Digital PCR system, target nucleic acid detection method, and target nucleic acid detection storage medium

Abstract

Provided is a digital PCR system or the like capable of measuring the concentration of a target nucleic acid with high accuracy even when an abnormality occurs in a well into which the target nucleic acid is introduced. A digital PCR system (200) according to the present invention is a system for detecting a target nucleic acid in a sample solution (108) by digital PCR, and comprises: an imaging device (fluorescence detector (202)) that captures an image including a plurality of wells (105) into which the sample solution (108) has been introduced; and a control unit (data analysis unit (203)) that recognizes the positions of the plurality of wells (105) in the image, recognizes the presence / absence of abnormalities included in each well (105) and the position of the abnormalities in the wells (105), calculates a representative value of the fluorescence intensity of each of the wells (105), corrects the representative value of the fluorescence intensity of the wells (105) having the abnormalities, and determines the presence or absence of the target nucleic acid in the plurality of wells (105).

Description

Digital PCR system, target nucleic acid detection method, and target nucleic acid detection storage medium

[0001] The present invention relates to a digital PCR system, a method for detecting a nucleic acid of interest, and a storage medium for detecting a nucleic acid of interest.

[0002] Digital PCR (Polymerase Chain Reaction) is a technology for detecting deoxyribonucleic acid (DNA) fragments, genes, etc. (hereinafter, these may be collectively referred to as "target nucleic acids") with high sensitivity. Digital PCR (hereinafter, sometimes referred to as "dPCR") can detect low-frequency gene mutations compared to conventional real-time quantitative PCR.

[0003] Digital PCR involves dividing a sample solution into tiny droplets, then performing PCR within the divided droplets to amplify and measure the target nucleic acid. Each tiny droplet contains one or more target nucleic acids, and some do not. Droplets in which the target nucleic acid has been amplified after PCR are counted as positive, while droplets that have not been amplified and do not contain the target nucleic acid are counted as negative. In digital PCR, the average copy number of the target nucleic acid per droplet, i.e., the concentration of the target nucleic acid, is calculated from the counting results using Poisson correction.

[0004] Patent Document 1 discloses a configuration for reducing quantification errors caused by optical artifacts in dPCR by identifying a reaction region as invalid when the optical signal is unevenly distributed in each reaction region and excluding the invalid reaction region from the calculation of the amount or concentration of the nucleic acid of interest. Patent Document 2 also discloses a method for quantifying the reaction volume in each reaction region to reduce quantification errors caused by reaction volume deviation in dPCR. Patent Document 3 discloses a highly robust sample solution separation technology in a sample fractionation device for dPCR by using a photocurable resin as a separation liquid. Note that, in Patent Document 3, "high robustness" means that the sample solution in the fractions (wells) does not leak out.

[0005] Japanese Patent No. 6985950 Japanese Patent No. 6936674 International Publication No. 2024 / 121959

[0006] In digital PCR, a sample solution containing the target nucleic acid is divided into multiple wells, and PCR is performed on each well to identify the type of DNA present in each well.Digital PCR is characterized by its ability to measure the target nucleic acid with high sensitivity by performing PCR after dividing the solution into individual wells.

[0007] One application of digital PCR is liquid biopsy, which detects minute amounts of cell-free DNA (cfDNA) present in blood. Cell-free DNA is DNA fragments derived from cells destroyed by the immune system or killed by apoptosis. Cell-free DNA may contain circulating tumor DNA (ctDNA), which is tumor-derived DNA. In the medical field, measuring the type and rate of ctDNA mutations is expected to be useful in cancer diagnosis, treatment selection, and treatment efficacy monitoring. Diagnostic applications require highly accurate measurement of target nucleic acids.

[0008] In digital PCR, the number of positive and negative droplets is measured for each fluorescent dye. Because each well contains zero, one, or multiple copies of the target nucleic acid, the proportion of positive droplets is fitted to a Poisson distribution to calculate the concentration of the target nucleic acid molecule (copies / μL). Therefore, to accurately determine the concentration of the target nucleic acid, it is necessary to accurately measure the proportion of positive droplets.

[0009] However, during research and development of digital PCR, the inventors discovered that abnormalities such as tiny bubbles can occur in the wells into which the target nucleic acid is introduced, which can affect the measurement results. In such cases, the fluorescence intensity of negative wells may be overestimated, resulting in a false positive determination. Furthermore, the fluorescence intensity of positive wells may be underestimated, resulting in a false negative determination. This can make it difficult to accurately measure the fluorescence intensity, making it difficult to accurately determine the concentration of the target nucleic acid.

[0010] The present invention has been made in view of the above. One object of the present invention is to provide a digital PCR system, a method for detecting a target nucleic acid, and a storage medium for detecting a target nucleic acid, which are capable of measuring the concentration of the target nucleic acid with high accuracy even if an abnormality occurs in the well into which the target nucleic acid is introduced. The above and other objects and novel features of the present invention will become apparent from the description of this specification and the accompanying drawings.

