JP2002335963A - System, method and program for identifying microorganism - Google Patents

Abstract

PROBLEM TO BE SOLVED: To identify a microorganism in a sample by collation with the database of the waveform data for luminance distribution obtained from an electrophoretic image. SOLUTION: The method for identifying the aimed microorganism comprises the following procedure: a DNA fragment in a sample is amplified using an SSC-PCR amplifier/analyzer 1 to obtain an electrophoretic image; the image is then converted into the corresponding image data with an electrophoretic image input unit 2, the data is taken in an analyzing computer 3, which, in turn, performs an operation for the correlation between a sample waveform data and the reference waveform data stored in a database memory 4 and selects a data of greater correlation (similarlity) and displays it on a result display means 5, wherein, among the sample waveform data, the correlation to be put to the operation is between portions with the peaks coinciding with each other and the whole reference waveform data or between only portions where the respective peaks of the sample waveform data and the reference waveform data coincide with each other.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to collating sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. The present invention relates to a technique for identifying microorganisms by using

[0002]

2. Description of the Related Art In recent years, garbage disposers for composting organic waste (garbage) and the like discharged from households and the like have been developed. In composting using a garbage disposer, the degree of composting is evaluated by monitoring the temperature, etc. during the organic matter decomposition process, and the state of the garbage disposal is adjusted based on this evaluation so that good quality fertilizer is produced. ing.

[0003] On the other hand, in order to produce higher quality fertilizers, it is necessary to have information on the microorganisms that function inside the garbage disposer, and to improve the soil by adding the produced fertilizers. Also, information on microbial groups in soil is needed.

On the other hand, pathogenic bacteria are isolated and cultured from patient samples, and the pathogenic bacteria are identified by biochemical tests and the like. At this time, if bacteria can be identified directly from DNA, it will be an important method in clinical medicine.

As a method for obtaining information on a group of microorganisms, there is a method of isolating individual bacteria contained in a group of bacteria and conducting a biochemical test. However, this method has a problem that it takes a long time and cannot be applied to bacteria that are difficult to isolate.

There is also known a method of obtaining information on a group of microorganisms by performing DNA analysis. To perform DNA analysis, it is necessary to amplify DNA, and a PCR (polymerase chain reaction) method is used. In the PCR method, using a primer having a base sequence complementary to the base sequence at both ends of the DNA to be amplified (template DNA) and a heat-resistant DNA polymerase, a cycle consisting of three steps of heat denaturation, annealing and extension is repeated. Can amplify a DNA fragment substantially the same as the template DNA. In particular,
A predetermined fragment in one DNA of a bacterium existing only in a small amount can be amplified, for example, 100,000 to 1,000,000 times. Regarding the PCR method, for example, USP 4,683,1
No. 95, etc.

However, in order to use the PCR method, the template D
It is premised that the nucleotide sequences of at least both ends of one region of NA are known, and a dilemma arises that the DNA fragments of these microorganisms cannot be amplified unless the types and sequences of the microorganisms in the garbage disposer are known.

[0008] Therefore, a single primer can be used to simultaneously convert many types of DNA from one type of DNA without knowledge of the base sequence.
RAPD (Randum Amplified Polymorph) for amplifying fragments
ic DNA) method or AP-PCR (Arbitrarity Primed)
-Polymerase Chain Reaction) method has been proposed. In these methods, the sequence specificity at the time of primer binding is reduced by lowering the annealing temperature of the primer during the PCR reaction and further increasing the magnesium ion concentration in the reaction solution. Then, the primer binds to the DNA of the microorganism with a mismatch, and the DNA fragment is replicated. Some DNA with a single primer
After the fragments have been amplified in large amounts, the amplified DNA fragments are separated by gel electrophoresis to obtain a DNA fingerprint, and by analyzing the DNA fingerprint, information on microorganisms can be obtained.

[0009]

However, when such a RAPD method or an AP-PCR method is applied to a microorganism group composed of a plurality of microorganisms such as a garbage disposer or soil, the DNA to be amplified is increased. Since the number of types of fragments is too large, it is difficult to distinguish microorganisms constituting a microorganism group from the amplified DNA fragments.

Of course, theoretically, a database of the fingerprints of DNA fragments whose base sequences are known is prepared as a standard fingerprint group, and the fingerprint of the sample is compared with the fingerprint of this database to identify microorganisms. Although it is possible to identify the fingerprint, the shape of the obtained fingerprint is various, and it is difficult to identify the fingerprint with high accuracy because it includes noise.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems of the prior art, and has as its object to print a fingerprint of a sample and a database, specifically, waveform data of luminance obtained from an electrophoretic image with high precision. The present invention provides an apparatus, a method, and a program that can identify microorganisms with high accuracy by collating the information.

[0012]

In order to achieve the above-mentioned object, the present invention provides a database storing sample waveform data indicating a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers. An apparatus for identifying a microorganism by comparing the sampled waveform data with the extracted standard waveform data, comprising: an extracting unit that extracts a portion of the sample waveform data that matches the standard waveform data; and a portion that matches the sample waveform data. Computing means for computing the correlation between the sample waveform data and the standard waveform data using

Here, it is preferable that the calculating means calculates a correlation between a portion where the sample waveform data matches and the standard waveform data.

Preferably, the calculating means calculates a correlation between a matching part of the sample waveform data and a matching part of the standard waveform data.

Further, the present invention provides a method for comparing microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. An extracting means for extracting a portion of the sample waveform data that matches the standard waveform data; a first arithmetic means for calculating a correlation between the sample waveform data and the standard waveform data; A second calculating means for calculating a correlation between the matched part of the sample waveform data and the standard waveform data, wherein matching is performed based on the correlation by the first calculating means and the correlation by the second calculating means. And

Further, the present invention provides a method for comparing microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. An extracting means for extracting a portion of the sample waveform data that matches the standard waveform data; a first arithmetic means for calculating a correlation between the sample waveform data and the standard waveform data; A second calculating means for calculating a correlation between a matching part of the sample waveform data and a matching part of the standard waveform data, and performing collation based on the correlation by the first calculating means and the correlation by the second calculating means. It is characterized by doing.

