JP2009270931A - Observation device of single nucleic acid molecule - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a nucleic acid analyzer such as a gene diagnosing/analyzing device for observing the fluorescence occurring on a reaction device for performing analysis and performing the analysis of nucleic acid or protein on the basis of the brightness of a fluorescent bright point. <P>SOLUTION: The precision of observation data is enhanced by extracting only target single emission from the false signal or the like contained in measured data. Unnecessary data is erased by taking the correlation of a fluorescence spectrum obtained in a time axis direction with teacher data, and the false signal of a time lag, contamination or the like in the measured data is reduced. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a nucleic acid analyzer, for example, a genetic diagnosis / analyzer for observing fluorescence generated on a reaction device for performing analysis and analyzing nucleic acid, protein, etc. on the basis of the fluorescence bright spot luminance. The present invention relates to a nucleic acid analyzer.

Development of technology for diagnosing diseases by measuring trace substances in living bodies is advancing.
As its measurement method, for example, in order to analyze the change in the expression level of mRNA of a plurality of genes in a biological tissue or mutation of a gene, the measurement is performed using a microarray using cDNA or oligo fragment as a probe. There are techniques for nucleic acid hybridization. In addition, a technique for measuring the amount of protein molecules in living tissue or serum using a protein array using antibodies or peptides as probes has been developed.

  There are several methods for detecting the substance to be analyzed obtained by the above-described measurement according to the application in order to cope with recent technological development, and each has been devised. For example, in Patent Document 1, it is intended to provide a rapid, simple, and highly sensitive detection method for a substance to be analyzed, and a method in which quantitative results are ensured in an analysis result even when the amount of sample handled is very small. A method for detecting the substance to be analyzed is considered. The method is to measure the intensity of a signal derived from a labeled analyte to be analyzed over time while increasing the intensity of the signal, and to express the intensity of the signal as a function of time. Comparison is made with a predetermined value, and quantification is performed using the intensity of the signal when the value is reached.

  In Patent Document 2, there is a method of mapping a cDNA sequence on a genome sequence at high speed so that large-scale sequence data can be processed at high speed.

WO05 / 052583 Publication JP 2005-176730 A

  However, when the method described in Patent Document 2 is used, in order to satisfy the k-mer monotonic increase equation, the emission of protein-based dust other than the intended fluorescence emission by mapping, the emission by the separated fluorescent substance Unnecessary information such as will also increase. In this case, when observing the luminescence intensity of the target sample, luminescence derived from such unnecessary information is mixed, which causes a shift in luminescence time and instability of the luminescence intensity, thereby causing the measurement data A pseudo signal such as deviation or contamination is generated, and the analysis accuracy is lowered. That is, it is necessary to perform the analysis again due to a decrease in the analysis accuracy, which also causes an increase in analysis time.

In single molecule DNA testing, it is common to analyze a large number of samples at once, and the processing speed depends greatly on the analysis process.
As described above, particularly in single-molecule DNA testing, a process capable of distinguishing only a target signal is required.

  The present invention has been made in view of the above problems, and provides a nucleic acid analyzer capable of improving the degree of purification of observation data by extracting only a single target light emission from a pseudo signal or the like included in measurement data. It is to provide.

  In order to solve the above problem, the apparatus of the present invention provides means for extracting a target light emission intensity by using a time-series change of the light emission intensity.

  That is, the nucleic acid analyzer of the present invention comprises a light source for irradiating a fluorescently labeled reaction device with excitation light, a detector for capturing the emission intensity from the reaction device excited by the light source, and the detection A single nucleic acid molecule observation device comprising: an analysis unit for analyzing the luminescence intensity signal captured by the detector; and the analysis unit chronologically analyzes the luminescence intensity signal captured by the detector. A time-series data creating means for creating the obtained emission intensity measurement data, and a plurality of emission intensity teacher data relating to the emission intensity, and an emission intensity phase relationship representing a correlation between each emission intensity teacher data and the emission intensity measurement data A light emission intensity correlation coefficient calculating means for calculating the number, and the light emission intensity measurement data is converted into the plurality of light emission intensity teacher data based on the light emission intensity correlation coefficient. Identifying any one of data, having a specifying means.

  In this case, the time series data creation means calculates statistical information of the created emission intensity measurement data, and the emission intensity correlation coefficient calculation means includes a plurality of emission intensity teacher data related to the emission intensity statistical information. And calculating a light emission intensity statistical correlation coefficient representing a correlation between the teacher data regarding the statistical information of each light emission intensity and the statistical information of the light emission intensity measurement data, and the specifying means is based on the light emission intensity statistical correlation coefficient Then, the statistical information of the emission intensity measurement data may be specified as any one of the emission intensity teacher data related to the plurality of statistical information.

  Alternatively, the emission intensity correlation coefficient calculating means calculates emission intensity correlation coefficients representing the correlation between each emission intensity teacher data and the emission intensity measurement data in reverse order in time series, and the specifying means in time series The emission intensity measurement data may be specified as any one of the plurality of emission intensity teacher data based on the emission intensity correlation coefficient calculated in reverse order.

  Alternatively, when the luminescence intensity measurement data includes a plurality of different luminescence intensity signals, the specifying unit obtains a plurality of different luminescence intensity signals included in the luminescence intensity measurement data based on the luminescence intensity correlation coefficient. Each of the emission intensity teacher data may be specified.

  Alternatively, when the emission intensity measurement data includes unnecessary information, the time-series data creation means calculates statistical information of the created emission intensity measurement data, and the emission intensity correlation coefficient calculation means calculates unnecessary information statistics. A plurality of emission intensity teacher data relating to information, calculating emission intensity statistical correlation coefficients representing correlation between emission intensity teacher data relating to statistical information of each unnecessary information and statistical information of the emission intensity measurement data; Based on the emission intensity correlation coefficient, unnecessary information included in the emission intensity measurement data is specified as any one of emission intensity teacher data regarding statistical information of the plurality of unnecessary information, and the emission intensity including the unnecessary information Only the bright spot data may be extracted from the measurement data.