[0011] A brief outline of a representative embodiment of the present disclosure is as follows: A digital PCR system according to a representative embodiment of the present disclosure is a system for detecting a target nucleic acid in a sample solution by digital PCR, and includes an imaging device that captures an image including a plurality of wells into which the sample solution has been introduced, and a control unit that recognizes the positions of the plurality of wells in the image, recognizes the presence or absence of an abnormality contained in each of the wells and the position of the abnormality within the well, calculates a representative value of the fluorescence intensity of each of the wells, corrects the representative value of the fluorescence intensity of wells having the abnormality, and determines the presence or absence of the target nucleic acid in the plurality of wells.

[0012] To briefly explain the effects obtained by representative embodiments of the inventions disclosed in this application, it is possible to provide a digital PCR system, a target nucleic acid detection method, and a target nucleic acid detection storage medium that can measure the concentration of a target nucleic acid with high accuracy even if an abnormality occurs in the well into which the target nucleic acid is introduced.

[0013] 8B is a conceptual diagram of a digital PCR device 101 used in a digital PCR system 200 according to the first embodiment. FIG. 8C is a configuration diagram of the digital PCR system 200 according to the first embodiment. FIG. 8D is a diagram showing an example of an image captured at high resolution of the digital PCR device 101 that has performed digital PCR using the digital PCR system 200 according to the first embodiment. FIG. 8E is a flowchart illustrating an example of a procedure for detecting the position of a well 105 from an image. FIG. 8F is a flowchart illustrating an example of a procedure for recognizing the position of a well 105. FIG. 8G is a flowchart illustrating an example of a procedure for recognizing an air bubble 110. FIG. 8H is a conceptual diagram illustrating an example of the configuration of a network. FIG. 8H is a conceptual diagram illustrating a network trained to obtain an output image (correct image) that is training data from an input image that is training data. FIG. 8I is a conceptual diagram illustrating inputting an input image containing an abnormal portion such as an air bubble 110 into a trained network, and obtaining an output image. FIG. 8I is a diagram comparing the output image (well image) and input image (well image) in FIG. 8B with an SSIM image. FIG. 8I is a diagram showing an example of an input image (well image) to be input to an SSIM autoencoder. FIG. 8I is an SSIM image obtained from the input image and output image of the autoencoder. 1 is a diagram showing pixels determined to be abnormal pixels when an SSIM image is binarized. FIG. 2 is an image in which pixels determined to be abnormal pixels are colored and overwritten on the original well image. FIG. 3 is a histogram showing the average intensity (average fluorescence intensity) of wells 105 determined to contain bubbles and wells 105 determined to contain no bubbles. FIG. 4 is a histogram showing the average intensity (average fluorescence intensity) of wells 105 determined to contain bubbles before and after correction for the influence of bubbles 110. FIG. 5 is a flowchart explaining the processing content of a digital PCR system 200 according to a second embodiment. FIG. 6 is an explanatory diagram explaining an example of the temperature dependency of fluorescence intensity. FIG. 7 is an explanatory diagram explaining an example of a differential curve obtained by differentiating a melting curve. FIG. 8 is a flowchart explaining the content of a target nucleic acid detection method according to the present embodiment. FIG. 9 is a configuration diagram explaining the configuration of a target nucleic acid detection storage medium 1400 according to the present embodiment and the configuration of a computer 10 constituting a data analysis unit.

[0014] A digital PCR system, a target nucleic acid detection method, and a target nucleic acid detection storage medium according to one embodiment of the present invention will be described below with reference to the accompanying drawings. For convenience, the following embodiments will be divided into multiple sections or embodiments. However, unless otherwise specified, they are not unrelated to each other, and one is related to the other in terms of partial or complete modifications, details, supplementary explanations, etc. Furthermore, in the following embodiments, when the number of elements (including numbers, numerical values, amounts, ranges, etc.) is mentioned, it is not limited to that specific number, and may be greater than or less than the specific number, unless otherwise specified or when it is clearly limited to a specific number in principle.

[0015] Furthermore, in the following embodiments, the components (including element steps, etc.) are not necessarily essential unless otherwise specified or considered to be clearly essential in principle. Similarly, in the following embodiments, when referring to the shape, positional relationship, etc. of components, etc., it is intended to include those that are substantially similar or similar to the shape, etc., unless otherwise specified or considered to be clearly not essential in principle. The same applies to the above-mentioned numerical values ​​and ranges. Furthermore, the description of the following embodiments is merely a typical example and does not limit the scope of the claims or application examples in any sense.

[0016] In addition, in all the drawings for explaining the embodiments, when there are multiple components having the same or similar functions, they may be described by adding different subscripts to the same reference numeral. Furthermore, when it is not necessary to distinguish between these multiple components, the subscripts may be omitted. Components having the same or similar functions may be given the same reference numeral, and duplicate descriptions may be omitted.

[0017] <Digital PCR System> (First Embodiment) FIG. 1 is a conceptual diagram of a digital PCR device 101 used in a digital PCR system 200 according to the first embodiment. FIG. 2 is a configuration diagram of the digital PCR system 200 according to the first embodiment. As described above, the digital PCR device 101 shown in FIG. 1 is used in the digital PCR system 200 shown in FIG. 2. As shown in FIG. 1, the digital PCR device 101 is composed of an inlet 102, an outlet 107, a flow channel 104, and wells (microchambers) 105a, 105b, 105c, etc. Although not particularly limited, each digital PCR device 101 has 10,000 to tens of thousands of wells 105. The volume of one well 105 is, for example, approximately 1 nL. The total volume of the sample solution 108 contained in the multiple wells 105 is, for example, 10 μL to several tens of μL.