Also, the present invention provides a method for comparing microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. An extracting means for extracting a portion of the sample waveform data that matches the standard waveform data; a first arithmetic means for calculating a correlation between the sample waveform data and the standard waveform data; Second calculating means for calculating a correlation between the matching part of the sample waveform data and the standard waveform data, and third calculating means for calculating a correlation between the matching part of the sample waveform data and the matching part of the standard waveform data Based on the correlation by the first computing means, the correlation by the second computing means, and the correlation by the third computing means. Characterized by collating Te.

In the apparatus, the extracting means may include means for normalizing at least one of the sample waveform data and the standard waveform data with a predetermined width and a predetermined height, and a means for normalizing the standardized waveform data and another waveform. It is preferable to have means for determining a match based on the result of multiplication with the data.

Further, in the present apparatus, it is preferable that at least one of the sample waveform data and the standard waveform data is shaped by a predetermined normalization function.

Further, the present invention provides a method for comparing microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Provide a way to identify The method includes extracting a portion of the sample waveform data that matches the standard waveform data, and calculating a similarity between the sample waveform data and the standard waveform data using the portion that matches the sample waveform data. And the step of performing.

Here, in the calculating step, it is preferable to calculate a similarity between a portion where the sample waveform data matches and the standard waveform data.

In the calculating step, it is preferable that a similarity between a matching part of the sample waveform data and a matching part of the standard waveform data is calculated.

Further, the present invention provides a method for comparing microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Provide a way to identify The method includes: extracting a portion of the sample waveform data that matches the standard waveform data; calculating a similarity between the sample waveform data and the standard waveform data; Calculating a similarity between the portion and the standard waveform data, and performing collation based on a plurality of similarities.

Further, the present invention provides a method for comparing microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting a portion of the sample waveform data that matches the standard waveform data; calculating a similarity between the sample waveform data and the standard waveform data; Calculating a similarity between a matching part of the data and a matching part of the standard waveform data, wherein the matching is performed based on a plurality of similarities.

The present invention also provides a method for comparing microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting a portion of the sample waveform data that matches the standard waveform data; calculating a similarity between the sample waveform data and the standard waveform data; Calculating a similarity between a matching part of the data and the standard waveform data; and calculating a similarity between a matching part of the sample waveform data and a matching part of the standard waveform data. The matching is performed based on the similarity.

In the method, the extracting may include normalizing at least one of the sampled waveform data and the standard waveform data with a predetermined width and a predetermined height; and extracting the standardized waveform data and another standardized waveform data. Determining a match based on the result of multiplication with the waveform data.

In this method, it is preferable that at least one of the sample waveform data and the standard waveform data is shaped by a predetermined normalization function.

Further, the present invention uses a computer to compare sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. To provide a program for identifying microorganisms. The program causes a computer to determine at least a match between the sample waveform data and the standard waveform data by a processor, and performs a correlation calculation of a part of the sample waveform data and all or part of the standard waveform data. Let the processor perform the correlation operation,
The processor is configured to compare a result of the correlation operation with a predetermined threshold value stored in a memory.

Further, the present invention uses a computer to compare sample waveform data showing a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. To provide a program for identifying microorganisms. This program causes the processor to determine at least the match between the sample waveform data and the standard waveform data, and causes the processor to execute a correlation operation between the sample waveform and the standard waveform data and store the memory in the memory. The processor causes the processor to execute a correlation operation on all or part of the matching portion of the sample waveform data and the standard waveform data and store the result in the memory. The threshold value is compared with a processor.

As described above, according to the present invention, the sample waveform data and the standard waveform data (the waveform data stored in the database) are compared with each other by performing the correlation calculation (correlation between the entire waveforms). Focusing on a portion that matches the standard waveform data, a correlation calculation is performed using the matching portion. There are two types of correlation calculation using the matching part, one is a correlation calculation of the matching part of the sample waveform data and the entire standard waveform data (that is, the correlation between the matching peak and the whole). Reference numeral 2 denotes a correlation operation (i.e., a correlation between matching peaks) between a matching portion of the sample waveform data and a matching portion of the standard waveform data. By performing a correlation operation by focusing on the matching part, the influence of noise can be suppressed, and the hit rate of the database, that is, the identification rate can be increased.

In the present invention, the correlation between the entire sample waveform data and the standard waveform data (similarity 1),
Three correlation values or similarities of the correlation between the matching peaks (similarity 2) and the correlation between the matching peak and the whole (similarity 3) are used. These correlation values or similarities are appropriately selected, or They can be used in combination.
As a mode of selection, only the similarity 2 or the similarity 3
Only can be used, and as a mode of combination,
There are similarity 1 and similarity 2, similarity 1 and similarity 3, similarity 2 and similarity 3, and similarity 1 and similarity 2 and similarity 3. Which one to use may be determined according to the sample to be identified.

In the present invention, when performing the correlation calculation, the calculation may be performed by focusing on the area of the peak where the waveform data coincides. For example, the ratio of the area of the peak corresponding to the sample waveform of the standard data waveform to the area of all the peaks of the standard waveform data indicates the degree of correlation or similarity between the sample waveform and the standard data waveform. Either the correlation value or the area ratio may be used, and the microorganisms may be comprehensively identified using both the correlation value and the area ratio.

Further, when performing the correlation calculation, the standard waveform data can be further processed and used. That is,
When the standard waveform data are similar to each other, there is a case where a large difference does not occur in the correlation value even when the correlation operation with the sample waveform is performed. In such a case, the standard waveform data is reconstructed by comparing the standard waveform data, removing peaks that match each other, and leaving only peaks that do not match each other. The standard waveform data (orthogonalized standard waveform data) thus obtained is standard waveform data having mutually different peaks. When the correlation calculation is performed with the sample waveform data, if the same microorganism is used, the correlation is significantly increased while the correlation is different. For microorganisms, the correlation is significantly reduced,
The identification of microorganisms is facilitated.

[0034]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of a microorganism identifying apparatus according to this embodiment. The microorganism identification device includes an SSC-PCR amplification / analysis device 1, an electrophoresis image input unit 2, an analysis computer 3, a database storage unit 4, and a result display unit 5.