  Alternatively, the light emission intensity correlation coefficient calculating means includes a plurality of base sequence teacher data relating to a base sequence, calculating reliability information between each base sequence teacher data and the light emission intensity measurement data, and the specifying means includes the The emission intensity measurement data may be specified as any one of the plurality of base sequence teacher data based on reliability information.

  Or, further, statistical information of the emission intensity measurement data is calculated, unnecessary data is deleted from the emission intensity measurement data based on the statistical information, unnecessary information deletion means using time series statistical information, It may be characterized by having.

  Or, further, there is provided unnecessary information deleting means using wavelength information, which acquires wavelength information from the emission intensity data and deletes unnecessary data from the emission intensity measurement data based on the wavelength information. Also good.

  Or, further, calculate statistical information of the emission intensity measurement data, and delete unnecessary data from the emission intensity measurement data based on the statistical information, unnecessary information deletion means using time-series statistical information, and There may be provided unnecessary information deletion means using wavelength information, which acquires wavelength information from the emission intensity data and deletes unnecessary data from the emission intensity measurement data based on the wavelength information.

  Or, further, automatic adjustment means for storing the measurement conditions of the single nucleic acid molecule observation apparatus, reading the measurement conditions, and automatically adjusting the measurement conditions of the single nucleic acid molecule observation apparatus using the read measurement conditions It is good also as having.

  Alternatively, the automatic adjustment means may automatically adjust the measurement conditions by flowing a cleaning liquid through the flow path for cleaning.

  Alternatively, the automatic adjustment means may automatically adjust the measurement conditions by irradiating the flow path with ultraviolet rays and cleaning.

  Alternatively, the automatic adjustment means may automatically adjust the measurement conditions by changing the sample concentration or reagent concentration.

  Alternatively, the automatic adjustment means may automatically adjust the measurement conditions by changing at least one of exposure time, detector sensitivity, and excitation light intensity.

  Alternatively, it may further include warning instruction means for presenting at least one of a warning and an instruction to the user.

  Alternatively, the warning instruction means may execute at least one of an operation of a single nucleic acid molecule observation device, a specific sound, LED lighting / extinguishing, and display on a display device with a figure or a character. Good.

  The analysis method of the present invention can extract only a single light emission, and has an effect of reducing pseudo signals such as time lag and contamination in measurement data.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, this embodiment is merely an example for realizing the present invention, and does not limit the technical scope of the present invention.

<Device configuration>
As shown in FIG. 1, the single nucleic acid molecule observation apparatus of this embodiment includes an autosampler unit 101 for introducing a detection target into the apparatus, a pump unit 102 for creating a flow to be introduced, and a measurement target with a fluorescent label. 7 units of the reaction unit 105 to be fixed, the irradiation unit 106 for exciting the fluorescent label with the excitation light, the detection unit 107 for capturing the fluorescence emission intensity, the analysis unit 108 for comparing the outputs of the detection unit 107, and the warning instruction operation unit 110. Consists of blocks.

  The autosampler unit 101 includes a sample introduction unit 701, a sample introduction end 702, a sample moving unit 703, a sample container 704, a buffer container 705, a reagent container 706, and a cleaning container 707. The sample introduction unit 701 has a sample introduction end 702. The sample container 704 is a container for holding a solution containing a plurality of trace measurement objects. For example, the sample container 704 is a resin sheet on a sample plate having 24 rows and 16 columns of wells capable of holding a sample of several tens of μL. It is configured to place a septa and sandwich it between a holder and a clip. The measurement object is, for example, a large number of nucleic acids having an appropriate length (size) in which four kinds of nucleoside base molecules are fluorescently labeled. The septa is for preventing evaporation of the sample in the well, and has a normally closed through hole at a position corresponding to the well. When the sample is introduced, the sample introduction end 702 is connected to the sample through the through hole. Contact. Also, a protective film may be attached to the upper surface of the septa to prevent sample evaporation. The buffer container 705 is a container that holds a buffer solution in which the sample introduction end 702 is immersed. The reagent container 706 is a container for holding a reagent and a reaction solution for fixing a measurement target with a fluorescent label to the reaction device 203. The cleaning container 707 is a container that holds a cleaning liquid for cleaning the sample introduction end 702.

  The pump unit 102 includes a pump 801 for flowing sample components and a waste liquid container 802 for discarding the waste liquid. The waste liquid container 802 is a container that holds a used buffer medium.

  The reaction unit 105 includes a constant temperature bath 201, a reaction chamber 202, and a reaction device 203.

  The irradiation unit 106 includes a laser light source 501 serving as a light source for excitation light, a first condenser lens 502, an irradiation filter 503, a second condenser lens 504, a return light prevention slit 505, and a sample to irradiate the sample with excitation light. The total reflection prism 506 is provided. The return light prevention slit 505 is a slit for preventing the light returning to the laser light source 501 by reflection and refraction when the excitation light is made incident on the total reflection prism 506 and not degrading the analysis performance.

  The detection unit 107 includes a first camera lens 601, a detection filter 602, a diffraction grating 603, a second camera lens 604, a two-dimensional detector 605, and an XYZ stage 606. The two-dimensional detector 605 converts the intensity of the collected light into a digital value. Note that the diffraction grating 603 may use wavelength dispersion means in which prisms, dichroic mirrors, and the like are appropriately combined. Instead of the two-dimensional detector 605, a one-dimensional detector, a photomultiplier, a photodiode, and the like, and an optical mechanism may be appropriately combined. The XYZ stage 606 is a moving unit that moves the detection unit 107 so that the target reaction grid 204 can be photographed. Furthermore, it is used to adjust the focus when shooting. Note that the XYZ stage 606 may use a moving means that appropriately combines an XY stage that is a biaxial mechanism and a moving mechanism. The reaction unit 105 may have a moving means.

  The analysis unit 108 includes a computer 901 and a database 902.

  The warning instruction operation unit 110 includes a display device 1101, a light emitting device 1102, a sound generation device 1103, and an operation device 1104. For example, there are a liquid crystal display and a cathode ray tube display as a display device, an LED and a lamp as a light emitting device, a speaker and a buzzer as a sounding device, and a push button, a keyboard and a mouse as an operation device.

  The autosampler unit 101 and the reaction unit 105 are connected by a first tube 803, and the reaction unit 105 and the pump unit 102 are connected by a second tube 804.