[0018] 2, the digital PCR system 200 includes a temperature controller 201, a fluorescence detector 202 (imaging device), and a data analysis unit 203 (control unit). The digital PCR device 101 is placed on the temperature controller 201.

[0019] The fluorescence detector 202 is composed of a light source 204, lenses 205, 208, 210, an excitation filter 206, a dichroic mirror 207, a fluorescence filter 209, and a light-receiving sensor 211. The fluorescence detector 202 captures an image including a plurality of wells 105 into which sample solutions 108 have been introduced. In the fluorescence detector 202, excitation light from the light source 204 passes through the lens 205, the excitation filter 206, and the dichroic mirror 207 and irradiates each well 105. The excitation light then excites a fluorescent substance contained in the sample solution 108 in the well 105, and the emitted fluorescence is detected by the light-receiving sensor 211 through the lens 208, the dichroic mirror 207, the fluorescence filter 209, and the lens 210. Fluorescence data detected by the fluorescence detector 202 is sent to the data analysis unit 203. The data analysis unit 203 may be a computer 10 (FIG. 14) connected to the digital PCR system 200 via a communication circuit 20 (FIG. 14) and a communication interface (I / F) 17 (FIG. 14).

[0020] The data analysis unit 203 recognizes the positions of the multiple wells 105 in the captured image. The data analysis unit 203 also recognizes the presence or absence of an abnormality in each well 105 and the position of the abnormality within the well 105. The data analysis unit 203 calculates a representative value of the fluorescence intensity of each well 105. Examples of the representative value of the fluorescence intensity include the average value and median value of the fluorescence intensity. Furthermore, the data analysis unit 203 corrects the representative value of the fluorescence intensity of the wells 105 that have the abnormality. The representative value can be corrected, for example, by removing the abnormal portion (abnormal pixel) of the well 105. In other words, the corrected representative value becomes the average value or median value of the fluorescence intensity of the well 105 after removing the abnormal portion. The data analysis unit 203 also determines the presence or absence of the target nucleic acid in the multiple wells 105. The presence or absence of the target nucleic acid can be determined by whether or not the fluorescence intensity is measured or whether or not the fluorescence intensity is higher than a predetermined threshold.

[0021] The sample solution 108 contains target nucleic acid, PCR amplification enzyme, buffer, primers, dNTPs (dATP, dTTP, dGTP, dCTP), and probes. The sample solution 108 is introduced into wells 105 in the digital PCR device 101. Specifically, the sample solution 108 is introduced into the wells 105 from the inlet 102 through the flow path 104. A separating liquid 109 such as oil or photocurable resin is then introduced into the flow path 104 to push out the sample solution 108, and the separating liquid 109 divides each well 105. In this way, the sample solution 108 is separated and introduced into the minute wells 105. The digital PCR device 101 with the sample solution 108 introduced therein undergoes thermal cycling by the temperature controller 201, and the target nucleic acid is amplified by PCR. The amplified target nucleic acid is subjected to fluorescence measurement using the fluorescence detector 202 as described above. Analysis is then performed by the data analysis unit 203 based on the measured data. By measuring the fluorescence of the target nucleic acid amplified by PCR, it is determined whether or not each well 105 contains the target nucleic acid. Fluorescence intensity can generally be used for this determination. If the measured fluorescence intensity is higher than a predetermined threshold, the well 105 can be determined as a positive well containing the target nucleic acid. On the other hand, if the measured fluorescence intensity is lower than the predetermined threshold, the well 105 can be determined as a negative well not containing the target nucleic acid.

[0022] FIG. 3 shows an example of a high-resolution image of a digital PCR device 101 that has undergone digital PCR using the digital PCR system 200 according to the first embodiment. As shown in FIG. 3 , digital PCR amplification products are observed in some wells 105, making them positive (bright) wells 105. On the other hand, no digital PCR amplification products are observed in some wells 105, making them negative (dark) wells 105. Furthermore, some wells 105 exhibited abnormalities, specifically air bubbles 110. Digital PCR involves thermal cycling, which changes the temperature of the sample solution 108 to cause the reaction. Therefore, when the temperature is increased, the sample solution 108 may evaporate, generating air bubbles 110. Furthermore, if the digital PCR device 101 is made of resin, the resin's water-absorbing properties may cause the sample solution 108 to absorb water, reducing the volume of the sample solution 108 relative to the volume of the wells 105 and generating air bubbles 110. When each well 105 of the digital PCR device 101 is sealed, the internal pressure decreases by the amount of sample solution 108 reduced, and this is observed as bubbles 110. One method for sealing each well 105 is to seal the well 105 using a solid. Specifically, a seal is used. Alternatively, when a photocurable resin is used as the separation liquid 109, the well 105 is sealed by hardening it. In these cases, bubbles 110 may be generated if the internal pressure of the well 105 decreases for some reason.