The SSC-PCR amplification / analysis device 1 is an SS
The DNA fragment was amplified using the C-PCR method and the resulting D
Create an electrophoretic image of the NA fragment. SSC-PCR (Si
The ngleStrain Counting Polymerase Chain Reaction) method is a method of amplifying a DNA fragment in a chain reaction from a group of microorganisms having an unknown base sequence or a single microorganism using a plurality of primers having a specific base sequence. This SSC-PCR method will be briefly described.

<SSC-PCR> In the SSC-PCR, amplification is performed by repeating the steps of thermal denaturation, annealing, and extension as in the conventional PCR. However, SS
In the C-PCR method, amplification is performed using a primer having a specific base sequence for a microorganism or a group of microorganisms having an unknown base sequence. In the heat denaturation step, the DNA or DNA fragment is denatured by heating to be single-stranded. In the annealing step, heat treatment is performed so that the primer binds to the end of the amplification region of the DNA strand. In the elongation reaction step, a complementary strand is synthesized by a polymerase starting from the primer,
It becomes a main chain. Thus, a DNA fragment is formed,
A new DNA fragment is formed from the NA fragment, and a DNA fragment is also formed from other DNAs of the same type, and the same reaction is continued in a chain reaction and amplified.

In the SSC-PCR method, by using primers each having a length of 12 base pairs (bp), a DNA fragment is stochastically generated once every 4 ¹² (about 10 ⁷ ) in each primer. It can be adjusted so as to appear, and can stochastically amplify only one type of DNA fragment from a microbial DNA generally having a DNA length of about 10 ⁷ base pairs. Therefore, it is possible to make the amplified DNA fragment correspond to the microorganism on a one-to-one basis.

The SSC-PCR amplification / analysis device 1 amplifies a DNA fragment as described above. And SSC
-Size (base pairs) of DNA fragments amplified in each reaction solution by electrophoresis using a PCR reaction solution
Fractionate each time. At this time, the concentration is known and quantifiable D
The NA size markers are electrophoresed simultaneously and fractionated by size. Next, the electrophoretic image obtained by the electrophoresis method is stained with a fluorescent dye, and a fluorescent image when ultraviolet light is irradiated is taken. For photographing, a CCD camera can be used. D
The NA fragment appears as a band in the electrophoresis image. The electrophoresis image is supplied to the electrophoresis image input unit 2. The electrophoresis method is used to separate DNAs having different sizes (number of bases (pairs), length, molecular weight, etc.). In addition, the luminance distribution (luminance value) is determined based on the final concentration (mol / l, g / l, etc.) or amount (mol,
g, etc.).

The electrophoresis image input unit 2 is constituted by a scanner or the like, converts an electrophoresis image into image data, and outputs the image data to the analysis computer 3.

After an electrophoretic image is obtained by electrophoresis, an image of the electrophoretic image is not obtained, but another method such as chromatography can be used to separate DNA having different sizes. In addition, the electrophoresis method also has various forms such as a gel form and a method of flowing electricity. For example, a gel filled in a capillary can be used in addition to the flat gel shown here. When these methods are used, the measured values differ depending on the method.
Is used, the DNA concentration is measured as a luminance value. In the case of a capillary gel, a sensor reading (sensor) is used to measure a sensor reading (fluorescence intensity or absorbance).
When measured by a CCD or a sensor, waveform data can be obtained directly without taking the form of an image.

The analysis computer 3 inputs the electrophoresis image data, and converts the concentration data depending on the size of the DNA for each primer, specifically, the waveform data indicating the brightness distribution (fluorescence brightness distribution) of the electrophoresis image. The generated waveform data is compared with the waveform data of known microorganisms stored in the database storage unit 4 in advance to identify the input waveform, that is, the microorganism of the sample. The result obtained by collating with the database storage unit 4 is output to a display such as a CRT or a result display unit 5 such as a printer. The generation of the waveform data in the analysis computer 3, the correction of the waveform, or the comparison with the waveform data in the database storage unit 4 will be further described later.

FIG. 2 is a block diagram showing the configuration of the analyzing computer 3 in FIG. The analysis computer 3 includes a CPU 310, a display 320, an input device 330, a ROM 340, a RAM 350, a recording medium driving device 360, a scanner 370, and an external storage device 380.

The display 320 is a liquid crystal display or a CR display.
T and the like, and can function as the result display means 5 in FIG. The input device 330 includes a keyboard, a mouse, and the like, and is used to input various data and various commands. A system program is stored in the ROM 340.

The recording medium drive 360 is a CD-ROM
It comprises a drive and the like, and reads / writes data from / to the recording medium 390. The recording medium 390 stores a microorganism identification program for performing the processing according to the present embodiment, and the CPU 310 sequentially executes the program.

The scanner 370 also functions as the electrophoresis image input unit 2 in FIG. 1, inputs an electrophoresis image as image data, and stores the image data in the external storage device 380.

The external storage device 380 is composed of a hard disk or the like, and the storage medium 3
The microbial identification program read from 90 is stored. In addition, the external storage device 380 can also function as the database storage unit 4 in FIG. 1 by storing waveform data of known microorganisms. Of course, the microorganism identification program can be stored in the external storage device 38 via the network.
0, or may be stored in the ROM 340 in advance.

FIG. 3 shows a processing flowchart of the present embodiment. First, a microorganism analysis experiment is performed by the SSC-PCR amplification / analysis device 1 as described above (S1). Here, in the analysis experiment, the isolated microorganism may be used as a sample, or a group of microorganisms including a plurality of microorganisms may be used as a sample. After performing the analysis experiment, the obtained electrophoresis image is taken into the analysis computer 3 (S2). The analysis computer 3 calculates D
A luminance distribution as a function of the NA size is generated as waveform data (S3), and the waveform data is further corrected (S3).
4). The waveform data is generated by plotting the luminance value with the DN size of the image data as the horizontal axis and the luminance value as the vertical axis. Considering the case where waveform data indicating the luminance distribution cannot be accurately generated due to the distorted electrophoretic image, the waveform data may be generated after performing various image corrections including affine transformation on the electrophoretic image. Good. On the other hand, the correction processing of the waveform data is executed as shown in FIG. Hereinafter, the waveform data correction processing will be described.