Next, details of the reaction unit 105 will be described with reference to FIG.
2A is an enlarged view of the periphery of the reaction device 203 in the reaction unit 105 and the total reflection prism 506 in the irradiation unit 106. FIG. The reaction device 203 includes an irradiation unit 401 that is irradiated with excitation light, a connection end 402 for introducing a sample, and an observation window 403 for observing emission intensity. A total reflection prism 506 is in close contact with the irradiation unit 401 of the reaction device 203.

  FIG. 2B is a view of the reaction device 203 as viewed from above. As shown in FIG. 2B, the reaction device 203 is composed of a plurality of reaction grids 204 arranged in a row, and a flow path 206 is connected to the reaction grids 204 in the same row. This is for analyzing a plurality of sample components simultaneously. The reaction device 203 employs a detachable exchange member in case the quality deteriorates due to a predetermined number of analyzes or the separation performance decreases. Further, even when the measurement technique is changed and the number of reaction grids needs to be changed, the reaction device 203 having a different number of reaction grids can be exchanged, and the number of reaction grids can be adjusted. In FIG. 2B, the reaction grids are arranged in rows, but they may not be arranged in rows, and the flow paths may not be connected to the reaction grids 204 in the same row. In FIG. 2B, 24 reaction grids 204 are included, but there may be one, four, eight, or the like.

  FIG. 2C is an enlarged view of the reaction grid 204 of FIG. 2B. As shown in FIG. 2C, reaction sites 205 are arranged in a lattice pattern inside the reaction grid 204 every 1 μm. In FIG. 2C, the reaction sites 205 are arranged in a grid pattern every 1 μm, but may be arranged in a square grid pattern, a triangular grid pattern, or a grid pattern at any distance according to the analysis method. You may arrange in.

  FIG. 2D is a diagram in which the fluorescent spectrum analysis fluorescent label 103 is labeled on the measurement object 304.

  FIG. 2E is a view of the reaction grid 204 of FIG. 2B as viewed from the side. As shown in FIG. 2E, in the reaction grid 204, a non-specific adsorption preventing layer 302 is coated on a quartz glass substrate 301, gold fine particles 303 of several tens of nanometers are fixed, and DNA for adsorbing the measurement target 304 on the gold fine particles 303 is obtained. A probe 305 is attached. Each reaction site 205 is arranged on the quartz glass substrate 301 substantially parallel to the glass substrate. “Substantially” means that it is within a precision error. In FIG. 2E, it is composed of gold fine particles 303, but it may be metal particles that generate an enhanced field by localized surface plasmon resonance such as silver.

<Analysis operation>
The single nucleic acid molecule observation apparatus of the present embodiment applies the excitation light irradiated from the irradiation unit 106 to the fluorescent label of the nucleic acid to the reaction device 203 to which the nucleic acid with the fluorescent label is fixed, and the emission intensity to the detection unit 107. Are obtained in chronological order, and the correlation is obtained and compared with a database prepared in advance to identify the emission intensity. For example, a base sequence is determined from a sample component having a DNA-containing sample.

  The flow of analysis by the single diffusion molecule observation apparatus in the present invention will be described with reference to FIG. FIG. 4A is a flowchart relating to the analysis method of the single nucleic acid molecule observation apparatus of the present invention. Each operating subject is the computer 901 in FIG.

  In step 11, the single nucleic acid molecule observation apparatus is initialized. This operation is an initialization operation unique to the single nucleic acid molecule observation apparatus. For example, it is confirmed whether the power supply voltage is stable and each unit is connected.

  In step 12, a single nucleic acid molecule observation apparatus is set up. From here, the details of step 12 will be described with reference to FIG. 4B.

  In step 21, in order to determine the measurement conditions of the single nucleic acid molecule observation apparatus, the previous measurement conditions stored in a storage unit (not shown) in the analysis unit are read. If the emission intensity acquired under the measurement conditions read out in this step is lower than the threshold value necessary for measurement, or on the contrary, strong or absent, the measurement conditions are determined by the following steps 22 to 26. Automatically adjusted. Specifically, the exposure time is lengthened, the detector sensitivity is lowered, the light source intensity is increased, or the stage is moved. Alternatively, it is conceivable to change the sample concentration or reagent concentration when there is no emission intensity. In step 27 below, if the measurement conditions are not sufficiently satisfied, a warning or instruction is issued to the user. As a method of warning or instruction, it is conceivable to operate the single nucleic acid molecule observation apparatus, a specific sound generated by the sound generation device 1103, or turn on or off the light emitting device 1102, or display it on the display device 1101 with a figure or a symbol.

  In step 22, if necessary, for example, to improve the noise level (S / N ratio) of the luminescence intensity measured according to the measurement conditions in step 21, the channel is washed or irradiated with ultraviolet rays. To clean the flow path.

  In step 23, the stage position is adjusted to observe the reaction grid 204.

  In step 24, the measurement object 304 and the reagent are introduced from the autosampler unit 101 into the reaction grid 204. The sample moving unit 703 is a moving unit for transferring the sample container 704, the buffer container 705, the reagent container 706, and the cleaning container 707 to the sample introducing unit 701. First, the sample moving unit 703 places the buffer container 705 in the sample introduction unit 701 and immerses the sample introduction end 702 in the buffer solution held in the buffer container 705. Then, the reaction grid 204, the first tube 803, and the second tube 804 are filled with the buffer medium. As a result, a flow path including the autosampler unit 101, the reaction unit 105, and the pump unit 102 is formed. The reason why the sample introduction end 702 is immersed in the cleaning liquid in the cleaning container is to clean the sample introduction end 702 and avoid contamination. Next, a component including the measurement target is dispensed from the sample container 704 into the flow path, and the sample component is allowed to flow into the reaction device 203 together with the reaction solution (fluorescent dNTP, polymerase). At this time, the temperatures of the reaction grid 204 and the flow path 206 that affect the reaction rate of the sample components are controlled by the thermostatic bath. Then, air maintained at a constant temperature by a temperature control mechanism such as Peltier is circulated in the thermostatic chamber by a blower mechanism such as a fan, and the reaction grid 204 and the flow path 206 are maintained at a predetermined temperature. Further, during the standby of the apparatus, it is transported to the sample introduction unit 701 in the same manner as in the analysis, and the sample introduction end 702 is immersed in a buffer solution, and the buffer medium in the autosampler unit 101, the reaction unit 105, and the pump unit 102 Prevent drying. The waste container 802 receives the used buffer medium that is discharged when the buffer medium is filled.

