorbance of absorber D(2,1), and detection region W(2,2) primarily detects the absorbance of absorber D(2,2); however, significant spectral crosstalk exists between them. Furthermore, the absorbance of absorbers D(1,1) and D(1,2) at absorbance point P(1) is also detected in detection regions W(2,1) and W(2,2), and similarly, the absorbance of absorbers D(2,1) and D(2,2) at absorbance point P(2) is also detected in detection regions W(1,1) and W(1,2), indicating significant spatial crosstalk.This is because a portion of the transmitted light from the lamp light LL that passes through absorbance point P(1) is detected in detection regions W(2,1) and W(2,2), and a portion of the transmitted light from the lamp light LL that passes through absorbance point P(2) is detected in detection regions W(1,1) and W(1,2). If the absorbances in detection regions W(1,1), W(1,2), W(2,1), and W(2,2) are X(1,1), X(1,2), X(2,1), and X(2,2), then, as shown in the lower part of Figure 8, time-series data of absorbances X(1,1) and X(1,2), X(2,1), and X(2,2) are obtained for each of the absorbance points P(1) and P(2) as a result of electrophoresis. The following explanation will use Figure 8, but it goes without saying that the same effect can be obtained using Figure 10.

[0054] Figure 11 shows raw data obtained by electrophoresis analysis using the model experimental system in Figure 8, after injecting known samples into capillaries Cap(1) and Cap(2). In both figures, the x-axis represents electrophoresis time (arbitrary units), and the y-axis represents fluorescence intensity (arbitrary units). Figures 11(a) and 11(b) show the same raw data obtained at emission point P(1), with (b) being an enlarged version of (a) on the y-axis scale. Similarly, Figures 11(c) and 11(d) show the same raw data obtained at emission point P(2), with Figure 11(d) being an enlarged version of Figure 11(c) on the y-axis scale. The x-axis scale in Figures 11(a) to (d) represents time from 0 to 500, the y-axis scale in Figures 11(a) and (c) represents fluorescence intensity from 0 to 250, and the y-axis scale in Figures 11(b) and (d) represents fluorescence intensity from -0.2 to 0.8. X(1,1) and X(2,1) are represented by solid lines, and X(1,2) and X(2,2) are represented by dotted lines. As shown in Figure 11, the sample was prepared so that phosphor D(1,1) (state as in Figure 9(a)) fluoresces independently at emission point P(1) at time 100, phosphor D(2,1) (state as in Figure 9(c)) at emission point P(2) at time 200, phosphor D(1,2) (state as in Figure 9(b)) at emission point P(1) at time 300, and phosphor D(2,2) (state as in Figure 9(d)) at emission point P(2) at time 400. The sample was prepared so that no other fluorescence is emitted. As can be seen in Figures 11(a) and 11(c), only the four large peaks corresponding to the fluorescence emission described above are observed (corresponding to (i) and (ii) in Figure 9). However, looking at Figures 11(b) and 11(d), four small peaks indicated by arrows are observed: phosphor D(2,1) at emission point P(2) at time 100, phosphor D(1,1) at emission point P(1) at time 200, phosphor D(2,2) at emission point P(2) at time 300, and phosphor D(1,2) at emission point P(1) at time 400 (corresponding to (iii) and (iv) in Figure 9). Therefore, it can be concluded that these four small peaks are due to spatial crosstalk and spectral crosstalk.For example, the small peak of phosphor D(2,1) at emission point P(2) at time 100 is the result of some of the fluorescence emission of phosphor D(1,1) at emission point P(1) at time 100 being detected as X(2,1) and X(2,2) due to spatial and spectral crosstalk (corresponding to (iii) and (iv) in Figure 9(a)).

[0055] If spatial crosstalk is not considered, that is, if the four small peaks in Figures 11(c) and 11(d) are ignored (only (i) and (ii) in Figure 9 are considered), then the spectral crosstalk of the four large peaks in Figures 11(a) and 11(c) can be resolved, i.e., color conversion can be performed using the conventional method. In this case, (Equation 6) is expressed as follows.

[0056]

number

number

[0057] Here, Z(1,1), Z(1,2), Z(2,1), and Z(2,2) represent the concentrations of the phosphors D(1,1), D(1,2), D(2,1), and D(2,2) at each time step. As shown in (Equation 18) and (Equation 19), in this embodiment, a 2x2 matrix Y and its inverse matrix Y -1 The matrix Y and its inverse matrix Y were the same at light emission point P(1) and light emission point P(2). However, in general, the matrix Y and its inverse matrix Y were the same at light emission point P(1) and light emission point P(2). -1 These may differ. Each element of matrix Y for emission point P(1) was determined by the intensity ratio of fluorescence intensities X(1,1) and X(1,2) from phosphor D(1,1) at time 100, and the intensity ratio of fluorescence intensities X(1,1) and X(1,2) from phosphor D(1,2) at time 300. Each element of matrix Y for emission point P(2) was determined by the intensity ratio of fluorescence intensities X(2,1) and X(2,2) from phosphor D(2,1) at time 200, and the intensity ratio of fluorescence intensities X(2,1) and X(2,2) from phosphor D(2,2) at time 400.

[0058] Figure 12 shows the color-transformed data obtained by performing color transformations (Equation 18) and (Equation 19) on the raw data from Figure 11 at each time step. The notation is the same as in Figure 11. As expected, the spectral crosstalk of the four large peaks in Figures 11(a) and 11(c) is resolved by the four large peaks in Figures 12(a) and 12(c). On the other hand, the spectral crosstalk of the four small peaks in Figures 11(b) and 11(d) is also resolved by the four small peaks (similarly indicated by arrows) in Figures 12(c) and 12(d), but these peaks themselves still remain, indicating that spatial crosstalk has not been resolved. This is a challenge of conventional methods.

[0059] Next, for Figure 11, we consider both spectral crosstalk and spatial crosstalk (considering all of (i) to (iv) in Figure 9), and perform both color conversion and spatial correction. In this case, (Equation 14) is expressed by the following formula.

[0060]

number

[0061] Matrix Y and its inverse matrix Y -1The format has been expanded from 2 rows and 2 columns in (Equation 18) and (Equation 19) to 4 rows and 4 columns. Also, while (Equation 18) and (Equation 19) performed color conversion independently for light source P(1) and light source P(2), (Equation 20) combines light source P(1) and light source P(2) and performs color conversion and spatial correction all at once. Each element of matrix Y in (Equation 20) was determined by the intensity ratios of fluorescence intensities X(1,1), X(1,2), X(2,1), and X(2,2) from phosphor D(1,1) at time 100, X(1,1), X(1,2), X(2,1), and X(2,2) from phosphor D(2,1) at time 200, X(1,1), X(1,2), X(2,1), and X(2,2) from phosphor D(1,2) at time 300, and X(1,1), X(1,2), X(2,1), and X(2,2) from phosphor D(2,2) at time 400, as shown in Figure 11. The top-left 2x2 and bottom-right 2x2 columns of matrix Y in (Equation 20) correspond to matrix Y in (Equation 18) and (Equation 19), and the values ​​of each element are approximately equal to each other. Similarly, matrix Y in (Equation 20) -1 The top left 2x2 and bottom right 2x2 rows are the matrices Y of (Equation 18) and (Equation 19). -1 These correspond to each other, and the values ​​of each element are approximately equal to each other. In other words, these elements are responsible for the color transformation that eliminates spectral crosstalk for the same light source. The reason why the values ​​of these elements do not perfectly match in (Equation 20), (Equation 18), and (Equation 19) is the difference in whether or not spatial crosstalk is considered, which will be explained next. In contrast, the matrices Y and Y in (Equation 20) -1 The top right 2x2 and bottom left 2x2 rows are responsible for spatial correction and color conversion, which eliminate spatial and spectral crosstalk between different light-emitting points, and are not present in (Equation 18) and (Equation 19). The above becomes clearer when (Equation 20) is transformed as follows.