<Correction of Waveform Data> Since uneven dyeing occurs when the electrophoretic image is dyed, if the unevenness occurs in the background of the electrophoretic image, the luminance distribution cannot be accurately detected.
In addition, when amplifying a DNA fragment and arranging the primer so as to strongly bind to a suitable position of a template DNA having a predetermined base sequence, the amplification efficiency of the DNA fragment is increased. The DNA fragment amplified by such a primer appears as a clear band with good reproducibility in the electrophoresis image, but when the bond between the primer and the compatible position of the template DNA is weak, it binds more than the compatible position of the template DNA. The primer binds to other positions where is strong. As described above, since the reaction proceeds competitively in the amplification reaction of the DNA fragment, the amplification efficiency of the DNA fragment is reduced when the binding between the primer and the template DNA is weak. A DNA fragment amplified by a weakly bound primer appears as an indistinct band in the electrophoresis image with low reproducibility, and the reliability is reduced.

Therefore, in the waveform correction processing, first, the background unevenness of the electrophoretic image is corrected (S40). Specifically, a rectangular region including a band is removed from the electrophoretic image, and interpolation of the removed rectangular region is performed based on luminance around the removed rectangular region. Then, based on the obtained background data, waveform data of a luminance distribution in which the influence of the luminance of the background is removed is generated by the following equation.

[0051]

## EQU1 ## Znew = Zmax. (Z1-Z0) / (Zm
ax-Z0) Here, Znew is the luminance value in the luminance distribution after correction, Zmax is the maximum luminance value before correction, Z0 is the minimum luminance value before correction, and Z1 is the luminance value at a certain point in the luminance distribution. In the above equation, (Z1-Z0) / (Zma
x-Z0) gives the ratio of the relative luminance to the whole of a certain point, and multiplying this by Zmax gives the relative luminance.

After correcting the background unevenness as described above,
By comparing the band emission intensity of the positive control with the emission intensity of the band of the DNA size marker whose amount is known in the waveform data, the DNA fragment amplified in the positive control is quantified, and the DNA fragment is quantified using the quantified value. The amplification efficiency of the amplification reaction is obtained, and the luminance of the waveform data is corrected based on the amplification efficiency (S
41). Here, the positive control refers to DNA
This is a sample used for a control experiment for removing an error generated in the fragment amplification step. In general, the amplification rate of a DNA fragment is affected by a concentration error of DNA polymerase or magnesium, a degree of activity of a used reagent, and a temperature error. In order to eliminate such an error, a DNA fragment amplified as a positive control is used. A primer that amplifies a DNA fragment of a known and quantifiable type;
A reaction solution containing a template DNA having a base sequence complementary to the primer is prepared. DNA fragments amplified in the positive control can be quantified by electrophoresis. Therefore, based on the amount of the DNA fragment amplified in the positive control,
The amplification efficiency of the fragment can be determined. The luminance can be corrected based on the amplification efficiency thus obtained.

After the luminance is corrected, the luminance of the waveform data corresponding to a plurality of types of reaction solutions for SSC-PCR is corrected using the luminescence intensity of the quantifiable DNA size marker band as a reference (S42). . Specifically, tone correction of the image data after electrophoresis is performed based on the emission intensity of the band of the DNA size marker whose concentration is known.

Then, axis alignment and concentration adjustment are performed based on the waveform data of the DNA size marker. That is, different gels used for electrophoresis have different band mobilities even under the same conditions. Therefore, the relationship between the band position and the DNA size is different between different gels. Therefore, the waveform data is standardized (axis alignment) so that the waveform data can be compared between different gels. Specifically, the alignment is performed by determining parameters of a predetermined conversion formula based on the waveform data of the DNA size marker, and correcting the waveform data for each gel based on the conversion formula. For example, the following equation can be used as the conversion equation.

[0055]

Y = m1 + log (m2−m3 · x) / (m4
−m5 · x) Here, x indicates the distance from the start position of electrophoresis, that is, the migration distance (pixels / 1000), and y indicates D
The size of NA is indicated by a logarithmic value (log (bp)).

On the other hand, near the middle of the migration distance, the relationship between the migration distance and the DNA size is close to a straight line. Increase. Therefore, when the axis of the waveform data is aligned by the above equation, “stretching” occurs in a region near both ends of the waveform data, and the width of the peak is widened. Further, the area of the peak also increases. Since the peak area corresponds to the DNA concentration, it is necessary to correct the peak height of the waveform data after the axis alignment so that the waveform data before the axis alignment and the peak area become equal. This correction is density adjustment.

The density adjustment can be performed as follows. That is, y = f (x) is obtained by substituting the parameters m1 to m5 determined in the axis alignment, and the first derivative y ′ = f ′ (x) is obtained. Further, a second derivative y ″ = f ″ (x) is obtained. Then, an inflection point x0 at which the second derivative becomes 0 is obtained. Slope f 'at inflection point x0
From (x0), the stretching rate f ′ (x) / f ″ at each point
(X0) is obtained. The luminance value at each point of the waveform data is divided by the stretching rate f ′ (x) / f ″ (x0). As a result, the height of the peak is reduced in a region near both ends of the waveform data, and the area of the peak is reduced. Correction is performed so that it becomes equal to that before the axis alignment.

Note that the processing in S40 to S43 is as follows.
Japanese Patent Application No. 2000-403350 previously filed by the present applicant.
It is also described in the issue, and please refer to it as appropriate.

After performing the above-described correction processing, the waveform data of a plurality of primers are connected to each other to make one waveform data the waveform data of the sample, and a DNA fragment of a known microorganism can be obtained by the same procedure. By comparing the waveform data stored in the database storage unit 4 with the waveform data, the waveform data of the sample can be identified. However, noise is included in the corrected waveform data, and the same condition cannot be satisfied. As described above, even if the obtained waveform data is compared with the waveform data in the database storage unit 4 as it is, there are many cases where a satisfactory result is not obtained.

Therefore, in the present embodiment, the waveform data is further corrected, so that higher-precision collation can be performed.

That is, with respect to the waveform data after the axis alignment and the density adjustment, the analysis computer 3 next fits each peak of the waveform data to a normalization function such as a Gaussian function (approximately by a Gaussian function), and positions the peaks. , Area and shape data are extracted (S44). Then, the original waveform is reconstructed by adding all the Gaussian functions of the respective peaks (S45).