  In step 25, the intensity of the laser light source 501 of the irradiation unit 106 is adjusted.

  In step 26, the detection sensitivity of the two-dimensional detector 605 is adjusted.

In step 27, a warning or instruction is given to the user. The display device 1101 displays an operation with a figure or a character or warns the user during operation. The light emitting device 1102 is turned on and off during operation, thereby indicating the state of the device and giving a warning. The sound generation device 1103 outputs an operation confirmation sound or a warning sound during operation. The user performs an operation at the time of operation using the operation device 1104.
The above is the setting of step 12.

  In step 13, measurement is performed with a single nucleic acid molecule observation apparatus. From here, the process of step 13 is demonstrated using FIG. 4C.

  In step 31, the irradiation unit 401 is irradiated with excitation light to excite the measurement object 304. The excitation light from the irradiation unit 106 is condensed by the first condenser lens 502, unnecessary wavelength components are cut by the irradiation filter 503, and the excitation light is condensed by the second condenser lens 504. The light collected by the second condenser lens 504 is applied to the total reflection prism 506. Since the observation window 403 of the reaction device 203 is installed on the opposite side of the irradiation unit 401, the excitation light totally reflects the outer surface of the irradiation unit 401 and propagates through the reaction device 203 to generate evanescent light. The evanescent light excites the sample in each reaction site 205 simultaneously. Due to the excitation light propagating in the reaction device 203, the sample in the reaction site 205 generates fluorescence in the range of several mm to several tens mm. Thus, in the present embodiment, a grid-like light emitting portion is formed.

  In step 32, the light emitted from the sample is measured by the two-dimensional detector 605 of the detection unit 107. When the sample component is irradiated with excitation light in the reaction device 203 in step 31, the phosphor labeled on the sample component emits light having a different wavelength for each sample component. This light is detected by the detection unit 107 for each reaction grid 204, and the light emission state is acquired as image information with respect to the time axis. FIG. 3 shows an example of an image obtained by the two-dimensional detector 605. In this image, reaction sites 205 are arranged in the vertical and horizontal directions, and the horizontal axis of one reaction site 205 is the wavelength dispersion direction. A black part indicates a light emitting point, and a white part indicates a non-light emitting point.

In the present embodiment, reaction grids 204 having a reaction site 205 interval of 1 μm are used. On the other hand, the wavelength distance between the peaks of the emission intensity by the phosphor is about 330 μm. That is, since the reaction site interval is smaller than the wavelength peak interval of the phosphor, the wavelength of the phosphor generated at each reaction site may be mixed, and it is necessary to specify which reaction site emitted the wavelength. Therefore, the wavelength indicating the center position of the reaction site 205 on the image of the two-dimensional detector 605 is obtained by obtaining emission intensity or spectral information from the image of each reaction site 205 that is emitting obtained in FIG. Obtainable. In this embodiment, by acquiring an image of a certain length in the X direction of the reaction site 205, not only can the position shift of the reaction site 205 be accurately obtained, but also the inclination angle of the reaction site 205 can be accurately determined. Obtainable. For example, when the reaction sites are arranged in a lattice pattern, the tilt and the position shift appear on the image, so that the position shift amount and the tilt can be obtained on the contrary.
The above is the flow until an image is obtained in units of reaction grids using the single nucleic acid molecule observation apparatus of the present invention.

  Next, a method for analyzing measurement data constituting an image obtained by the secondary detector and removing unnecessary information from the obtained measurement data will be described. In the present embodiment, a case where four color fluorescent dyes are used for one reaction grid 204 will be described. Each operation subject is the computer 901 in FIG.

  In step 33, a first unnecessary information removal process (unnecessary information removal using time-series statistical information) is performed. The purpose here is to identify the reaction sites that were not correctly formed in the manufacturing stage and the reaction sites where hybridization did not occur in the analysis stage, and apply general statistical techniques to perform computer processing on the location. Is to perform data processing quickly and efficiently. Here, unnecessary information is deleted based on statistical information with respect to the time axis of the non-light emitting point. The measurement data of the reaction site that was not accurately formed in the manufacturing stage can be recorded at the time of wavelength calibration, and can be removed from the measurement object when performing the measurement. Reaction sites in which hybridization did not occur in the analysis stage can be removed by performing statistical processing that takes a noise ratio and measurement data of a portion that is not a bright spot. A detailed description of the first unnecessary information removal method will be given later (FIG. 5).

  In step 34, a second unnecessary information removal process (specific) unique to the present invention is performed. The object here is to remove many light components other than those derived from the fluorescence emission of the sample components contained in the measurement data, for example, light emission of protein-based dust and light emission by separated fluorescent substances. In addition, even if it is derived from the sample component fluorescence emission, it may be considered that simultaneous emission by a plurality of fluorescent substances is included, and the emission intensity signal is specified one by one from the measurement data including the plurality of emission intensity signals. Here, by obtaining a correlation between the emission intensity signal acquired in the time axis direction and the teacher data, first, unnecessary information is specified and deleted, and then a plurality of emission intensity signals are specified. Details will be described later (FIGS. 6A and 6B). Teacher data is mainly created from measurement data and artificially created. As another extraction method, it is possible to use a general statistical method that obtains correlations such as signal average, variance, and S / N ratio. For example, there is a method in which the case where the degree of dispersion is the same as that of regular white noise is set as a non-light emitting point.

  In step 35, a third unnecessary information removal process (unnecessary information removal using wavelength information) is performed. The purpose here is to remove many light emission components other than those derived from the sample component fluorescence emission contained in the acquired measurement data, as in the second unnecessary information removal. However, here, the wavelength information is acquired from the measurement data and compared with the wavelength information of the phosphor added to the measurement object 304, thereby deleting the emission intensity signal derived from the wavelength other than the phosphor as unnecessary information. A detailed description of the third unnecessary information removing method will be given later (FIG. 7). This process is auxiliary, and is executed to remove those that could not be removed by the second unnecessary information removal method.