[0023] The photocurable resin may be, for example, a hydrophobic, liquid photocurable resin having a molecular skeleton of a cycloalkane, a normal alkane with six or more carbon atoms, or an isoalkane. Such a photocurable resin is less likely to inhibit the PCR reaction. The photocurable resin preferably has a viscosity of, for example, 500 mPa·s or less. This allows the photocurable resin to be easily introduced into the inlet 102, flow channel 104, etc. of the digital PCR device 101. The photocurable resin is preferably used in conjunction with any photoinitiator that initiates the photopolymerization reaction. Photocuring may be performed, for example, by irradiating with ultraviolet light having a wavelength of 250 to 600 nm. In this embodiment, the photocurable resin is introduced into the flow channel 104 in a liquid state and then cured by exposure to ultraviolet light. This seals the well 105. For details on the photocurable resin and its use, see International Publication No. 2024 / 121959. Examples of oils that may be used include silicone-based oils and Fluorinert-based oils.

[0024] If bubbles 110 are formed as described above, excitation light is scattered at the interface of the bubbles 110 when fluorescence measurement is performed. Furthermore, the bubbles 110 do not contain sample solution 108, so the fluorescence intensity is weakened. Therefore, the bubbles 110 cause variations in the fluorescence intensity emitted by the wells 105, making it difficult to accurately determine whether the result is positive or negative.

[0025] In the digital PCR system 200 of this embodiment, by recognizing tiny bubbles 110 in the well 105 and correcting the representative value of the fluorescence intensity of that well 105, it is possible to accurately calculate the fluorescence intensity and the concentration of the target nucleic acid even if bubbles 110 are generated.

[0026] Next, with reference to FIGS. 4 to 6, a procedure for accurately calculating the fluorescence intensity and the concentration of the target nucleic acid when a microbubble 110 is present in the well 105 will be described. FIG. 4 is a flowchart illustrating an example of a procedure for detecting the position of the well 105 from an image. As shown in FIG. 4, first, the data analysis unit 203 (control unit) removes rotational components from the captured image (S401). That is, the image is rotated so that the edges of the well 105 are aligned vertically and horizontally. This removes the rotational components, as described above. Next, the data analysis unit 203 removes distortion components (S402). That is, the image is corrected so that the edges of the well 105 are linear. Next, the data analysis unit 203 removes background (S403). This allows for the removal of influences other than those of the well 105, such as noise. Next, the data analysis unit 203 detects the position of the well 105 after these procedures have been performed (S404).

[0027] FIG. 5 is a flowchart illustrating an example of a procedure for recognizing the positions of wells 105. As shown in FIG. 5, the data analysis unit 203 (control unit) projects the image in the first axis direction and creates a projection vector (S501). Here, creating a projection vector means aligning (aligning) the images of the wells 105 in each column and row (images including the edges of the wells 105) vertically if they are aligned vertically, or horizontally if they are aligned horizontally. This can be easily understood by comparing, for example, FIG. 3 (before alignment) with FIG. 9A (after alignment), which will be described later. In this way, areas where wells 105 are present appear brighter, and areas where no wells 105 are present appear darker.

[0028] Next, the data analysis unit 203 identifies the frequency of the fundamental wave of the vector (S502). When a projection vector is created as described above, in this embodiment, the rectangular well 105 appears with a certain frequency. This results in a signal that is somewhat periodic, and this frequency is identified. Note that if the edges of the well 105 are not aligned vertically and horizontally, a clear periodicity cannot be obtained. Therefore, the image of the well 105 is rotated as described with reference to FIG. 4, and a projection vector is created as described above, and its period is determined.

[0029] Next, the data analysis unit 203 detects a local peak of the fundamental wave (S503). The local peak corresponds to a dark area where there are no wells 105. This allows the recognition of, for example, a row of wells 105. Next, the data analysis unit 203 searches before and after the local peak to detect the end points of the wells 105 (S504). This allows the data analysis unit 203 to detect the positions of the wells 105 in the first axial direction. Next, the data analysis unit 203 similarly detects the positions of the wells 105 in the second axial direction (S505). That is, the data analysis unit 203 similarly performs S501 to S504 for the second axial direction to detect the positions of the wells 105 in the second axial direction. Next, the data analysis unit 203 aligns the positions of the wells 105 in the first axial direction and the second axial direction to recognize the positions of the wells 105 (S506).

[0030] In this embodiment, an image of the position of the recognized well 105 is used to recognize an abnormality in the well image. Note that, for example, an air bubble 110 is an example of the abnormality. Below, an abnormality will be described using the air bubble 110 as an example. Furthermore, the well image is an image of the position of the recognized well 105 described above, that is, an image in which each well 105 is aligned in the column direction and the row direction (see, for example, FIG. 9A). An example of a method for recognizing an air bubble 110 is as follows. FIG. 6 is a flowchart explaining an example of the procedure for recognizing an air bubble 110. FIG. 7 is a conceptual diagram explaining an example of a network configuration.