FIG. 5 shows how the normalization function is approximated and how the original waveform is reconstructed using the normalization function.
In the figure, the horizontal axis represents DNA size, and the vertical axis represents luminance. (A) is original waveform data obtained using a certain primer, and (b) is waveform data obtained by adjusting the axis and adjusting the concentration of the waveform of (a).
(C) is a waveform when each peak is approximated by a Gaussian function to the waveform data of (b). The position and area of each peak expressed by the Gaussian function, and the shape (average value and variance) of the Gaussian function are calculated and stored. (D) is waveform data reconstructed by adding the Gaussian function of each peak. Comparing (b) and (d), it can be seen that the shape of each peak has a normal distribution, and small irregularities have been removed. With the above processing, the correction processing of the waveform data is completed.

Referring back to FIG. 3, after the waveform data has been corrected as described above, the waveform data for a plurality of primers are linked to form one waveform data (S
5). Then, the created waveform data is collated with the waveform data stored in the database storage unit 4 in advance, and it is determined which waveform data corresponds (S6).

Here, it is not always necessary to link the waveform data, and the average (or sum, maximum value, minimum value, standard deviation, etc.) of the results of each primer is used without linking, or processing is performed in parallel for each primer. It is also possible. Therefore, in S5, the waveform data is stored in a specific location, and is stored as a waveform data set in the database storage unit 4.
May be compared with the waveform data stored in the.

Normally, the waveform data itself created in S5 is compared with the waveform data itself stored in the database storage unit 4, and the similarity is evaluated by performing a predetermined correlation operation, and the waveform having the highest similarity is evaluated. Identify with data,
In the present embodiment, similarity is determined with higher accuracy by performing more various correlation operations.

FIG. 6 shows a flowchart of a database search process in this embodiment. First, the analysis computer 3 calculates a similarity 1 between the sample waveform data and the waveform data (standard waveform data) in the database storage unit 4 (S60). The similarity 1 is a result of a correlation operation between the entire sample waveform data and the entire standard waveform data. Specifically, the sample waveform data is Si, the standard waveform data is Di, and n is the number of points of each waveform.

(Equation 3) Is performed, and the similarity between the two waveform data is calculated. The similarity 1 obtained by the calculation is stored in the analysis computer 3.

Next, the analysis computer 3 calculates the similarity 2 for the sample waveform data and the standard waveform data (S61). The similarity 2 is a correlation calculation result obtained by extracting only the peak portions that match each other from the sample waveform data and the standard waveform data and performing the matching on the matching peak portions, and the peak portion Sci that matches Si and Di in the above equation. , Dci.
The method of extracting the coincident peak will be described later.

Further, the analysis computer 3 calculates the similarity 3 for the sample waveform data and the standard waveform data (S62). The degree of similarity 3 is a correlation calculation result obtained by extracting only the peak portion that matches the standard waveform data from the sample waveform data and performing the matching peak portion and the entire standard waveform data. This is the similarity when Sci is used.

Hereinafter, the similarities 1 to 3 will be described in more detail.

<Similarity 1> FIG. 7 schematically shows a correlation calculation for calculating the similarity 1. In the figure, the horizontal axis represents DNA size, and the vertical axis represents luminance. The thin line is the sample waveform data, and the thick line is the standard waveform data. A correlation operation is performed on the entire sample waveform data and the entire standard waveform data, and a similarity 1 is calculated. In this case, for example, similarity 1 =
It is calculated as 0.600. The similarity 1 is a similarity generally used when quantitatively evaluating the similarity between a certain waveform and another waveform, and represents the similarity when both waveforms are viewed as a whole. Therefore, the value of similarity 1 decreases if a certain peak portion is very well matched but many other portions are not matched, and a peak that is generally similar even if there is no very well matched peak. As the number increases, the value of the similarity 1 increases.

<Similarity 2> FIG. 8 is a flowchart showing a process of calculating the similarity 2 executed by the analysis computer 3. First, the analysis computer 3 normalizes the width and height of the peak by setting a bar having a height of 1 at the peak position with respect to the sample waveform (S610). FIG. 9 schematically shows this processing. (A) is the sample waveform data, (b) is the height 1 at the peak position.
The data is standardized by setting bars. (B)
This is data indicating the position of the peak of the sample waveform data.

Next, as shown in FIG. 8, with respect to the standard waveform data, the width of each peak is determined based on a predetermined error coefficient, and its height is normalized to 1 (S
611). FIG. 10 schematically shows this processing. (A) is standard waveform data, for example, E. coli E.
This is E. coli waveform data. (B) is an error coefficient, which is an experimental error considered to be mixed in a series of experimental procedures until an electrophoretic image is obtained, and can be obtained experimentally. The error coefficient takes an arbitrary value between 0 and 1. In the figure, it is assumed that the error is such that the DNA size is small in the middle region and larger at both ends. (C) is waveform data in which the width is adjusted by the error coefficient shown in (b) and the height is normalized to 1. Specifically, the width may be adjusted by multiplying the original width shown in FIG. (C)
Is data indicating the position of the peak of the standard waveform data and the width of the peak.

Next, as shown in FIG. 8, the sample waveform data obtained in S610 and the standard waveform data obtained in S611 are multiplied by all combinations of peaks, and both peaks are determined based on the multiplication result. It is determined whether they match (S612). FIG. 11 schematically shows this processing. Since the heights of both the sample waveform data and the standard waveform data are normalized to 1, the result of the multiplication is 1 when both peaks match exactly, and from 0 to 0 depending on the degree of matching when both peaks do not match. It will take a value between 1. Therefore, a certain threshold value is provided, and when the threshold value is equal to or larger than the threshold value, the two peaks match. The result of the match / mismatch is stored in the RAM of the analysis computer 3 as a flag, for example.
350.