Note that all three unnecessary information removal methods from the first to the third may be executed, or one or two methods may be selectively executed.
The above is the measurement flow of step 13.

  In step 14, when continuous measurement is required, repeated measurement is performed. For example, in FIG. 2B, after measurement of one reaction grid 204, when it is desired to continuously measure other reaction grids, the process returns from this step to step 12, and the same process is repeated again.

In step 15, an analysis for determining a nucleic acid molecule is performed based on the obtained measurement result. Here, the base conversion process that determines the nucleic acid from the color of the fluorescent dye, the process of searching for the same thing to increase the accuracy of the nucleic acid determined by the base conversion process, and the base sequence determination process that compares with the database are performed. Is called.
The above is the main flow in performing the single nucleic acid molecule analysis of the present invention.

  Then, the unnecessary information removal method from the first to the third necessary for solving the problem of the present invention performed by the measurement in the above-described step 13 will be described below.

  A flowchart of the first unnecessary information removing step 33 is shown in FIG.

  In step 41, statistical information of the emission intensity signal with respect to the time axis direction is extracted. Here, the noise level is measured mainly by a general statistical method.

  In step 42, bright spot data and noise data with low emission intensity are removed. Specifically, the average value and variance of the statistical information obtained in step 41 are calculated, and those with an average value that is too low and those with a too large variance are removed.

  A flowchart of the second unnecessary information removing step 34 is shown in FIG. 6A.

  In step 51, as preparation before performing DP (dynamic programming) matching processing, normalization is performed after obtaining the moving average of the bright spot data. DP matching is a matching technique used in the field of speech recognition. Details will be described in step 54 described later.

  In step 52, the distance by DP is calculated. FIG. 6B shows a procedure for calculating the DP distance. In this embodiment, the section fixing method for fixing the start point and end point is used, but the section free method for not fixing the start point and end point may be used.

  In step 61, a local distance calculation is performed. Here, teacher data and measurement data are compared using a comparison method. The error is calculated by the absolute value error method with likelihood using signal intensities at time t1 when the teacher data is present and at time t2 when the measurement data is present. The likelihood represents an allowable amount of deviation (error) between signals. In general, since white noise is included in a signal, the likelihood may be the degree of likelihood of white noise or the degree of signal stability allowed as device performance. In the present embodiment, first, an absolute value error is calculated, and the likelihood is set to 0.1, and if it is less than that, the error value is 0, and if it is more than that, 0.1 is subtracted. This makes it possible to perform error comparison in consideration of signal noise. The t1 and t2 are shifted and compared in the entire section by the comparison method to obtain the local distance. In this embodiment, the likelihood is set to 0.1, but this value may be a fixed value such as 0.05 or 0.2, or a variable value that varies depending on the section or the light emission intensity.

  In step 62, a global distance calculation is performed. Here, the local distance is compared using the DP path. The distance that makes the local distance closest at time t1 with the teacher data and at time t2 with the measurement data is selected. By changing the DP path, it is possible to reduce time correspondence or to ignore specific noise. Finally, the normalized DP distance is obtained by dividing by the time of all sections of teacher data and measurement data. As a result, in the classification 54 based on the DP distance, which will be described later, it is possible to classify according to which pattern the one with the smallest DP distance applies.

  In step 63, backtrace shape comparison is performed. Here, in the case of measurement data that cannot be classified only by the DP distance, the shape of the backtrace, which is the DP distance calculation process, is compared with the teacher data 1001. In general, measurement data having a plurality of emission intensities is difficult to compare with teacher data. However, this method compares the measurement data at the end point where all light emission ends with the teacher data in reverse order in the time direction, not from the measurement data at the start point, which often includes a spectrum that emits light at other stages. . As a result, it is possible to improve the accuracy of specifying the emission intensity signal in the classification based on the DP distance (step 54) described later.

The procedure for performing backtrace shape comparison will be described below.
FIG. 11A shows an example of measurement data when backtrace shape comparison is performed. FIG. 11B shows an example of the result of backtrace shape comparison. In FIG. 11A, since the back trace of FIG. 11B continuously refers to the test signal, insertion occurs for the bright spot. This can be determined from the fact that the same value continues in the horizontal frame number direction in FIG. 11A. In order to detect a frame that has been quenched in multiple stages from the measurement data in FIG. 11A, a frame that has been quenched in reverse order in the time direction is compared with the teacher data. As a result of the comparison, a predicted spectrum signal corresponding to the identified teacher data is subtracted from the original measurement data. By repeating this for the number of times of quenching while going back to the light emission start frame, a plurality of fluorescent spectra are specified one by one in the teacher data.
The above is the flow of distance calculation by DP.

  In step 53, if there are a plurality of teacher data, the process is repeated.

  In step 54, the light emission patterns are classified based on the DP distance calculation performed in step 52. Classification may be made by comparing only DP distances, but can also be performed by using a discriminant analysis method or the like depending on the combination of DP distances. Although DP matching is used here, methods such as template matching, map matching, hidden Markov model matching, etc. may be used. The emission intensity signal is identified by matching the peak on the vertical axis of the obtained measurement data with the teacher data. In addition, even if it is derived from the sample component fluorescence emission, it can be considered that simultaneous emission by a plurality of fluorescent substances is included, and the plurality of emission intensity signals are specified one by one from the measurement data after unnecessary information is removed. The computer 901 calculates reliability information based on the measurement data acquired from the detection unit 107, and determines the base sequence by comparing with the database. Hereinafter, a method for classifying measurement data based on teacher data will be described.

  FIG. 8 shows a pattern in which a sample component emits light obtained as a result of the measurement. The horizontal axis of the graph of FIG. 8 is the number of frames, and the vertical axis is the intensity of the bright spot. One frame on the horizontal axis means data measured every 1/100 second. When measuring, it may be every 1/50 second or every 1/200 second in addition to every 1/100 second.