[0031] As shown in FIG. 6 , in this embodiment, first, as preparation, a network is trained using training data (S601). That is, in this embodiment, the presence or absence of bubbles 110 and their positions are recognized by machine learning, in which the data analysis unit 203 inputs images of normal wells 105 and trains them, thereby recognizing wells 105 containing bubbles 110 and their positions. Note that, for example, an SSIM (Structural Similarity Index Measure) autoencoder network as shown in FIG. 7 may be used as the network. As shown in FIG. 7 , the data analysis unit 203 first reduces the dimension of the input image using an encoder. Specifically, image features are extracted using a convolutional neural network (CNN) or the like that uses, for example, variance μ (standard deviation), population mean σ, random sampling, or the like as feature quantities (latent variables). The decoder restores the latent representation obtained by the encoder to the original image. Training is performed to maximize structural similarity with the original image. In this embodiment, the SSIM index is used as a loss function, and learning is performed taking into account structural information about the image and correlations between images. Images without bubbles 110 (referred to as normal images) are used as the learning data. By learning using normal images as input images for the learning data, an output image (learning data) with no bubbles 110 in the wells 105 and uniform fluorescence intensity becomes the correct image ( FIG. 8A ). Note that FIG. 8A is a conceptual diagram illustrating a network trained to obtain an output image (correct image) as learning data from an input image as learning data.

[0032] Next, in this embodiment, a well image for detection is input to the trained network, and an output image is obtained (S602). In this way, normal components are extracted as the output image (FIG. 8B). That is, an image of a well in which abnormalities such as the air bubble 110 have been removed and is similar to the surrounding area is output. Note that FIG. 8B is a conceptual diagram illustrating how an input image containing an abnormality such as the air bubble 110 is input to the trained network, and an output image is obtained. Next, in this embodiment, the difference between the input image and the output image of the captured well image is evaluated using SSIM (S603). Here, for example, the SSIM of the image near each pixel is calculated for each pixel. Pixels with a large difference have low similarity and therefore a small SSIM. Pixels with low similarity (small SSIM) are then recognized as abnormal (air bubble 110) (S604). Therefore, an SSIM image can be obtained from the input and output images of the autoencoder. Furthermore, binarizing the SSIM image makes it easier to identify abnormal pixels, i.e., air bubbles 110 (FIG. 8C). FIG. 8C is a diagram comparing the output image (well image) and input image (well image) in FIG. 8B with an SSIM image.

[0033] FIG. 9 shows an example of recognizing a bubble 110 from a well image. FIG. 9A shows an example of an input image (well image) input to an SSIM autoencoder. FIG. 9A displays images of wells 105 arranged side by side, normalized for each well 105. As shown in FIG. 9A, the wells 105 are arranged regularly (periodically). FIG. 9B shows an SSIM image obtained from the input and output images of the autoencoder. As shown in the SSIM image in FIG. 9B, non-uniform, abnormal components (abnormal pixels), such as areas where bubbles 110 are present, can be extracted. FIG. 9C shows pixels determined to be abnormal pixels after binarizing the SSIM image. As shown in FIG. 9C (and as described with reference to FIG. 8C), binarizing the SSIM image makes it easier and more reliable to recognize abnormal pixels, i.e., bubbles 110. FIG. 9D shows an image in which pixels determined to be abnormal pixels are colored and overwritten on the original well image. As shown in FIG. 9D, it can be seen that the abnormal pixel containing the air bubble 110 can be recognized in the original well image.

[0034] If the air bubble 110 can be recognized in this way, it is possible to correct its influence. For example, pixels determined to be abnormal pixels are excluded and a representative value of the fluorescence intensity of the well 105 is determined. Note that the representative value may be, for example, the average value or the median value. Alternatively, the representative value may be extracted using the output image of the autoencoder as the well image.

[0035] FIG. 10A is a histogram showing the average intensity (average fluorescence intensity) of wells 105 determined to contain bubbles and wells 105 determined to contain no bubbles. FIG. 10B is a histogram showing the average intensity (average fluorescence intensity) of wells 105 determined to contain bubbles before and after correction for the influence of bubbles 110. As shown in FIG. 10A, it can be seen that the distribution of average intensity differs depending on whether or not bubbles 110 are present. Here, the representative values ​​of wells 105 determined to contain bubbles are corrected as described above. As shown in FIG. 10B, this correction results in the distribution of average intensity (before correction) of wells 105 determined to contain bubbles becoming similar to the distribution of average intensity (after correction) of wells 105 determined to contain no bubbles shown in FIG. 10A. This demonstrates that the influence of bubbles 110 has been eliminated by the correction.

[0036] As described above, the digital PCR system 200 determines the presence or absence of target nucleic acid in multiple wells 105, recognizes abnormalities (air bubbles 110), removes their influence, and corrects the representative value of the fluorescence intensity of the abnormal wells 105. This improves the accuracy of calculating the fluorescence intensity of the wells 105. In other words, correcting the representative value makes it equivalent to measuring the fluorescence intensity with high accuracy. Therefore, the accuracy of determining whether a result is positive or negative is also improved. Furthermore, since the digital PCR system 200 can use wells 105 containing air bubbles 110 for analysis, the number of wells that can be analyzed can be maintained. Therefore, the digital PCR system 200 can measure the concentration of the target nucleic acid with high accuracy even if an abnormality occurs in the well 105 into which the target nucleic acid is introduced.