[0062]

number

number

[0063] The first term on the right-hand side of equation (21) and the second term on the right-hand side of equation (22) are responsible for the color conversion in the conventional method, corresponding to equations (18) and (19), respectively. In contrast, the second term on the right-hand side of equation (21) and the first term on the right-hand side of equation (22) are responsible for spatial correction and color conversion, which are not handled in the conventional method. Thus, color conversion and spatial correction and color conversion can be separated and performed individually, or only one of them can be performed. Alternatively, it is possible to change the processing depending on the light source, for example, by applying spatial correction and color conversion to light source P(1) and only color conversion to light source P(2).

[0064] Figure 13 shows the color-transformed and spatially corrected data obtained by performing color transformation + spatial correction (Equation 20) on the raw data from Figure 11 at each time step. The notation is the same as in Figure 11. First, as with Figure 12, it can be seen that the spectral crosstalk of the four large peaks in Figures 11(a) and 11(c) is resolved in the four large peaks in Figures 13(a) and 13(c). Furthermore, as expected, the spatial and spectral crosstalk of the four small peaks in Figures 11(b) and 11(d) is also resolved in Figures 13(c) and 13(d), and the peaks disappear (indicated by arrows).

[0065] In summary, this disclosure demonstrates that it is possible to eliminate or reduce, through computational processing, spectral crosstalk between multiple emission points, spatial crosstalk between multiple emission points, and spectral crosstalk that occur in any optical system for detecting the fluorescence of multiple types of phosphors emitted from multiple emission points.

[0066] [Example 3] Next, based on the experimental results of [Example 2], an unknown sample was analyzed. Figure 14 shows the raw data obtained by electrophoretic analysis of the unknown sample after injecting it into capillaries Cap(1) and Cap(2) using the model experimental system of Figure 8. Since the same model experimental system as in [Example 2] is used, equations (18) to (22) can be used as is. The notation is the same as in Figure 11. However, the vertical axis scale in Figures 11(b) and (d) has been reduced to fluorescence intensity -0.5 to 2.0. At times 100, 200, 300, and 400, a large peak was observed at emission point P(1), while a small peak (indicated by the arrow) was observed at emission point P(2).

[0067] Figure 15 shows the color-converted data obtained by performing color conversions (Equation 18) and (Equation 19) on the raw data from Figure 14 at each time point. The notation is the same as in Figure 14. As shown in Figure 15(a), at emission point P(1), the emission fluorescence of phosphor D(1,1) was detected alone at times 100 and 300, and the emission fluorescence of phosphor D(1,2) was detected alone at times 200 and 400. On the other hand, as shown in Figure 15(d), the identity of the four small peaks (indicated by arrows) detected at emission point P(2) at times 100, 200, 300, and 400 remains unknown. In other words, it was unclear whether the origin of these fluorescence was (1) fluorescence emission from a mixture of phosphors D(2,1) and D(2,2) at emission point P(2), (2) fluorescence emission from impurities other than phosphors D(2,1) and D(2,2) at emission point P(2), or (3) spatial crosstalk of fluorescence emission from phosphors D(1,1) and D(1,2) at emission point P(1).

[0068] Figure 16 shows the color-converted and spatially corrected data obtained by applying color conversion + spatial correction (equation 20) to the raw data in Figure 14 at each time step. The notation is the same as in Figure 14. As shown in Figure 16(a), similar to Figure 15(a), at emission point P(1), the emission fluorescence of phosphor D(1,1) was detected alone at times 100 and 300, and the emission fluorescence of phosphor D(1,2) was detected alone at times 200 and 400. In addition, as shown in Figure 16(d), spatial and spectral crosstalk between the fluorescence emission of phosphors D(1,1) and D(1,2) at emission point P(1) and the detection regions W(2,1) and W(2,2) was eliminated. As a result, it was found that at emission point P(2), weak fluorescence emission from phosphor D(2,2) was detected independently at times 100 and 300, and weak fluorescence emission from phosphor D(2,1) was detected independently at times 200 and 400. The peak intensities of these weak fluorescence emission were slightly less than 1% of the peak intensity of fluorescence emission observed at emission point P(1). On the other hand, as can be seen from the results in Figure 12, the spatial crosstalk ratio generated in the model experimental system in Figure 8 is also slightly less than 1%. Therefore, in Figure 15, true weak fluorescence emission and false weak fluorescence emission due to spatial crosstalk were mixed together, making it impossible to distinguish between the two.

[0069] The results above demonstrate that spatial crosstalk can raise the detection limit in detecting light emission from each light source. As in the example above, if the spatial crosstalk ratio is 1%, even if the detection limit for a single light source is 0.1%, signals below 1% cannot be distinguished as either a true signal or a false signal due to spatial crosstalk, so the effective detection limit rises to 1%. In other words, compared to detecting light emission from a single light source, the detection sensitivity and dynamic range are both an order of magnitude lower when detecting light emission from multiple light sources. This disclosure solves such problems and can avoid the decrease in detection sensitivity and dynamic range when detecting light emission from multiple light sources.

[0070] [Example 4] Next, based on the experimental results of [Example 2], a substance labeled with an unknown phosphor D(1,3) together with phosphors D(1,1) and D(1,2) was subjected to electrophoretic analysis at emission point P(1). Figure 17 shows the raw data obtained by electrophoretic analysis of both capillary caps (1) and (2) using the model experimental system of Figure 8, where the sample was injected only into capillary cap (1) and not into capillary cap (2). The notation method is the same as in Figure 14. The sample was prepared so that phosphor D(1,1) fluoresced individually at emission point P(1) at time 100, phosphor D(1,2) fluoresced individually at emission point P(1) at time 200, phosphor D(1,1) fluoresced individually at emission point P(1) at time 300, and phosphor D(1,3) fluoresced individually at emission point P(1) at time 400. The sample was prepared so that no other fluorescence was emitted. As shown in Figure 17(a), only the four large peaks corresponding to the fluorescence emission described above were observed. The spectral crosstalk ratio of phosphor D(1,3) was similar to that of phosphor D(1,2), but slightly different. On the other hand, the four small peaks indicated by arrows in Figure 17(d) can be judged to be the result of the spatial and spectral crosstalk of the emission fluorescence that gives rise to the four large peaks in Figure 17(a).