  There are 12 types of measurement data obtained from (1) to (12) in FIG. 8, where S is a portion with high emission intensity and B is a portion with low emission intensity. Specifically, (1) one-step extinction with S being stable, (2) S with extinction rising to the right, (3) extinction extinguishing with S and B, and (4) starting 2-4 frames Extinction, (5) extinction at the first frame, (6) extinction at the first stage at 5 frames or less after the start, second stage extinction after 5 frames, (7) quenching with unstable S, (8) twice Light emission (B is the same and S is long), (9) Double light emission (B is different and S is long), (10) Double light emission (B is the same and S is short), (11) Start-up Light emission delayed in time, (12) Strong light emission before quenching.

  If the obtained measurement data is highly reliable and consists of a single light emission (1), direct DP matching is used. A conceptual diagram of DP matching is shown in FIG. DP matching is composed of teacher data 1001, measurement data 1002, DP path, and comparison method 1004. The greatest feature is that even if the shape expands or contracts in the time direction, it can be specified without registering a pattern corresponding to it.

  The teacher data 1001 in FIG. 9 is a waveform serving as a reference for specifying the obtained measurement data, and gives time on the horizontal axis and normalized signal strength on the vertical axis. Here, the intensity at the point where the slope of the signal changes is called in order in the time axis direction, starting from frame 0. For example, in the case of DNA-A, T, G, and C of the teacher data in the upper right diagram in FIG. 9, the values are 0.0, 0.0, 1.0, 0.3, and 0.0.

  The measurement data 1002 is obtained by normalizing the emission intensity signal acquired in time series in the signal intensity direction. The horizontal axis represents time, and the vertical axis represents normalized signal strength.

  The DP path determines how much time correspondence between signals is allowed when performing DP matching. The DP path has a weight, and the weight usually increases according to a time correspondence shift. Fig. 10 schematically shows the internal calculation method used in the DP path. This is a method for allowing noise included in signals when the values are greatly different when the signals are compared. In the figure, white circles indicate coordinate points, lines indicate allowable error paths, and numbers indicate path weights (difficulty). In this embodiment, the standard type of FIG. 10A is used, but other than the standard type, the arrow type of FIG. 10B, the reference priority type of FIG. 10C, the test priority type of FIG. 10D, the time-adaptive free type of FIG. .

  The comparison method 1004 is a method for comparing signals when calculating the local distance by the DP path. The signal comparison is mainly performed by error calculation. Although the absolute value error method with likelihood is used in the present embodiment, a comparison method such as a difference error method, a square error method, or a Mahalanobis distance method may be used.

  In FIG. 9, the measurement data 1002 corresponds to a DNA-G pattern in which the DP distance 1005 calculated from each teacher data 1001 is the smallest value.

  However, when the obtained measurement data has the patterns (2) to (12) in FIG. 8 including a plurality of emission intensity signals, it is difficult to directly execute DP matching. A method for performing DP matching in these cases will be described.

  In the case of (2) and (3) in Fig. 8, since the sawtooth waveform is added to the pattern of (1), DP matching is possible if the sawtooth waveform is registered in the teacher data 1001. It is. The sawtooth waveform is approximated by the fact that the starting point rises from near 0.5 to 1.0 and extinguishes (2), and the starting point rises from near 0.5 to 1.0 and then extinguishes and rises again to near 0.5 (3). The

  In the cases of (4) and (5) in FIG. 8, this is possible by using short-term DP matching. DP matching using only the first 5 frames or less is called short-term DP matching. The light emission state in this section is compared. Three values of 0.0, 1.0, and 0.0 are used as teacher data.

  In the case of (6) in FIG. 8, DP matching is possible after the backtrace shape comparison described above.

  In the case of (7) in FIG. 8, it is possible by changing the likelihood of the comparison method 1004. In this embodiment, the likelihood is set to 0.1. However, when the likelihood is set to 0.2, when the DP distance becomes small, it means that S in the pattern of (1) is unstable. Thus, by adjusting the likelihood, (7) becomes DP matching as (1).

  In the cases of (8), (9), and (10) in FIG. 8, classification is possible by registering a plurality of rectangular waves in the teacher data 1001. (8) and (10) use 0.0, 1.0, 0.0, 1.0, and 0.0 as teacher data. (9) uses 0.0, 1.0, 0.5, 1.0, and 0.0 as teacher data. This allows DP matching. Alternatively, if each light emission of two times of light emission is stable, DP matching can be performed by separating the two times of light emission at a time.

  In the case of (11) in FIG. 8, DP matching is possible for a short period by registering a waveform with a weak signal intensity at the initial stage in the teacher data 1001. Three values of 0.0, 0.1, and 0.0 are used as teacher data.

  In the case of (12) in FIG. 8, DP matching is possible if a waveform with a medium signal intensity at the initial stage is registered in the teacher data 1001. Three values of 0.5, 0.1, and 0.0 are used as teacher data.

With these methods, DP matching is possible for the 12 types of measurement data in FIG. 8, and even with measurement data including a plurality of emission intensity signals, the emission intensity signals can be specified one by one.
The above is the flow of the second unnecessary information removal in step 34.

  Next, a flowchart of the third unnecessary information removing step 35 is shown in FIG.

  In step 71, a bright spot spectrum is extracted from the measurement data.

  In step 72, the phosphor or illuminant is specified by judging the bright spot spectrum of the measurement data, and when it is judged as unnecessary information, unnecessary information is removed. Hereinafter, a specific analysis method in this step will be described.

  First, a stable bright spot is selected from the measurement data. This is because when a bright spot spectrum is used, spectrum analysis should be performed using light emission in a section where the intensity of the bright spot is stable. There are three methods for selecting a stable bright spot: 1. a bright spot in the middle of the stable section, 2. a bright spot after a specific frame from the start of stability, and 3. binning of the bright spot in the stable section.

  1. The selection method by “bright spot in the middle of the stable section” is that if S is within the section error ± 0.05 from the average value of the section in the specific section, the section is assumed to be stable, and in the stable section In this method, fluorescence spectrum analysis is performed using the bright spot in the middle frame. This interval error may not be a value other than ± 0.05.

  2. The selection method based on “bright spot after specific frame from start of stable” is a method of searching for a stable section and then performing fluorescence spectrum analysis using, for example, a bright spot after 10 frames after the start of stable section. The delay frame may not be 10.