After judging the coincidence / non-coincidence of each peak, as shown in FIG. 8, only the peak portions that match the original sample waveform data and the standard waveform data are extracted (S613). This extraction can be performed based on the match / mismatch for each peak stored in the RAM 350 (leave the peak with the “match” flag and remove the peak with the “mismatch” flag). Then, a correlation operation is performed between the coincident peak portions to calculate the similarity 2 (S614). FIG. 12 schematically shows this processing. (A) shows a peak portion where original sample waveform data (waveform data before standardization) and original standard waveform data (waveform data before standardization) coincide with each other. It is a coincident peak portion.
(B) shows the waveform data of the matching peak portion of the sampled waveform data and the waveform data of the matching peak portion of the standard waveform data. For example, by performing a correlation operation between the matching peak portions, Similarity 2
= 0.847 is obtained. Since the similarity 2 is a result of the correlation operation between the matching peak portions, the similarity 2 increases when there is a matching peak in the entire waveform and when the matching is very good. At this time, there is no problem for the peaks that do not match. In other words, even if the number of matching peaks is small as a whole, the similarity 2 increases when a certain peak locally matches very well.

<Similarity 3> FIG. 13 is a flowchart showing a process of calculating the similarity 3 executed by the analysis computer 3. S620 to S622 correspond to S61 described above.
0 to S612, the sample waveform data and the standard waveform data are normalized, and it is determined whether or not both peaks match. After judging the match / mismatch for each peak, only the matching peak is extracted from the original sample waveform data (S623). Then, a correlation operation is performed on the matching peak portion of the sample waveform data and the entire original standard waveform data to calculate the similarity 3 (S62).
4). FIG. 14 schematically shows this processing.
By performing a correlation operation between the matching peak portion of the sample waveform data and the entire standard waveform data, for example, the similarity 3 = 0.5
82 is obtained. Since the similarity degree 3 targets only the matching peak portion of the sample waveform data and targets the entire waveform of the standard waveform data, the portion where both waveforms match very well for the matching peak portion and do not match However, the similarity increases when they are approximate to some extent.

The similarities 1 to 3 are calculated as described above, and are stored in the RAM 350 of the analysis computer 3. Then, after calculating and storing the similarities 1 to 3, the similarities 1 to 3 are again stored in the RAM 3 as shown in FIG.
50, and sequentially reads the waveform data stored in the database storage unit 4, selects the standard waveform data most similar to the sample waveform data, and outputs the result (S63).

FIG. 15 is a detailed flowchart of a process for identifying a sample based on the similarities 1 to 3. First, the analysis computer 3 compares the similarity 1 with a predetermined first threshold (S631). When the similarity 1 is equal to or more than the first threshold, the microorganism of the standard waveform data is specified as the microorganism of the sample (S632). If the similarity 1 for a plurality of standard waveform data is equal to or greater than the threshold, the microorganisms of all the standard waveform data that have exceeded the threshold are output.

On the other hand, if the similarity 1 is smaller than the predetermined threshold, it is determined whether the similarity 3 is equal to or larger than the second threshold (S633). If the similarity 3 is equal to or greater than the second threshold, the microorganism of the standard waveform data is specified as the microorganism of the sample (S632). Of course, a plurality of microorganisms may be applicable.

If it is determined that the similarity 3 is also smaller than the second threshold, it is determined whether the similarity 2 is equal to or greater than a third threshold (S634). When the similarity 2 is equal to or more than the third threshold, the microorganism of the standard waveform data is specified as the microorganism of the sample (S632). When the similarity 2 is smaller than the third threshold, that is, all the similarities 1 to 3
If any of them is smaller than the threshold value, it is determined that the waveform data of the corresponding microorganism does not exist in the database storage unit 4 and the fact is output and the process is terminated.

As described above, in the present embodiment, not only the correlation calculation of the entire sample waveform data and the entire standard waveform data is performed, but the correlation calculation of the matching portions is performed by focusing on the portions where the peaks of both waveforms match. By doing so, it is possible to eliminate the influence of noise and perform more accurate collation with the database.

In the present embodiment, all the similarities 1 to 3 are calculated, and these three similarities 1 to 3 are compared with the threshold in the order of similarity 1, similarity 3, and similarity 2. The microorganisms may be specified in a different order. For example, similarity 2
→ Similarity 3 → Similarity 1 and so on.

Further, instead of using all of the similarities 1 to 3, the microorganisms of the sample can be specified using only the similarity 2 or only the similarity 3 or only the similarities 2 and 3. . For example, only the similarity 3 is compared with the threshold value, and when the similarity 3 indicates a large value, it can be estimated that there is a high possibility that the microorganism of the sample is the same as the microorganism (standard strain) in the database. When the similarities 2 and 3 are almost the same, it can be estimated that the sample is highly likely to be the same as the standard strain. Further, when the similarity 3 is much smaller than the similarity 2, it can be estimated that there is a high possibility of the allogeneic strain.

Further, in the present embodiment, the standard waveform data is stored in the database storage unit 4 and collated with the sample waveform data. However, characteristic parameters for defining the waveform data (for example, peak position, peak value, etc.) , Peak width, peak area) and the like are stored in the database storage unit 4,
You can also check against a sample.

Hereinafter, the correlation calculation will be described in detail using the similarity 3 as an example.

FIG. 16 shows a sample (sample 1).
Is similar to the standard waveform data in the database 3
3 shows the result of performing the correlation calculation. The correlation operation uses waveform data obtained by linking the waveform data obtained by 48 types of primers to each other, and shows the result of comparing Sample 1 with 23 types of microorganisms in the database. I have. In the figure, the left column of the table represents the species (Species) of the microorganisms in the database, and the numerical value in the right column of the table represents the correlation value (PC) between Sample 1 and each microorganism in the database. The higher the correlation value, the higher the correlation value. For example, the first in the table
The correlation value with Escherichia coli K12 is 0.914, 1 in the table.
The third correlation value with Vibrio parahaemolyticus GTC40 was 0.122, and the lowest in the table was Campylobacter coli GTC258.
Is 0.000. As can be seen from this table, samples 1 and the correlation value between the Escherichia coli K12 is significantly higher than the 0.914 and other microorganisms, thus the sample 1 can be identified as Escherichia coli K12.

The correlation value takes a numerical value of 0.0 to 1.0, and it is generally possible to judge that there is a correlation if the correlation value is 0.5 or more. Are not always 0.0 even if they are completely different types, and even if they are the same type, the correlation value is not always 1.0 due to the influence of waveform noise. Therefore, it is preferable that the identification of the microorganism is performed by relative evaluation in addition to the absolute value of the correlation value. In FIG. 16, Aeromonas caviae GTC46 having the second highest correlation value
The correlation value of 5 is as low as 0.305, and the sample 1 is identified as Escherichia coli K12 by the relative evaluation.