[0037] In the above embodiment, the abnormality in the well 105 is described as an air bubble 110, but this is not limited thereto, and any substance different from the sample solution 108 can be recognized in a similar manner. For example, if oil, which is the separation liquid 109, has entered a part of the well 105, more accurate fluorescence intensity can be obtained by recognizing and correcting the abnormal pixel related to the oil in the same manner as described for the air bubble 110 in this embodiment. Therefore, even in such a case, this embodiment can measure the concentration of the target nucleic acid with high accuracy.

[0038] Second Embodiment A digital PCR system 200 according to the second embodiment will be described with reference to FIG. 11 . FIG. 11 is a flowchart illustrating the processing steps performed by the digital PCR system 200 according to the second embodiment. The method of recognizing bubbles 110 and correcting the representative value of the fluorescence intensity, as described in the first embodiment, is also useful when measuring the melting temperature of each well 105, as described with reference to FIG. 11 . By performing the processing steps shown in FIG. 11 , it is possible to measure the melting temperature of each well 105 after PCR. The configuration of the digital PCR system 200 according to the second embodiment is the same as that according to the first embodiment.

[0039] 11 , first, the sample solution 108 is introduced into each well 105 and divided (S1101). Then, PCR is performed to amplify the target nucleic acid (S1102). Thereafter, the temperature of the digital PCR device 101 is controlled by the temperature controller 201, and fluorescence measurement is performed to measure the temperature dependence of the fluorescence image of each well 105 (S1103).

[0040] The temperature dependence of a fluorescence image can be measured as follows. As described above, the sample solution 108 contains the target nucleic acid, PCR amplification enzyme, buffer, primers, dNTPs, and a probe. In this embodiment, asymmetric PCR is performed by varying the concentrations of the forward primer and reverse primer to generate an amplification product 1201 (see FIG. 12A ) having a sequence complementary to the probe. FIG. 12A is an explanatory diagram illustrating an example of the temperature dependence of fluorescence intensity. As shown in the explanatory diagram in the upper left of FIG. 12A , the amplification product 1201 hybridizes with a probe 1202 at low temperatures. The probe 1202 has a fluorescent dye 1203 and a quencher 1204 attached to its end. When the probe 1202 binds to the amplification product 1201, the fluorescent dye 1203 and the quencher 1204 separate, causing the probe 1202 to emit fluorescence. In other words, the fluorescence intensity increases. 12A, when the temperature of the sample solution 108 rises and exceeds the melting temperature determined by the sequences of the amplified product 1201 and the probe 1202, the amplified product 1201 and the probe 1202 dissociate. As a result, the quencher 1204 is positioned near the fluorescent dye 1203 of the probe 1202, reducing the fluorescence intensity.

[0041] As described above, the melting temperature is determined by the sequences of the amplification product 1201 of the target nucleic acid and the probe 1202. Therefore, the target nucleic acid can be identified by measuring the melting curve and calculating the melting temperature. As an example, by amplifying wild-type and mutant forms of the target nucleic acid and using probes with complementary sequences to the wild-type and mutant forms, melting curves 1205 and 1206, which have different melting temperatures, as shown in FIG. 12A are obtained. In this embodiment, differential curves 1207 and 1208 are obtained by differentiating these melting curves (FIG. 12B), and melting temperatures 1209 and 1210 are calculated as their peaks (S1105). Note that FIG. 12B is an explanatory diagram illustrating an example of a differential curve obtained by differentiating the melting curve. This allows the target nucleic acid introduced into each well 105 to be identified (S1106).

[0042] The digital PCR system 200 described in the first embodiment is also useful for calculating the melting temperature in the second embodiment. Specifically, as in the first embodiment, the presence or absence of bubbles 110 and abnormal pixels in each well 105 are identified from the fluorescence image (well image), and then a representative value of the fluorescence intensity of each well 105 is calculated so as to eliminate the influence of the bubbles 110. In this embodiment, this process is repeated for the fluorescence image corresponding to the measured temperature. In this way, the melting curve, differential curve, and melting temperature are calculated while excluding the influence of the bubbles 110, thereby enabling accurate calculation of the melting temperature. If the influence of the bubbles 110 is present, noise is superimposed on the measured melting curve due to the influence of scattered light and the movement of the bubbles 110 during temperature changes. This results in errors in the calculated melting temperature. This may result in a decrease in the accuracy of identifying the target nucleic acid. In the second embodiment, the target nucleic acid can be accurately identified by accurately calculating the fluorescence intensity and melting temperature while eliminating the influence of the bubbles 110. For example, wild-type and mutant types can be accurately identified.

[0043] <Target Nucleic Acid Detection Method> Next, one embodiment of a target nucleic acid detection method (hereinafter, sometimes referred to as "this method") will be described with reference to Fig. 13. Fig. 13 is a flowchart illustrating the content of the target nucleic acid detection method according to this embodiment. This method detects a target nucleic acid in a sample solution 108 by digital PCR.