  3. The selection method based on “Binning of bright spots in the stable interval” is a method of searching for the stable interval, integrating all the spectra in the stable interval, and analyzing the spectrum. This method enables accurate fluorescence spectrum analysis even when the emission intensity is low.

  Next, since the emission from the fluorescent label reflects a waveform with a specific peak in the measurement data, the wavelength information of the peak is obtained, and the wavelength information and correlation coefficient of the reference fluorescence spectrum are obtained to obtain the DNA sample component The four color fluorescence spectra are identified from the detected light, and four types of DNA are identified. FIG. 12 shows a reference fluorescence spectrum acquired as wavelength calibration data. Fluorescence from the four color fluorescent dyes is wavelength-dispersed by the diffraction grating 603 and becomes a spectrum.

  In this manner, the fluorescence spectra of four colors can be separated from the detection light of the DNA sample component, and the four types of DNA can be accurately identified. Then, the correlation coefficient to the reference fluorescence spectrum can be calculated and used as the reliability information when searching the database.

  On the other hand, if any of the four types of dyes has a low correlation coefficient, it means that the bright spot is not derived from a fluorescent label, and is excluded as unnecessary information. An example of the fluorescence spectrum of unnecessary information is shown in FIG.

<Effect of this embodiment>
With the analysis means of the present invention described above, only single emission can be extracted from measurement data having a plurality of emission intensity signals, and there is an effect of reducing pseudo signals such as time lag and contamination in the measurement data. In addition, there is an effect of using the result to automate the operation of the apparatus and present information to the user.

It is a figure which shows the outline of a single nucleic acid molecule observation apparatus. It is a figure which shows the outline of the reaction device used with a single nucleic acid molecule observation apparatus. It is a figure which shows the outline of the reaction device used with a single nucleic acid molecule observation apparatus. It is a figure which shows the outline of the reaction device used with a single nucleic acid molecule observation apparatus. It is a figure which shows the outline of the reaction device used with a single nucleic acid molecule observation apparatus. It is a figure which shows the outline of the reaction device used with a single nucleic acid molecule observation apparatus. It is a figure which shows the luminescent spot image of 1 frame measured with a single nucleic acid molecule observation apparatus. It is a figure which shows the operation | movement procedure of the apparatus at the time of analyzing with a single nucleic acid molecule observation apparatus. It is a figure which shows the operation | movement procedure of the apparatus at the time of analyzing with a single nucleic acid molecule observation apparatus. It is a figure which shows the operation | movement procedure of the apparatus at the time of analyzing with a single nucleic acid molecule observation apparatus. It is a figure which shows the procedure of removing 1st unnecessary information with a single nucleic acid molecule observation apparatus. It is a figure which shows the procedure of removing 2nd unnecessary information with a single nucleic acid molecule observation apparatus. It is a figure which shows the procedure of removing 2nd unnecessary information with a single nucleic acid molecule observation apparatus. It is a figure which shows the procedure of removing 3rd unnecessary information with a single nucleic acid molecule observation apparatus. It is a figure which shows the characteristic waveform of the emitted light intensity with respect to the time direction measured with a single nucleic acid molecule observation apparatus. It is a figure which shows the characteristic waveform of the emitted light intensity with respect to the time direction measured with a single nucleic acid molecule observation apparatus. It is a figure which shows the characteristic waveform of the emitted light intensity with respect to the time direction measured with a single nucleic acid molecule observation apparatus. It is a figure which shows the characteristic waveform of the emitted light intensity with respect to the time direction measured with a single nucleic acid molecule observation apparatus. It is a figure which shows the concept of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows DP path | pass of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows DP path | pass of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows DP path | pass of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows DP path | pass of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows DP path | pass of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows the back trace of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows the back trace of DP matching used with a single nucleic acid molecule observation apparatus. It is a figure which shows the wavelength characteristic of the fluorescent dye used with a single nucleic acid molecule observation apparatus. It is a figure which shows the example of the fluorescence spectrum of the unnecessary information measured with a single nucleic acid molecule observation apparatus.

Explanation of symbols

11 Initialization steps
12 Setup steps
13 Measurement steps
14 Continuous measurement judgment step
15 Nucleic acid molecule analysis step
21 Previous measurement result reading step
22 Channel cleaning step
23 Stage movement step
24 Measurement target injection step
25 Light source intensity determination step
26 Detector sensitivity determination step
27 Warning instructions step
31 Excitation light irradiation / excitation step
32 Measuring steps with the detector
33 First unnecessary information removal step
34 Second unnecessary information removal step
35 Third unnecessary information removal step
41 Statistical data extraction step
42 Steps for removing statistical unnecessary information
51 Normalization step after moving average of bright spots
52 DP distance calculation step
53 Iteration step
54 Classification steps by DP distance
61 Local distance calculation step
62 Global distance calculation step
63 Backtrace shape comparison step
71 Bright spot spectrum extraction step
72 Spectral unwanted information removal step
101 Autosampler unit
102 Pump unit
103 Fluorescence spectrum analysis Fluorescent label
105 reaction units
106 Irradiation unit
107 detection unit
108 analysis units
201 temperature chamber
202 Reaction site center reaction chamber
203 reaction devices
204 Fluorescence intensity response grid
205 reaction sites
206 flow path
301 quartz glass substrate
302 Non-specific adsorption prevention layer
303 Gold fine particles
304 Single nucleic acid molecule observation device
305 DNA probe
401 Excitation light irradiation unit
402 Connection end
403 Observation window
501 Laser light source
502 First condenser lens
503 Irradiation filter
504 Second condenser lens
505 Return light prevention slit
506 Total reflection prism
601 First camera lens
602 detection filter
603 diffraction grating
604 Second camera lens
605 2D detector
606 XYZ stage
701 Sample introduction part
702 Sample introduction end
703 Sample moving part
704 Sample container
705 Buffer container
706 Reagent container
707 Cleaning container
801 pump
802 Waste liquid container
803 first tube
804 second tube
901 calculator
902 database
1001 Teacher data
1002 Measurement data
1003 DP pass
1004 Comparison method
1101 Display device
1102 Light emitting device
1103 Sound generator
1104 Operating device

Claims (16)