[0071] Figure 18 shows the color-converted and spatially corrected data obtained by applying color conversion + spatial correction (Equation 20) to the raw data in Figure 17 at each time step. The notation is the same as in Figure 14. As shown in Figure 18(a), at emission point P(1), the emission fluorescence of phosphor D(1,1) was detected alone at times 100 and 300, and the emission fluorescence of phosphor D(1,2) was detected alone at time 200. However, as indicated by the arrow, the emission fluorescence of phosphor D(1,3) was detected alone at time 400, but since the spectral crosstalk ratio was different from both phosphor D(1,1) and phosphor D(1,2), spectral crosstalk was not resolved. On the other hand, as shown in Figure 18(d), although the spatial and spectral crosstalk due to the fluorescence emission of phosphors D(1,1) and D(1,2) at emission point P(1) was resolved at times 100, 200, and 300, it was found that the spatial and spectral crosstalk due to the fluorescence emission of phosphor D(1,3) at emission point P(1) at time 400 remained, as indicated by the arrows, and a peak persisted. This is because, in [Example 2], equation (20) was derived for the fluorescence emission of phosphors D(1,1) and D(1,2) at emission point P(1), and phosphors D(2,1) and D(2,2) at emission point P(2), and the characteristics of the spatial and spectral crosstalk of the fluorescence emission of D(1,3) at emission point P(1) are different from those of equation (20). Therefore, the above problem can be solved by re-acquiring (Equation 20) for the emission fluorescence of phosphor D(1,3) at emission point P(1). The above phenomenon represents one aspect of this disclosure.

[0072] [Example 5] Next, building on the experimental results of [Example 2], an experiment was conducted to induce fluorescence emission by gradually increasing the concentration of a substance labeled with phosphor D(1,1) at emission point P(1). Figure 19 shows the raw data obtained by electrophoretic analysis of both capillary cap(1) and cap(2) using the model experimental system of Figure 8, where the sample was injected only into capillary cap(1) and not into capillary cap(2). The notation method is the same as in Figure 14. However, the vertical axis scale in Figures 19(a) and (c) has been reduced to fluorescence intensity 0 to 400. At times 100, 200, 300, and 400, phosphor D(1,1) fluoresced independently at emission point P(1), and the samples were prepared so that the concentration and fluorescence intensity of phosphor D(1,1) increased stepwise. Furthermore, the samples were prepared so that no other fluorescence was emitted. As shown in Figure 19(a), four large peaks corresponding to the fluorescence emission described above were observed. However, since the sensor S used in the model experimental system in Figure 8 saturates at a fluorescence intensity of 200, three peaks at times 200, 300, and 400 were detected as saturated. In contrast, the four small peaks shown in Figure 19(d), which are caused by spatial and spectral crosstalk of the fluorescence emission of phosphor D(1,1) at emission point P(1), were observed without saturation due to their low fluorescence intensity.

[0073] Figure 20 shows the color-converted data obtained by performing color conversions (Equation 18) and (Equation 19) on the raw data from Figure 19 at each time step. The notation is the same as in Figure 19. As shown in Figure 20(a), the emission fluorescence of phosphor D(1,1) at time 100 at emission point P(1) was detected independently, with the spectral crosstalk eliminated. However, the emission fluorescence of phosphor D(1,1) at times 200, 300, and 400 was saturated, and therefore the spectral crosstalk was not eliminated. This is because, as is clear from Figure 19(a), when the fluorescence intensity is saturated, the spectral crosstalk ratio derived from the intensity ratio of X(1,1) and X(1,2) changes, deviating from the spectral crosstalk ratio defined by (Equation 18) and (Equation 19). In contrast, as shown in Figure 20(d), since the fluorescence intensity of phosphor D(2,1) is not saturated, the spectral crosstalk was eliminated for each of the four small peaks indicated by the arrows. The intensity ratios of the four peaks shown in Figure 20(d) represent the intensity ratios of the fluorescence emission of phosphor D(1,1) at emission point P(1) at times 100, 200, 300, and 400.

[0074] Figure 21 shows the color-transformed and spatially corrected data obtained by performing color transformation + spatial correction (Equation 20) on the raw data from Figure 19 at each time step. The notation is the same as in Figure 19. In Figure 21(a), the same results are obtained for the same reasons as in Figure 20(a). In contrast, as shown in Figure 21(d), both spatial and spectral crosstalk were eliminated for the emission fluorescence of phosphor D(2,1) at time 100. However, for the emission fluorescence of phosphor D(2,1) at times 200, 300, and 400, both spatial and spectral crosstalk were not eliminated. This is because, at times 200, 300, and 400, the fluorescence intensity of X(1,1) or X(1,2) shown in Figure 19(a) saturated, resulting in a deviation from the spatial crosstalk ratio defined by (Equation 20) for the ratio of the fluorescence intensities of X(2,1) and X(2,2) shown in Figure 19(d) to those X(1,1) and X(1,2). Furthermore, the degree of this deviation increases with the degree of saturation, so in Figure 21(d), the fluorescence intensity of phosphor D(2,1) increased from time 200 to time 300, and from time 300 to time 400.

[0075] [Example 6] Here, we describe a method for determining matrix Y (Equation 20) using a different method than in [Example 2]. In [Example 2], as shown in Figure 11(a), matrix Y was determined by making each of the phosphors D(1,1), D(2,1), D(1,2), and D(2,2) fluoresce individually in that order, that is, by alternately causing fluorescence emission at emission point P(1) and emission point P(2). However, if the electrophoretic velocity of either capillary Cap(1) or Cap(2) deviates from the expected value, there is a possibility that fluorescence will be emitted simultaneously at emission point P(1) and emission point P(2), in which case matrix Y cannot be determined.

[0076] Figure 22 shows a more practical method to avoid the above problems. First, as shown in Figure 22(a), the sample is injected only into capillary Cap(1), and electrophoretic analysis is performed on capillary Cap(1) and Cap(2) to cause the phosphors D(1,1) and D(1,2) to fluoresce individually, and fluorescence detection is performed in the detection regions W(1,1), W(1,2), W(2,1), and W(2,2). Next, as shown in Figure 22(b), the sample is injected only into capillary Cap(2), and electrophoretic analysis is performed on capillary Cap(1) and Cap(2) to cause the phosphors D(2,1) and D(2,2) to fluoresce individually, and fluorescence detection is performed in the detection regions W(1,1), W(1,2), W(2,1), and W(2,2). By this method, matrix Y can be determined in the same way as in [Example 2]. Furthermore, this method allows for a simpler and more reliable determination of matrix Y than in [Example 2]. In Figure 22, for clarity, the phosphors D(1,1) and D(1,2) in capillary Cap(1) and D(2,1) and D(2,2) in capillary Cap(2) are depicted as fluorescing multiple times, but fluorescing once each is sufficient. Also, in the above, the next electrophoresis analysis is performed only after the first electrophoresis analysis is complete, but it is not necessary to extend the interval between electrophoresis analyses that far. The same effect as in Figure 22 can be achieved by appropriately staggering the timing of sample injection into capillary Cap(1) and capillary Cap(2). For example, after injecting the sample into capillary Cap(1), electrophoresis can be performed for a short time, and then the sample can be injected into capillary Cap(2) before resuming electrophoresis. This reduces the time required to determine matrix Y. Furthermore, this method allows the sample injected into capillary Cap (1) and the sample injected into capillary Cap (2) to have the same composition or be identical. This simplifies the preparation of samples and reduces associated costs.