FIG. 17 shows another sample (sample 2).
Is similar to the standard waveform data in the database 3
Are shown. In this table, the correlation value between the sample 2 and Staphylococcus aureus GTC286 significantly high as 0.882, Sample 2 Staphylococcus
aureus GTC286 . In addition,
In FIG. 17, Bacillus cereu having the second highest correlation value
s GTC 418 has a low correlation value of 0.250, which is more clearly discernible than in sample 1.

FIG. 18 shows the result of calculating the correlation of the similarity 3 with the standard waveform data in the database for still another sample (sample 3). In this table, the correlation value between Sample 3 and Escherichia coli K12 is 0.1 .
796, and a correlation value of 0.76 with Vibrio cholerae GTC37.
7 is significantly higher. Therefore, sample 3 is Escherichi
a coli K12 and Vibrio cholerae GTC37 .

When identifying microorganisms by performing a correlation operation on the similarity 3, that is, the correlation between the matching peak portion of the sample waveform data and the entire original standard waveform data,
Instead of calculating the correlation value by comparing the data values of the waveform data, it is also possible to use the area or the integral value of the waveform data. Specifically, aj is the area of the waveform peak that matches the sample on the database side, and bj is the area of all the peaks of the waveform on the database side.

(Equation 4) To calculate the area ratio. Here, k is the number of coincident peaks on the database side, and m is the total number of peaks on the database side. It can be understood that the higher the correlation between the sample waveform and the database waveform, the larger the area of the molecule, and therefore the degree of correlation between the two waveform data can be evaluated by the area ratio. By calculating the area ratio, it is expected that a difference due to noise in both waveform data is absorbed, and a correlation value excellent in noise resistance is obtained.

FIG. 19 shows the area ratio (AR) of the sample 1 described above with respect to the standard waveform in the database. For example, the area ratio with Escherichia coli K12 is 0.866, and the second Campylobacter jejuni GTC
The area ratio with 259 is 0.237. When comparing the correlation value shown in FIG. 16 with the first and second values, the first value is: Escherichia coli K12 : 0.914 (correlation value)
→ 0.866 (area ratio) 2nd: 0.305 (correlation value) → 0.237 (area ratio), and the difference between the evaluation values is larger in the area ratio. Therefore, with respect to this sample 1, the area ratio makes it easier to identify microorganisms more reliably.

FIGS. 20 and 21 show the area ratios of Sample 2 and Sample 3, respectively. FIG.
At 0, comparing the correlation values in FIG. 17 with the first and second values, the first one is: Staphylococcus aureus GTC286 : 0.882
(Correlation value) → 0.946 (area ratio) 2nd: 0.250 (correlation value) → 0.138 (area ratio), and the area ratio has a larger difference in evaluation value, and the correlation value The microorganisms can be more reliably identified by using the area ratio than in the case of (1).

In FIG. 21, when compared with the correlation value of FIG. 18, Esherichia coli K12 : 0.796 (correlation value) → 0.8
21 (area ratio) Vibrio cholerae GTC37 : 0.767 (correlation value) → 0.
827 (area ratio) Camphylobacter fetus GTC260 : 0.576 (correlation value)
→ 0.335 (area ratio), the difference in the evaluation value is larger in the area ratio, and the microorganism can be more reliably identified by using the area ratio in the sample 3 as well.

As described above, it is possible to identify microorganisms contained in a sample by using the correlation value or the area ratio. However, as a result of conducting the experiment, the present applicant clearly identified the microorganisms in some samples. We have found that in some cases it cannot be identified. Specifically, this is a case where a sample contains a large number of microorganisms.
When more than one species of microorganisms are contained, there may be no significant difference between the microorganisms in the correlation value or the area ratio.

FIGS. 22 and 23 show the results of the correlation value and the area ratio for a certain sample (sample 4). Sample 4 actually contains 10 microorganisms, and the correlation value and area ratio of these 10 microorganisms are relatively high, but the correlation value or area of the 11th microorganism is high. The difference with the ratio is not great.

That is, regarding the correlation value, the tenth fine
The correlation value between the organism and the 11th microorganism is Clostridium perfringens GTC166 : 0.537Aeromonas caviae GTC465 : 0.480, and the difference is only 0.053. Also, the area ratio
Also, the area ratio of the 10th and 11th microorganisms isPseudomonas aeruginosa GTC2 : 0.329Campylobacter coli GTC258 : 0.310, and the difference is only 0.019.

In such a case, it becomes difficult to determine the boundary between the microorganisms contained in the sample and the microorganisms not contained, that is, where to set the threshold value, and there is a possibility of erroneous determination. Therefore, when it is assumed that the number of microorganism species contained in the sample is large, it is preferable to further devise the correlation calculation.

For example, when there is no large difference between the correlation values,
Since it is considered that there is no significant difference between the standard waveforms in the database, it is possible to perform a process of comparing the standard waveform data with each other, removing the matching peaks from the standard waveform data, and leaving only different independent peaks. Conceivable. The waveform data obtained by removing the coincident peaks in this manner is considered to be waveform data unique to the type, and it is expected that the difference will be clear when a correlation operation is performed with the sample waveform data. In the present embodiment, such conversion of the waveform data in the database into waveform data having only mutually independent peaks is referred to as “orthogonalization”. Correlation between the orthogonalized database waveform and the sample waveform is performed, and the correlation value or the area ratio is calculated to identify microorganisms contained in the sample.

FIGS. 24 and 25 are graphs showing the correlation values and the area ratios obtained as a result of performing a correlation operation on the sample 4 with the orthogonalized database. In both figures, the horizontal axis is the number indicating the microorganism in the database, and is the number according to the order of FIG. 22 and FIG. 23, that is, in FIG. 24, 1 is Plesio.
monas shigelloides GTC410 , 2 is Pasteurella multoci
da GTC1698 , 3 is Serratia marcescens GTC135 ..., and in FIG. 25, 1 is Plesiomonas shigelloides
GTC410 , 2 for Paeteurella multocida GTC 1698 , 3 for S
erratia marcescens GTC135 ... In both figures, in addition to the orthogonalized database, a non-orthogonal database (Norma)
The results using l) are also shown. 24 in FIG.
Focusing on the third microorganism, Clostridium perfringens GTC166 and the eleventh microorganism Aeromonas caviae GTC465 , the correlation value with the tenth microorganism increases when the orthogonalized database is used, while the eleventh correlation value decreases, It can be seen that the difference between the values has increased. FIG.
The same applies to the case of the area ratio.