[0044] As shown in FIG. 13 , this method includes the following steps a) to g). Step a) is a division step (S1301). In step a), the sample solution 108 is divided into multiple wells 105. The multiple wells 105 are divided using a separation liquid 109, such as oil or photocurable resin. PCR is performed between step a) and the following step b). Step b) is an imaging step (S1302). In step b), an image including the multiple wells 105 is captured. The image is captured by the fluorescence detector 202 (imaging device). Step c) is a first recognition step (S1303). In step c), the positions of the multiple wells 105 in the captured image are recognized. Step c) is performed by steps S401 to S404 and steps S501 to S506, as described with reference to FIGS. 4 and 5 . Step d) is a second recognition step (S1304). In step d), the presence or absence of anomalies (air bubbles 110) in each well 105 and the location of the anomalies within the well 105 are recognized. As described with reference to FIG. 6, step d) is performed by steps S601 to S604. Step e) is a calculation step (S1305). In step e), a representative value of the fluorescence intensity of each well 105 is calculated. Examples of the representative value of the fluorescence intensity include the average value and median value of the fluorescence intensity. Step f) is a correction step (S1306). In step f), the representative value of the fluorescence intensity of the wells 105 containing anomalies is corrected. The correction of the representative value can be performed, for example, by removing the abnormal portion (abnormal pixel) of the well 105. In other words, the corrected representative value becomes the average value or median value of the fluorescence intensity of the well 105 after removing the abnormal portion. Step g) is a discrimination step (S1307). In step g), the presence or absence of the target nucleic acid in multiple wells 105 is discriminated. The presence or absence of the target nucleic acid can be determined by whether or not the fluorescence intensity is measured or whether or not the fluorescence intensity is higher than a predetermined threshold. Note that steps c) to g) are performed by the data analysis unit 203 (control unit).

[0045] As described above, this method determines the presence or absence of target nucleic acid in multiple wells 105, recognizes abnormalities (air bubbles 110), removes their influence, and corrects the representative value of the fluorescence intensity of the abnormal wells 105. This improves the accuracy of calculating the fluorescence intensity of the wells 105. In other words, correcting the representative value makes it equivalent to measuring the fluorescence intensity with high accuracy. Therefore, the accuracy of determining whether a result is positive or negative is also improved. Furthermore, this method allows wells 105 containing air bubbles 110 to be used for analysis, so the number of wells that can be analyzed can be maintained. Therefore, this method can measure the concentration of the target nucleic acid with high accuracy even if an abnormality occurs in the well 105 into which the target nucleic acid is introduced.

[0046] <Target Nucleic Acid Detection Storage Medium> Next, with reference to FIG. 14 , one embodiment of a target nucleic acid detection storage medium (hereinafter, sometimes referred to as "the storage medium") 1400 will be described. FIG. 14 is a configuration diagram illustrating the configuration of the target nucleic acid detection storage medium 1400 according to this embodiment and the configuration of a computer 10 constituting a data analysis unit. The storage medium 1400 is used to detect target nucleic acids in a sample solution 108 by digital PCR. The storage medium 1400 is readable by the computer 10 and stores a program for causing the computer 10, specifically the data analysis unit 203 (control unit), to execute each of the steps shown in FIG. 14 after capturing an image including a plurality of wells 105 into which the sample solution 108 has been introduced. Note that steps c) to g) (S1403 to S1407) in the storage medium 1400 are similar to steps c) to g) (S1303 to S1307) of the method described above.

[0047] First, the computer 10 will be described. As shown in FIG. 14 , the computer 10 includes a CPU 11, a RAM 14, a ROM 15, a HDD 16, a communication I / F 17, an input / output I / F 18, and a media I / F 19. The communication I / F 17 is connected to a communication circuit 20. The input / output I / F 18 is connected to an input / output device 21. The media I / F 19 reads and writes data from a recording medium 22. The ROM 15 stores application programs executed by the CPU 11, various data, and the like. The HDD 16 stores image data, various data, control programs, and the like as appropriate. The HDD 16 may be a solid-state drive (SSD), for example. The CPU 11 executes application programs loaded into the RAM 14 to realize various functions. The digital PCR system 200 (not shown in FIG. 14 ) may be connected to one computer 10 or to multiple computers 10 via the communication circuit 20 and the communication I / F 17.

[0048] The computer 10 executes a program using a processor (e.g., a CPU, a GPU) and performs processing defined by the program using storage resources (e.g., a memory) and interface devices (e.g., a communication I / F 17). Therefore, the entity that executes the program and performs the processing may be the processor. Similarly, the entity that executes the program and performs the processing may be a controller, device, system, computer, or node that has a processor. The entity that executes the program and performs the processing may be a control unit, and may include a dedicated circuit that performs specific processing. Examples of the dedicated circuit include an FPGA (Field Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), and a CPLD (Complex Programmable Logic Device).

[0049] Various programs, including the program stored in the storage medium 1400, may be installed on the computer 10 from a program source. The program source may be, for example, a program distribution server or a computer-readable storage medium (this storage medium may include the storage medium 1400). When the program source is a program distribution server, the program distribution server may include a processor and storage resources for storing the program to be distributed, and the processor of the program distribution server may distribute the program to be distributed to other computers. In addition, in an embodiment, two or more programs may be realized as one program, or one program may be realized as two or more programs.