[0077] Figure 23 shows a method to further simplify the determination of matrix Y. Here, even though the sample injection into capillary Cap(1) and capillary Cap(2) is performed at the same time, the samples are prepared so that the phosphors D(1,1) and D(1,2) in capillary Cap(1) and the phosphors D(2,1) and D(2,2) in capillary Cap(2) fluoresce at different times, similar to Figure 22. In other words, the composition of the sample injected into capillary Cap(1) and the sample injected into capillary Cap(2) should be changed so that there is a difference in the electrophoretic rate of the substances labeled with each phosphor. For example, the sample to be injected into capillary Cap(1) may contain a 50-base DNA fragment labeled with phosphor D(1,1) and a 60-base DNA fragment labeled with phosphor D(1,2), while the sample to be injected into capillary Cap(2) may contain a 70-base DNA fragment labeled with phosphor D(2,1) and an 80-base DNA fragment labeled with phosphor D(2,2).

[0078] Alternatively, Figure 23 can also be achieved by setting different electrophoretic conditions for capillaries Cap(1) and Cap(2), even if samples of the same composition are injected into Capillaries Cap(1) and Cap(2) at the same time. For example, this could be done by temporarily lowering the applied voltage to Capi(2) during electrophoresis, or by lowering the temperature of Capi(2) during electrophoresis.

[0079] The above assumed the preparation of a dedicated sample for determining matrix Y, but this is not always necessary. If there is a state in the raw electrophoretic data of the actual sample to be analyzed in which fluorescence emission occurs independently in each capillary, and this can be identified, then matrix Y can be determined using the raw data at that time.

[0080] [Example 7] Figure 24 is a diagram of a multi-capillary electrophoresis apparatus, which is an example of an analytical instrument. Multi-capillary electrophoresis apparatuses are widely used as analytical instruments for DNA sequencing and DNA fragment analysis. As shown in Figure 24, the multi-capillary electrophoresis apparatus comprises capillary 24-1, cathode 24-4, anode 24-5, cathode buffer 24-6, anode buffer 24-7, pump block 24-9, syringe 24-11, and laser light source 24-12. In this example, four capillaries 24-1 were used, and DNA sequencing of different samples was performed in each capillary 24-1. The DNA sequencing samples consisted of DNA fragments labeled with four types of phosphors. One analysis session was performed by the following steps (1) to (6). (1) First, the sample injection ends 24-2 of the four capillaries 24-1 were immersed in the cathode buffer 24-6, and the sample elution ends 24-3 were immersed in the anode buffer 24-7 via the polymer block 24-9. (2) Next, the valve 24-10 of the pump block 24-9 was closed, and the polymer solution inside was pressurized by pushing down the piston of the syringe 24-11 connected to the pump block 24-9, filling the polymer solution into each capillary 24-1 from the sample elution end 24-3 toward the sample injection end 24-2. (3) Subsequently, the valve 24-10 was opened, and different samples were injected into each capillary 24-1 from the sample injection end 24-2 using an electric field injection. Then, capillary electrophoresis was started by applying a high voltage between the cathode 24-4 and the anode 24-5 using the power supply 24-8. DNA fragments labeled with four types of phosphors were electrophoresed from the sample injection end 24-2 toward the sample elution end 24-3. (4) In parallel, the position of each capillary 24-1 electrophoresed a certain distance from the sample injection end 24-2 was designated as the emission point 24-14, and a laser beam 24-13 with a wavelength of 505 nm emitted from the laser light source 24-12 was simultaneously irradiated onto each emission point 24-14. Here, the coating of each capillary 24-1 near the emission point 24-14 was removed in advance, and each capillary 24-1 near the emission point 24-14 was arranged on the same plane. The laser beam 24-13 was focused and then introduced from the side of the above-mentioned arrangement plane along the arrangement plane.(5) Then, DNA fragments labeled with four types of phosphors underwent electrophoresis inside each capillary 24-1, and were excited by the irradiation of the laser beam 24-13 as they passed through the emission point 24-14, causing them to emit fluorescence. In other words, four types of phosphors emitted fluorescence from the four emission points, and the fluorescence intensity of each changed moment by moment in conjunction with electrophoresis. (6) Finally, the fluorescence emitted from each emission point was detected in multiple colors by a sensor (not shown), and the DNA sequence of the samples injected into each capillary was performed by analyzing the time-series data obtained by a computer (not shown). The analysis session consisting of the above steps (1) to (6) can be repeated multiple times. For example, by analyzing samples (1) to (4) in the first analysis session, analyzing samples (5) to (8) in the second analysis session, and so on, a large number of different samples can be analyzed.

[0081] In this embodiment, the multi-color detection described in (6) above was performed using the optical system described in Patent Document 1. Specifically, the fluorescence emitted from four emission points P(1) to P(4) was collimated with one focusing lens, passed through one transmission-type diffraction grating, and the first-order diffracted light from each fluorescence was imaged onto a two-dimensional sensor using one imaging lens. Figure 25 shows the acquisition result of a two-dimensional sensor image including the wavelength dispersion image of Raman scattered light emitted from four emission points P(1) to P(4) by laser beam irradiation, when four capillaries Cap(1) to Cap(4) are filled with a standard solution. The horizontal axis is the direction of the arrangement of the four capillaries, and the vertical axis is the wavelength direction. The four streaky images extending vertically, indicated by arrows, are the wavelength dispersion images of the Raman scattered light from capillaries Cap(1) to Cap(4), respectively. Based on these results, wavelength calibration of the wavelength dispersion image on the two-dimensional sensor image was performed for each emission point, and the relationship between pixel position and wavelength was determined. Based on these results, as shown in Figure 26, B=20 detection regions with different wavelength bands were set for each of the light emission points P(1) to P(4) on the 2D sensor image. Each of the 20 detection regions was set to detect light emission in the 500-700 nm range, divided into wavelength bands at 10 nm intervals. For example, the detection regions W(1,1), W(1,2), ..., W(1,20) set for light emission point P(1) were set to detect light emission in the wavelength bands of 500-510 nm, 510-520 nm, ..., 690-700 nm, respectively, and their respective signal intensities were set to X(1,1), X(1,2), ..., X(1,20). The same procedure was followed for light emission points P(2) to P(4). Here, as long as the positions of each light emission point P(1) to P(4) and the optical system are fixed, the relationship between the pixel position and wavelength for each capillary in the 2D sensor image is maintained, so the detection region set in Figure 26 is valid for multiple different light emission detections or multiple different analysis sessions.

[0082] In this example, dR110, dR6G, dTAMRA, and dROX were used as four types of phosphors, and each was used to label DNA fragments with terminal base species T, C, A, and G prepared by the Sanger reaction. The maximum fluorescence emission wavelengths of dR110, dR6G, dTAMRA, and dROX are 541 nm, 568 nm, 595 nm, and 618 nm, respectively. Therefore, the fluorescence emission of each is detected with the strongest intensity in the detection regions of wavelengths 540-550 nm, 560-570 nm, 590-600 nm, and 610-620 nm. For example, the fluorescence emission of dR110, dR6G, dTAMRA, and dROX at emission point P(1) is mainly detected in detection regions W(1,5), W(1,7), W(1,10), and W(1,12), respectively. However, each fluorescence emission is detected in other detection regions of emission point P(1) due to spectral crosstalk, and is also detected at a weak intensity in the detection regions of emission points P(2) to P(3) due to spatial and spectral crosstalk. Hereafter, dR110, dR6G, dTAMRA, and dROX at emission point P(1) are denoted as D(1,1), D(1,2), D(1,3), and D(1,4), respectively, and their concentrations are denoted as Z(1,1), Z(1,2), Z(1,3), and Z(1,4). The same applies to emission points P(2) to P(4).