A light source for irradiating the fluorescence-labeled reaction device with excitation light, a detector for capturing emission intensity from the reaction device excited by the light source, and an emission intensity signal captured by the detector as data A single nucleic acid molecule observation apparatus comprising an analysis unit for analysis,
The analysis unit is
Creating luminescence intensity measurement data obtained in chronological order the luminescence intensity signals captured by the detector, time-series data creating means;
A plurality of emission intensity teacher data relating to the emission intensity, and an emission intensity correlation coefficient calculating means for calculating an emission intensity correlation coefficient representing a correlation between each emission intensity teacher data and the emission intensity measurement data;
Identifying means for identifying the emission intensity measurement data as any one of the plurality of emission intensity teacher data based on the emission intensity correlation coefficient;
An apparatus for observing a single nucleic acid molecule, comprising: In the single nucleic acid molecule observation apparatus according to claim 1,
The time series data creation means calculates statistical information of the created emission intensity measurement data,
The emission intensity correlation coefficient calculating means includes a plurality of emission intensity teacher data relating to the statistical information of the emission intensity, and represents a correlation between the teacher data relating to the statistical information of each emission intensity and the statistical information of the emission intensity measurement data. Calculate the emission intensity statistical correlation coefficient,
The specifying means specifies, based on the emission intensity statistical correlation coefficient, statistical information of the emission intensity measurement data as any one of emission intensity teacher data related to the plurality of statistical information. One nucleic acid molecule observation device. In the single nucleic acid molecule observation apparatus according to claim 1,
The emission intensity correlation coefficient calculating means calculates the emission intensity correlation coefficient representing the correlation between each emission intensity teacher data and the emission intensity measurement data in reverse order in time series,
The specifying means specifies the emission intensity measurement data as one of the plurality of emission intensity teacher data based on the emission intensity correlation coefficient calculated in reverse order in time series. One nucleic acid molecule observation device. In the single nucleic acid molecule observation apparatus according to claim 1,
When the emission intensity measurement data includes a plurality of different emission intensity signals,
The specifying means specifies each of a plurality of different emission intensity signals included in the emission intensity measurement data in each of the emission intensity teacher data based on the emission intensity correlation coefficient. Nucleic acid molecule observation device. In the single nucleic acid molecule observation apparatus according to claim 1,
When the emission intensity measurement data includes unnecessary information,
The time series data creation means calculates statistical information of the created emission intensity measurement data,
The light emission intensity correlation coefficient calculating means includes a plurality of light emission intensity teacher data related to statistical information of unnecessary information, and a light emission representing a correlation between the light emission intensity teacher data related to statistical information of each unnecessary information and the statistical information of the light emission intensity measurement data. Calculate the strength statistical correlation coefficient,
The specifying means specifies unnecessary information included in the emission intensity measurement data as any one of emission intensity teacher data related to statistical information of the plurality of unnecessary information based on the emission intensity correlation coefficient, and the unnecessary A single nucleic acid molecule observation apparatus, wherein only bright spot data is extracted from emission intensity measurement data including information. In the single nucleic acid molecule observation apparatus according to claim 1,
The emission intensity correlation coefficient calculating means includes a plurality of base sequence teacher data relating to a base sequence, and calculates reliability information between each base sequence teacher data and the emission intensity measurement data,
The single nucleic acid molecule observation apparatus, wherein the specifying means specifies the luminescence intensity measurement data as any one of the plurality of base sequence teacher data based on the reliability information. The single nucleic acid molecule observation device according to claim 1, further comprising:
Statistical information of the emission intensity measurement data is calculated, unnecessary data is deleted from the emission intensity measurement data based on the statistical information, unnecessary information deletion means using time series statistical information,
An apparatus for observing a single nucleic acid molecule, comprising: The single nucleic acid molecule observation device according to claim 1, further comprising:
Wavelength information is acquired from the emission intensity data, unnecessary data is deleted from the emission intensity measurement data based on the wavelength information, unnecessary information deletion means using wavelength information,
An apparatus for observing a single nucleic acid molecule, comprising: The single nucleic acid molecule observation device according to claim 1, further comprising:
Statistical information of the emission intensity measurement data is calculated, unnecessary data is deleted from the emission intensity measurement data based on the statistical information, unnecessary information deletion means using time series statistical information,
Wavelength information is acquired from the emission intensity data, unnecessary data is deleted from the emission intensity measurement data based on the wavelength information, unnecessary information deletion means using wavelength information,
An apparatus for observing a single nucleic acid molecule, comprising: The single nucleic acid molecule observation device according to claim 1, further comprising:
Automatic adjustment means for storing measurement conditions of a single nucleic acid molecule observation apparatus, reading the measurement conditions, and automatically adjusting the measurement conditions of the single nucleic acid molecule observation apparatus using the read measurement conditions;
An apparatus for observing a single nucleic acid molecule, comprising: In the single nucleic acid molecule observation apparatus according to claim 10,
The single nucleic acid molecule observation apparatus, wherein the automatic adjustment means automatically adjusts the measurement conditions by flowing a cleaning liquid through the flow path for cleaning. In the single nucleic acid molecule observation apparatus according to claim 10,
The single nucleic acid molecule observation apparatus, wherein the automatic adjustment means automatically adjusts the measurement conditions by irradiating the flow path with ultraviolet rays and cleaning. In the single nucleic acid molecule observation apparatus according to claim 10,
The single nucleic acid molecule observation apparatus, wherein the automatic adjustment means automatically adjusts measurement conditions by changing a sample concentration or a reagent concentration. In the single nucleic acid molecule observation apparatus according to claim 10,
The single nucleic acid molecule observation apparatus, wherein the automatic adjustment means automatically adjusts measurement conditions by changing at least one of exposure time, detector sensitivity, and excitation light intensity. The single nucleic acid molecule observation device according to claim 1, further comprising:
A warning instruction means for presenting at least one of a warning and an instruction to the user;
An apparatus for observing a single nucleic acid molecule, comprising: In the single nucleic acid molecule observation apparatus according to claim 15,
The warning instruction means executes at least one of an operation of a single nucleic acid molecule observation device, a specific sound, an LED on / off, and a display or a figure on a display device. Nucleic acid molecule observation device.

Images