[0050] Next, steps c) to g) (S1403 to S1407) of the storage medium 1400 will be described. As shown in FIG. 14, step c) is the first recognition step (S1403). In step c), the positions of multiple wells 105 in the captured image are recognized. As described with reference to FIGS. 4 and 5, step c) is performed through steps S401 to S404 and steps S501 to S506. Step d) is the second recognition step (S1404). In step d), the presence or absence of anomalies (air bubbles 110) in each well 105 and the position of the anomalies within the well 105 are recognized. As described with reference to FIG. 6, step d) is performed through steps S601 to S604. Step e) is a calculation step (S1405). In step e), a representative value of the fluorescence intensity of each well 105 is calculated. The representative value of the fluorescence intensity may be, for example, the average or median of the fluorescence intensity. Step f) is a correction step (S1406). In step f), the representative value of the fluorescence intensity of the wells 105 having an abnormality is corrected. The correction of the representative value can be performed, for example, by removing the abnormal portion (abnormal pixel) of the well 105. In other words, the representative value after correction becomes the average or median of the fluorescence intensity of the well 105 from which the abnormal portion of the well 105 has been removed. Step g) is a determination step (S1407). In step g), the presence or absence of the target nucleic acid in multiple wells 105 is determined. The presence or absence of the target nucleic acid can be determined by whether or not the fluorescence intensity is measured or whether or not the fluorescence intensity is higher than a predetermined threshold.

[0051] The storage medium 1400 may be provided inside the computer 10 or may be provided outside the computer 10 via the communication I / F 17, the media I / F 19, or the like.

[0052] As described above, the storage medium 1400 executes the computer 10 to determine the presence or absence of target nucleic acid in multiple wells 105, recognize abnormalities (air bubbles 110), eliminate their influence, and correct the representative value of the fluorescence intensity of the abnormal wells 105. This improves the accuracy of calculating the fluorescence intensity of the wells 105. In other words, correcting the representative value makes it equivalent to measuring the fluorescence intensity with high accuracy. Therefore, the accuracy of determining whether a result is positive or negative is also improved. Furthermore, the storage medium 1400 can maintain the number of analyzable wells because wells 105 containing air bubbles 110 can also be used for analysis by the computer 10. Therefore, the storage medium 1400 can measure the concentration of the target nucleic acid with high accuracy even if an abnormality occurs in the well 105 into which the target nucleic acid is introduced.

[0053] The digital PCR system 200, the target nucleic acid detection method, and the target nucleic acid detection storage medium 1400 according to the present invention have been described in detail above using embodiments. However, the present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0054] 10 Computer 101 Digital PCR device 105, 105a, 105b, 105c Well 108 Sample solution 110 Air bubble (abnormal) 200 Digital PCR system 202 Fluorescence detector (imaging device) 203 Data analysis unit (control unit) 1209 Melting temperature 1210 Melting temperature 1400 Target nucleic acid detection storage medium (this storage medium)

Claims

1. A system for detecting a target nucleic acid in a sample solution by digital PCR, comprising: an imaging device that captures an image including a plurality of wells into which the sample solution has been introduced; and a control unit that recognizes the positions of the plurality of wells in the image, recognizes the presence or absence of an abnormality in each of the wells and the position of the abnormality within the well, calculates a representative value of the fluorescence intensity of each of the wells, corrects the representative value of the fluorescence intensity of wells having the abnormality, and determines the presence or absence of the target nucleic acid in the plurality of wells.

2. A digital PCR system as described in claim 1, wherein the control unit recognizes the presence or absence of an abnormality and the location of the abnormality using machine learning learned by inputting images of normal wells, and recognizes the wells containing an abnormality and the location of the abnormality.

3. A digital PCR system according to claim 1, wherein a photocurable resin is used as a separating liquid for dividing the sample solution into the plurality of wells, and the wells are divided by hardening the photocurable resin.

4. A digital PCR system according to claim 1, wherein the control unit acquires the temperature dependency of the image and calculates the melting temperature of the target nucleic acid using a representative value of the corrected fluorescence intensity.

5. A method for detecting a target nucleic acid in a sample solution by digital PCR, comprising: a) dividing the sample solution into a plurality of wells; b) capturing an image including the plurality of wells; c) recognizing the positions of the plurality of wells in the image; d) recognizing the presence or absence of an abnormality in each of the wells and the position of the abnormality within the well; e) calculating a representative value of the fluorescence intensity of each of the wells; f) correcting the representative value of the fluorescence intensity of wells having the abnormality; and g) determining the presence or absence of the target nucleic acid in the plurality of wells.

6. A computer-readable storage medium for detecting a target nucleic acid in a sample solution by digital PCR, having recorded thereon a program for causing a computer to execute the following steps after capturing an image including a plurality of wells into which the sample solution has been introduced: c) recognizing the positions of the plurality of wells in the image; d) recognizing the presence or absence of an abnormality in each of the wells and the position of the abnormality within the well; e) calculating a representative value of the fluorescence intensity of each of the wells; f) correcting the representative value of the fluorescence intensity of wells having the abnormality; and g) determining the presence or absence of the target nucleic acid in the plurality of wells.

Images