[0083] In this case, (Equation 10) to (Equation 12) in (Equation 9) become the following equations.

number

number

number

[0084] Here, X is an 80x1 matrix, Y is an 80x16 matrix, and Z is a 16x1 matrix. By substituting (Equation 23) to (Equation 25) into (Equation 14) and performing spatial correction and color transformation, both spatial and spectral crosstalk were eliminated, and time-series data of the concentrations of four types of phosphors in each capillary could be obtained. As a result, DNA sequencing of different samples injected into each capillary could be performed. Note that the conventional method corresponding to this embodiment involves performing the color transformation (Equation 8) for each capillary. The conventional method cannot eliminate spatial crosstalk.

[0085] Equations (23) to (25), and (14) derived therefrom, are valid for multiple different emission detections or multiple different analysis sessions, provided that the position of each emission point, the optical system, and the phosphor used are fixed.

[0086] The method for determining matrix Y is as already described. However, it is not always necessary to set all 80 × 16 elements. For example, elements with sufficiently small absolute values ​​compared to other elements can be replaced with zero, simplifying the related calculations. Also, if the range affected by spatial crosstalk is limited, (Equation 9) to (Equation 12) can be defined within that limited range. In this case, the limited range can be sequentially shifted as needed to perform a wider-range analysis. For example, if the range affected by spatial crosstalk can be limited to adjacent capillaries, spatial and spectral crosstalk for capillaries two or more distances away can be ignored, and the related calculations can be omitted. In this case, it is also possible to simplify the process of emitting light individually at each emission point to determine matrix Y to a process of emitting light simultaneously at multiple emission points that do not affect each other with spatial crosstalk.

[0087] [Example 8] In this example, DNA sequencing was performed using the multi-capillary electrophoresis apparatus shown in Figure 24, similar to Example 7. However, the fluorescence emission of four types of phosphors emitted from four emission points was detected in multiple colors at each time point using the optical system described in Patent Document 2. Figure 27 shows the configuration of this optical system and the four-part image of the four emission points acquired by this optical system. The optical system in Figure 27 comprises a focusing lens array 27-1, a long-pass filter 27-3, dichroic mirrors 27-4, 27-5, 27-6, and 27-7, and a two-dimensional sensor 27-8. As shown in Figures 27(a) and 27(b), first, the fluorescence emitted from A=4 emission points P(1) to P(4) was individually collimated by the four focusing lenses 27-2 constituting the focusing lens array 27-1 to form four light beams 27-9, and these were then transmitted together through a single long-pass filter 27-3 to cut off the laser beam light. Next, each beam of light 27-9 was incident on a set of dichroic mirror arrays consisting of four dichroic mirrors 27-4, 27-5, 27-6, and 27-7, thereby splitting each beam of light 27-9 into four divided beams 27-10, 27-11, 27-12, and 27-13, respectively, which were then emitted from the dichroic mirror array. Here, by adjusting the spectral characteristics of each dichroic mirror, the divided beams 27-10, 27-11, 27-12, and 27-13 were made to have light in the wavelength bands of 520-550 nm, 550-580 nm, 580-610 nm, and 610-640 nm, respectively. Finally, the generated A×B=4×4=16 divided light beams were incident on the 2D sensor 27-8, and 16 divided images W(1,1)~W(4,4) were obtained as shown in Figure 27(c). Here, the arrangement direction of the four capillaries Cap(1)~Cap(4), i.e., the arrangement direction of the light emission points P(1)~P(4), and the division direction by the dichroic mirror array, i.e., the wavelength direction, are perpendicular to each other. Therefore, as shown in Figure 27(c), the divided images W(1,1)~W(4,4) were detected in the 2D sensor image 27-14 in an aligned state without overlapping.For example, the fluorescence emitted from emission point P(1) was detected as split images W(1,1), W(1,2), W(1,3), and W(1,4), each possessing light components in the wavelength bands of 520-550 nm, 550-580 nm, 580-610 nm, and 610-640 nm, respectively. The same applies to emission points P(2) to P(4). Therefore, the split images W(1,1) to W(4,4) in the 2D sensor image 27-14 were set as detection regions. The signal intensities detected in each detection region were denoted as X(1,1) to X(4,4), respectively. Here, as long as the positions of each emission point P(1) to P(4) and the optical system are fixed, the relationship between the pixel position and wavelength for each capillary in the 2D sensor image 27-14 is maintained, so the detection regions set in Figure 27(c) are valid for multiple different emission detections or multiple different analysis sessions.

[0088] In this example, similar to [Example 7], dR110, dR6G, dTAMRA, and dROX were used as C=4 types of phosphors, and each was used to label DNA fragments with terminal base species T, C, A, and G prepared by the Sanger reaction. The maximum fluorescence emission wavelengths of dR110, dR6G, dTAMRA, and dROX are 541 nm, 568 nm, 595 nm, and 618 nm, respectively. Therefore, the emission fluorescence of each is detected with the strongest intensity in the detection region of the resolution image having light components in the wavelength bands of 520-550 nm, 550-580 nm, 580-610 nm, and 610-640 nm. For example, the emission fluorescence of dR110, dR6G, dTAMRA, and dROX at emission point P(1) is mainly detected in detection regions W(1,1), W(1,2), W(1,3), and W(1,4), respectively. However, each emission fluorescence is detected in other detection regions of emission point P(1) due to spectral crosstalk, and in the detection regions of emission points P(2) to P(3) due to spatial crosstalk and spectral crosstalk. Hereafter, dR110, dR6G, dTAMRA, and dROX at emission point P(1) are denoted as phosphors D(1,1), D(1,2), D(1,3), and D(1,4), respectively, and their concentrations are denoted as Z(1,1), Z(1,2), Z(1,3), and Z(1,4). The same applies to emission points P(2) to P(4).

[0089] In this case, (Equation 10) to (Equation 12) in (Equation 9) become the following equations.

number

number

number

[0090] Here, X is a 16x1 matrix, Y is a 16x16 matrix, and Z is a 16x1 matrix. By substituting (Equation 26) to (Equation 28) into (Equation 14), both spatial and spectral crosstalk were eliminated, and time-series data of the concentrations of four types of phosphors in each capillary could be obtained. As a result, DNA sequencing of different samples injected into each capillary could be performed. Note that the conventional method corresponding to this embodiment involves performing the color conversion (Equation 8) for each capillary. The conventional method cannot eliminate spatial crosstalk.

[0091] Equations (26) to (28), and (14) derived therefrom, are valid for multiple different emission detections or multiple different analysis sessions, provided that the position of each emission point, the optical system, and the phosphor used are fixed.

[0092] [Example 9] Figure 28 shows a diagram of the configuration of a multi-channel DNA extension reactor, in which multiple reaction vessels are arranged on a plane, and complementary strand extension reactions are performed in single-base units in each channel using different DNA fragments as templates. DNA sequencing of each DNA fragment is performed by multi-color detection of the emission fluorescence of four types of C=4 types of phosphors labeled on the four types of bases incorporated into the complementary strand during the extension reaction. The diagram also shows schematic diagrams of two-dimensional sensor images obtained from multiple two-dimensional sensors. The multi-channel DNA extension reactor comprises a laser light source 28-1, a dichroic mirror 28-3, a lens 28-4, dichroic mirrors 28-8, 28-9 and 28-10, a lens 28-11, a first two-dimensional sensor 28-12, a lens 28-13, a second two-dimensional sensor 28-14, a lens 28-15, a third two-dimensional sensor 28-16, a lens 28-17, and a fourth two-dimensional sensor 28-18. The maximum fluorescence emission wavelengths of the phosphors labeled with base species T, C, A, and G were set to 535 nm, 565 nm, 595 nm, and 625 nm, respectively. First, a laser beam 28-2 emitted from a laser light source 28-1 was passed through a dichroic mirror 28-3, focused by a lens 28-4, and irradiated onto a multi-channel sample 28-5. Next, the fluorescence emission 28-6 of each phosphor excited by the laser beam irradiation in each channel was collimated collectively by the lens 28-4, and the resulting light beam 28-7 was reflected by the dichroic mirror 28-3. The dichroic mirror 28-3 has spectral properties that transmit laser beam light and reflect emission fluorescence. Subsequently, the light beam 28-7 was split into four light beams having four different wavelength components using three types of dichroic mirrors 28-8, 28-9, and 28-10. Then, the first divided light beam was imaged onto the first two-dimensional sensor 28-12 by lens 28-11, the second divided light beam was imaged onto the second two-dimensional sensor 28-14 by lens 28-13, the third divided light beam was imaged onto the third two-dimensional sensor 28-16 by lens 28-15, and the fourth divided light beam was imaged onto the fourth two-dimensional sensor 28-18 by lens 28-17.

[0093] Sample 28-19 is a schematic diagram of Sample 28-5 observed from the front. Five channels (A) each formed emission points P(1) to P(5). In reality, many more channels exist on the sample, and the fluorescence emission from each channel was detected collectively by the optical system described above. However, here, only the emission point P(3) of interest and the emission points P(1), P(2), P(4), and P(5) surrounding P(3), which directly affect P(3) with spatial crosstalk, are analyzed. Therefore, only emission points P(1) to P(5) are depicted in Sample 28-19. If an emission point other than P(3) is to be the focus, that emission point and its surrounding emission points can be analyzed similarly. In any case, when focusing on any emission point, the same fluorescence detection data can be analyzed individually after fluorescence detection. The first two-dimensional sensor image 28-20 is a segmented image of sample 28-19 acquired by the first two-dimensional sensor 28-12, with detection regions W(1,1) to W(5,1) set at each imaging point of emission points P(1) to P(5), and their respective signal intensities denoted as X(1,1) to X(5,1). The first two-dimensional sensor image 28-20 detected fluorescence components in the wavelength band 520 to 550 nm. The second two-dimensional sensor image 28-21 is a segmented image of sample 28-19 acquired by the second two-dimensional sensor 28-14, with detection regions W(1,2) to W(5,2) set at each imaging point of emission points P(1) to P(5). The second two-dimensional sensor image 28-21 detected fluorescence components in the wavelength band 550 to 580 nm. The third two-dimensional sensor image 28-22 is a segmented image of sample 28-19 acquired by the third two-dimensional sensor 28-16, with detection regions W(1,3) to W(5,3) set at each imaging point of emission points P(1) to P(5). The third two-dimensional sensor image 28-22 detected fluorescence components in the wavelength band 580 to 610 nm. The fourth two-dimensional sensor image 28-23 is a segmented image of sample 28-19 acquired by the fourth two-dimensional sensor 28-18, with detection regions W(1,4) to W(5,4) set at each imaging point of emission points P(1) to P(5). The first two-dimensional sensor image 28-23 detected fluorescence components in the wavelength band 610 to 640 nm.Specifically, the fluorescence emission from each emission point was detected using four 2D sensors, resulting in B=4 colors. For example, the fluorescence emission of the four types of phosphors at emission point P(3) is primarily detected in detection regions W(3,1), W(3,2), W(3,3), and W(3,4), respectively, in order of wavelength. However, each fluorescence emission is also detected in other detection regions of emission point P(3) due to spectral crosstalk, and in the detection regions of emission points P(1), P(2), P(4), and P(5) due to spatial and spectral crosstalk. Hereafter, the four types of phosphors at emission point P(3) are denoted as D(3,1), D(3,2), D(3,3), and D(3,4), respectively, in order of wavelength, and their concentrations are denoted as Z(3,1), Z(3,2), Z(3,3), and Z(3,4). The same applies to emission points P(1), P(2), P(4), and P(5).

[0094] In this case, (Equation 10) to (Equation 12) in (Equation 9) become the following equations.

number

number

number

[0095] Here, X is a 20x1 matrix, Y is a 20x20 matrix, and Z is a 20x1 matrix. By substituting (Equations 29) to (Equation 31) into (Equation 14), both spatial and spectral crosstalk were eliminated, and time-series data of the concentrations of the four types of phosphors in each channel could be obtained. However, in this embodiment, since only the emission point P(3) is being focused on, only the elements Z(3,1), Z(3,2), Z(3,3), and Z(3,4) were extracted from matrix Z. As a result, DNA sequencing could be performed in the channel of emission point P(3). A similar method can be used when focusing on other emission points and performing DNA sequencing in the corresponding channels.

[0096] [Example 10] In this embodiment, the conventional method and the method disclosed herein are summarized in flowcharts. Figure 29 is a flowchart of one analysis session of the conventional method. As shown in Figure 29, the analysis session of the conventional method is performed by an analysis system having an analyzer, a computer, a display device, and a database to analyze sample 29-1. The analyzer has a sensor (not shown) into which light from sample 29-1 is incident. First, sample 29-1 of type A (A is an integer of 2 or more) labeled with type C (C is an integer of 1 or more) phosphors is input to the analyzer. Next, in the analyzer, each sample is analyzed in parallel in step 29-2, and for each analysis and at each time step, emission fluorescence is detected in the wavelength band of type B (B is an integer of 1 or more), and time-series raw data of A × B fluorescence intensities X(a,b) are obtained, and these time-series raw data are sent to the computer. Here, a = 1, 2, ..., and A, b = 1, 2, ..., and B. Next, in the computer, step 29-3 is performed to analyze (a0) the matrix Y of row C and column B stored in the database. - The spectral crosstalk is eliminated by a color conversion (a0) using (a0), and at each time step, the concentration Z(a0,c) of C types of phosphors is determined from the fluorescence intensity X(a0,b). Here, a0 = 1, 2, ..., or A, c = 1, 2, ..., and C. Finally, the display device outputs the time-series color conversion data of Z(a0,c) in step 29-4. Steps 29-3 and 29-4 are performed for all a0. As shown in Figure 29, the conventional method is characterized by the fact that A analyses and interpretations are performed independently. Note that in (Equation 1) to (Equation 4), fluorescence intensity X(a,b) is denoted as X(b) and concentration Z(a0,c) is denoted as Z(c).

[0097] Figure 30 is a flowchart of one analysis session of this method. Similar to Figure 29, the analysis session of this embodiment shown in Figure 30 is performed by an analysis system having an analyzer, computer, display device, and database to analyze sample 29-1. The analyzer has a sensor (not shown) into which light from sample 29-1 is incident. First, sample 30-1 of type A (A is an integer of 2 or more) labeled with type C (C is an integer of 1 or more) phosphors is input to the analyzer. Next, in step 30-2, each sample is analyzed in parallel in the analyzer, and for each analysis and at each time step, emission fluorescence is detected in the wavelength band of type B (B is an integer of 1 or more), and time-series raw data of A × B fluorescence intensity X(a,b) are acquired, and these time-series raw data are sent to the computer. Here, a = 1, 2, ..., and A, b = 1, 2, ..., and B. However, unlike in Figure 29, as shown by the dotted arrow in step 30-2, there is significant spatial crosstalk of emission fluorescence between different analyses. Next, in the computer, step 30-3 calculates the matrix Y (A×C) rows and (A×B) columns stored in the database for all analyses (a). - By using color conversion and spatial correction, spectral and spatial crosstalk are eliminated, and at each time step, the concentrations Z(a,c) of type C phosphors for type A samples or analyses are determined collectively from the fluorescence intensity X(a,b). Here, a=1,2,..., and A, c=1,2,..., and C. Finally, in the display device, step 30-4 outputs time-series color conversion + spatial correction data of Z(a,c) for each analysis (a). As shown in Figure 30, a feature of this method is that the color conversion + spatial correction in step 30-3 is performed collectively for all analyses. Furthermore, a matrix Y common to all analyses is used for color conversion + spatial correction. - Another characteristic is the use of [a specific method / technique].

[0098] Figure 31 shows the matrix Y to be stored in the database, which is done before Figure 30. -This is a flowchart showing how to determine the matrix Y. First, a matrix Y determination sample 31-1 of type A (A is an integer of 2 or more) labeled with C types of phosphors (C is an integer of 1 or more) is input to the analyzer. This sample is prepared so that the fluorescence emission of phosphor (c) in analysis (a) occurs individually. That is, a=1,2,..., and A, c=1,2,..., and C, and fluorescence emission is produced for all A×C combinations, but fluorescence emission does not occur for two or more combinations simultaneously. Here, the above is achieved by the matrix Y determination sample 31-1, but it may also be achieved by setting the analyzer. Next, in step 31-2, the analyzer detects the above A×C individual fluorescence emissions in the wavelength band of type B (B is an integer of 1 or more), and time-series raw data of A×B fluorescence intensity X(a,b) is obtained, and this time-series raw data is sent to the computer. Next, in the computer, a (A×B) x (A×C) matrix is ​​derived from the (A×B) x (A×C) data obtained in step 31-3, and further, the generalized inverse of Y, a (A×C) x (A×B) matrix Y, is obtained. - We find the matrix Y. - The matrix Y is stored in the database and used in subsequent analyses such as Figure 30. - Make use of it.

[0099] Figure 32 shows the matrix Y obtained in Figure 31. - This flowchart shows the process of repeating the analysis sessions 30-2 to 30-4 for sample 30-1 of type A, as shown in Figure 30, multiple times, utilizing the following: Here, the sample 30-1 of type A being analyzed in each analysis session is different from one another. As shown in Figure 32, the same matrix Y stored in the database is used for multiple different analysis sessions. - A distinctive feature of this law is its utilization of [a specific resource / method].

[0100] Figure 33 shows an example of the computer configuration. The computer is connected to the analytical instrument. The computer performs not only data analysis but also controls the analytical instrument. In Figures 29 to 31, the database and display device are depicted outside the computer, but in Figure 33, they are depicted inside the computer. Data analysis conditions and analytical instrument control conditions are set via the keyboard, which is the input unit. Time-series raw data of fluorescence intensity X(a,b) output from the analytical instrument is sequentially stored in memory. Also, a matrix Y of (A×C) rows and (A×B) columns is stored in the database inside the HDD. - The fluorescence intensity X(a,b) and Y stored in memory are stored in memory. The CPU processes the fluorescence intensity X(a,b) and Y. - The product of the two values ​​is calculated to derive time-series color transformation + spatial correction data for the phosphor concentration Z(a,c), which is then sequentially stored in memory and simultaneously displayed on the monitor. Furthermore, the analysis results can be compared with information on the network via the network interface NIF.

[0101] [Differentiation] This disclosure is not limited to the embodiments described above, but includes various modifications. For example, the embodiments described above are described in detail for the purpose of illustrating this disclosure, and do not necessarily have to include all the configurations described. Furthermore, parts of one embodiment can be replaced with the configurations of another embodiment. Furthermore, configurations of other embodiments can be added to the configuration of one embodiment. Furthermore, parts of the configuration of each embodiment can be added, deleted, or replaced with parts of the configurations of other embodiments. [Explanation of Symbols]

[0102] 1-1 Pinhole plate 1-2 Light-emitting point side aperture plate 1-3 Focusing lens 1-4 Sensor-side opening plate 1-5 Color Glass Filters 1-6 2D sensors 1-7 Halogen lamp light 1-8 Light-emitting points 1-9 light 1-10 Luminous Images Cap(1) Capillary 1 Cap(2) Capillary 2 P(1) Light-emitting point on Cap(1) P(2) Emitting point on Cap(2) Phosphor 1 or absorber 1 on D(1,1) P(1) D(1,2) P(1) phosphor 2 or absorber 2 Phosphor 1 or absorber 1 on D(2,1) P(2) Phosphor 2 or absorber 2 on D(2,2) P(2) Phosphor 3 or absorber 3 on D(1,3) P(1) Detection area that primarily detects W(1,1) and D(1,1) Detection area that primarily detects W(1,2) and D(1,2) Detection area that primarily detects W(2,1) and D(2,1) Detection area that primarily detects W(2,2) D(2,2) Signal strength of X(1,1) W(1,1) Signal strength of X(1,2) W(1,2) Signal strength of X(2,1) W(2,1) Signal strength of X(2,2) W(2,2) Concentration of Z(1,1) D(1,1) Concentrations of Z(1,2) and D(1,2) Concentration of Z(2,1) D(2,1) Concentration of Z(2,2) D(2,2) S sensor LB laser beam LL lamp light 24-1 Capillary 24-2 Sample injection end 24-3 Sample elution edge 24-4 Cathode 24-5 Anode 24-6 Cathode buffer solution 24-7 Anode side buffer solution 24-8 Power supply 24-9 Polymer Block 24-10 Valve 24-11 Syringe 24-12 Laser light source 24-13 Laser beam 24-14 Light source Cap(a) Capillary a (a=1,2,3,and 4) P(a) Light-emitting point on capillary a (a=1,2,3, and 4) Detection region for the wavelength band b of emission from W(a) P(a) (a=1,2,..., and 4) (b=1,2,..., and 20) 27-1 Focusing lens array 27-2 Focusing lens 27-3 Long-pass filter 27-4, 27-5, 27-6, and 27-7 Dichroic Mirrors 27-8 2D Sensor 27-9 Luminous flux 27-10, 27-11, 27-12, and 27-13 divided luminous beams 27-14 2D sensor image P(a) Light emission point a (a=1,2,3,4,and 5) Detection region for the wavelength band b of emission from W(a,b) P(a) (a=1,2,3,4, and 5) (b=1,2,3, and 4) 28-1 Laser light source 28-2 Laser beam 28-3, 28-8, 28-9, and 28-10 Dichroic Mirrors 28-4, 28-11, 28-13, 28-15, and 28-17 lenses Samples 28-5 and 28-19 28-6 Fluorescence 28-7 Luminous flux 28-12, 28-14, 28-16, and 28-18 2D sensors 28-20, 28-21, 28-22, and 28-23 2D sensor images 29-1 Sample 29-2 Process in analytical instruments 29-3 Processes in Computers 29-4 Process in a display device 30-1 Sample 30-2 Process in analytical instruments 30-3 Processes in Computers 30-4 Process in a display device 31-1 Sample 31-2 Process in analytical instruments 31-3 Processes in Computers

Claims

1. A multi-capillary electrophoresis apparatus configured to perform parallel electrophoretic analysis of A (A is an integer of 2 or more) samples S(a) (a = 1, 2, ..., A) containing components labeled with C (C is an integer of 1 or more) types of phosphors D(a, c) (c = 1, 2, ..., C) using A capillaries, A number of light-emitting points P(a) (a = 1, 2, ..., A) fixed in position on the A capillaries, wherein fluorescence is emitted from the A light-emitting points P(a) when one or more laser beams are irradiated onto the A light-emitting points P(a) and the C type phosphors D(a, c) moving within the A capillaries are excited; An optical system configured to detect A × B (b = 1 or more integers) raw signals X(a, b) (b = 1 or more integers) of fluorescence emitted from the A light-emitting points P(a) of the C type phosphor D(a, c) in a wavelength band W(a, b) (b = 1, 2, ..., B) of B (b = 1, 2, ..., B); A computer that applies batch processing to the A × B raw signals X(a,b) to reduce spatial and spectral crosstalk between the A × B raw signals X(a,b), and outputs A × C processed signals Z(a,c) corresponding to the concentrations of the C type phosphor D(a,c) at the A emission points; A power supply configured to apply voltage to both ends of each of the aforementioned capillaries A; A multi-capillary electrophoresis apparatus equipped with the following features.

2. A multi-capillary electrophoresis apparatus according to claim 1, The optical system described above is Multiple dichroic mirrors configured to divide the fluorescence emitted from each of the A light-emitting points P(a) into B wavelength bands of light, thereby forming A × B wavelength-divided luminous beams; One or more area sensors or multiple line sensors in which A × B wavelength-resolved images of fluorescence emitted from the A light-emitting points P(a) are individually formed; Equipped with, On one or more area sensors or on multiple line sensors, A × B pixel sections corresponding to the B wavelength bands W(a, b) of the A light-emitting points are defined on the A × B wavelength division images. The combined signal of the A × B pixel sections corresponds to the A × B raw signals X(a, b), Multi-capillary electrophoresis apparatus.

3. A multi-capillary electrophoresis apparatus according to claim 2, Let b = b0 (where b0 is a constant) and c = c0 (where c0 is a constant). The spatial crosstalk between A processed signals Z(a, c0) is relatively smaller than the spatial crosstalk between A raw signals X(a, b0). Multi-capillary electrophoresis apparatus.

4. A multi-capillary electrophoresis apparatus according to claim 2 or 3, Assuming a = a0 (where a0 is a constant), the spectral crosstalk between C processed signals Z(a0, c) is relatively smaller than the spectral crosstalk between B raw signals X(a0, b). Multi-capillary electrophoresis apparatus.

5. A multi-capillary electrophoresis apparatus according to claim 2, A matrix X with (A × B) rows and 1 column, whose elements are the aforementioned A × B raw signals X(a,b), [Math 1] Let the above A × C processing signals Z(a, c) be the elements of a (A × C) matrix Z, [Math 2] Let Y be a matrix of (A×B) x (A×C) columns having (A×B)×(A×C) elements Y(a,b)(a,c) that satisfies the relationship X=Y×Z. [Math 3] In that case, The computer determines the generalized inverse of matrix Y to be Y, which has (A × C) rows and (A × B) columns. － Let's find Z = Y beforehand. － The system is configured to perform the multiplication by X operation as the batch process, or to perform an operation equivalent to the operation as the batch process. Multi-capillary electrophoresis apparatus.

6. A multi-capillary electrophoresis apparatus according to claim 2 or 5, In the electrophoretic analysis of the A samples S(a) mentioned above, the A × B raw signals X(a,b) change over time. The computer is configured to execute the batch processing at each time point and output a time series of the A × C processing signals Z(a, c). Multi-capillary electrophoresis apparatus.

7. A multi-capillary electrophoresis apparatus according to claim 2, 5, or 6, When the analysis is performed multiple times at different times, the computer is configured to perform the common batch processing for each electrophoretic analysis. Multi-capillary electrophoresis apparatus.

8. A multi-capillary electrophoresis apparatus according to claim 1, The optical system described above is A lenses that individually collimate the fluorescence emitted from the A light-emitting points P(a) to form A light beams; B dichroic mirrors configured to collectively divide each of the A luminous beams into B wavelength bands of light, thereby forming A × B wavelength-divided luminous beams; One or more area sensors or multiple line sensors, on which A × B wavelength-divided beams are projected, and A × B wavelength-divided images are individually formed; Equipped with, On one or more area sensors or on multiple line sensors, A × B pixel sections corresponding to the B wavelength bands W(a, b) of the A light-emitting points are defined on the A × B wavelength division images. The combined signal of the A × B pixel sections corresponds to the A × B raw signals X(a, b), Multi-capillary electrophoresis apparatus.

9. A multi-capillary electrophoresis apparatus according to claim 8, Let b = b0 (where b0 is a constant) and c = c0 (where c0 is a constant). The spatial crosstalk between A processed signals Z(a, c0) is relatively smaller than the spatial crosstalk between A raw signals X(a, b0). Multi-capillary electrophoresis apparatus.

10. A multi-capillary electrophoresis apparatus according to claim 8 or 9, Assuming a = a0 (where a0 is a constant), the spectral crosstalk between C processed signals Z(a0, c) is relatively smaller than the spectral crosstalk between B raw signals X(a0, b). Multi-capillary electrophoresis apparatus.

11. A multi-capillary electrophoresis apparatus according to claim 8, A matrix X with (A × B) rows and 1 column, whose elements are the aforementioned A × B raw signals X(a,b), [Math 4] Let the above A × C processing signals Z(a, c) be the elements of a (A × C) matrix Z, [Math 5] Let Y be a matrix of (A×B) x (A×C) columns having (A×B)×(A×C) elements Y(a,b)(a,c) that satisfies the relationship X=Y×Z. [Math 6] In that case, The computer determines the generalized inverse of matrix Y to be Y, which has (A × C) rows and (A × B) columns. － Let's find Z = Y beforehand. － The system is configured to perform the multiplication by X operation as the batch process, or to perform an operation equivalent to the operation as the batch process. Multi-capillary electrophoresis apparatus.

12. A multi-capillary electrophoresis apparatus according to claim 8 or 11, In the electrophoretic analysis of the A samples S(a) mentioned above, the A × B raw signals X(a,b) change over time. The computer is configured to execute the batch processing at each time point and output a time series of the A × C processing signals Z(a, c). Multi-capillary electrophoresis apparatus.

13. A multi-capillary electrophoresis apparatus according to claim 8, 11, or 12, When the analysis is performed multiple times at different times, the computer is configured to perform the common batch processing for each electrophoretic analysis. Multi-capillary electrophoresis apparatus.

Images