FIGS. 26 and 27 show FIGS. 24 and 25, respectively.
The correlation value and the area ratio when the results of (1) and (2) are included in the microorganisms (In) and the microorganisms (Out) not included in the sample are shown. As described above, the difference between In and Out is larger for both the correlation value and the area ratio when orthogonalization is performed than when no orthogonalization is performed. By removing the matching peak in the database, the contribution of the matching peak to the correlation value can be removed, so that it can be interpreted that the difference is enlarged. Therefore, when the microorganisms cannot be identified by using the correlation value or the area ratio, the possibility of identifying the microorganisms of the sample increases by orthogonalizing the database and calculating the correlation value or the area ratio.

The embodiments of the present invention have been described above, but the present invention is not limited to these embodiments, and various modifications can be made. For example, in the present embodiment, when performing a correlation operation between the sample waveform data and the standard waveform data in the database, the waveform data is quantized for each DNA size having a predetermined width, and the luminance value is assigned for each quantized DNA size, and a matrix is assigned. It is also possible to perform the correlation operation while holding the data as a matrix (1 × N matrix). For example, FIG.
As shown in FIG.
It is assumed that the microorganisms specified by 1, 23, 59, 72, and 93 showed waveform data as shown in the figure. The horizontal axis is the DNA size (log scale), and the vertical axis is the luminance value. As shown in FIG. 29, such waveform data is quantized into a DNA size (logarithmic scale) with a predetermined width, and a matrix is formed by assigning an average luminance value within the size to the quantized size. In the drawing, the magnitude of the luminance value of each matrix is indicated by shading. By standardizing the standard waveform data and the sample waveform data in the database together and performing a correlation operation between the matrices, the processing time on the computer can be reduced. In the case of matrixing, in addition to matrixing the waveform data after waveform shaping by Gaussian distribution, raw waveform data without waveform shaping can be directly matrixed and used.

[0101]

As described above, according to the present invention,
Microorganisms can be accurately identified.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram of a microorganism identification device according to an embodiment.

FIG. 2 is a configuration block diagram of an analysis computer in FIG. 1;

FIG. 3 is an overall processing flowchart of the embodiment.

FIG. 4 is a flowchart of a waveform data correction process.

FIG. 5 is an explanatory diagram of a waveform data correction process.

FIG. 6 is a flowchart of a database search process.

FIG. 7 is an explanatory diagram of calculating a similarity 1;

FIG. 8 is a flowchart of a process for calculating a similarity 2;

FIG. 9 is an explanatory diagram of normalization of sample waveform data.

FIG. 10 is an explanatory diagram of standardization of standard waveform data.

FIG. 11 is an explanatory diagram of peak coincidence / mismatch determination.

FIG. 12 is an explanatory diagram of calculating a similarity 2;

FIG. 13 is a flowchart of a calculation process of a similarity 3;

FIG. 14 is an explanatory diagram of calculating a similarity 3;

FIG. 15 is a flowchart of a sample estimation process.

FIG. 16 is an explanatory diagram of a correlation value of sample 1;

17 is an explanatory diagram of a correlation value of Sample 2. FIG.

18 is an explanatory diagram of a correlation value of Sample 3. FIG.

19 is an explanatory diagram of an area ratio of Sample 1. FIG.

20 is an explanatory diagram of an area ratio of a sample 2. FIG.

21 is an explanatory diagram of an area ratio of a sample 3. FIG.

FIG. 22 is an explanatory diagram of a correlation value of sample 4.

FIG. 23 is an explanatory diagram of an area ratio of a sample 4;

FIG. 24 is a graph showing a correlation value between the non-orthogonalization database and the orthogonalization database.

FIG. 25 is a graph showing the area ratio between the non-orthogonalized database and the orthogonalized database.

FIG. 26 is an explanatory diagram of a correlation value difference between the non-orthogonalization database and the orthogonalization database.

FIG. 27 is an explanatory diagram of the area ratio difference between the non-orthogonalized database and the orthogonalized database.

FIG. 28 is an explanatory diagram of waveform data in a certain primer.

FIG. 29 is an explanatory diagram in which the waveform data of FIG. 28 is matrixed.

[Explanation of symbols]

1 SSC-PCR amplification / analysis device, 2 electrophoresis image input unit, 3 analysis computer, 4 database storage unit, 5 result display means.

Continued on the front page F term (reference) 4B024 AA11 CA01 GA19 HA11 4B029 AA07 AA23 BB01 BB20 CC01 FA15 4B063 QA01 QA18 QQ05 QQ42 QR32 QR38 QS16 QS17 QS20 QS25 QS36 QS40 QX01

Claims (30)

[Claims] An apparatus for identifying microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting means for extracting a portion of the sample waveform data that matches the standard waveform data, and calculating a correlation between the sample waveform data and the standard waveform data using a portion of the sample waveform data that matches. A microorganism identification device, comprising: 2. The microorganism identifying apparatus according to claim 1, wherein the calculating means calculates a correlation between a portion where the sample waveform data matches and the standard waveform data. 3. The microorganism identifying apparatus according to claim 1, wherein said calculating means calculates a correlation between a matching part of said sample waveform data and a matching part of said standard waveform data. 4. An apparatus for identifying microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. An extracting unit for extracting a portion of the sample waveform data that matches the standard waveform data; a first arithmetic unit for calculating a correlation between the sample waveform data and the standard waveform data; and the sample waveform data. And a second calculating means for calculating a correlation between the matched part of the reference waveform data and the standard waveform data, wherein the microorganism is collated based on the correlation by the first calculating means and the correlation by the second calculating means. Identification device. 5. An apparatus for identifying microorganisms by comparing sample waveform data indicating a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting means for extracting a portion of the sample waveform data that matches the standard waveform data; first calculating means for calculating a correlation between the sample waveform data and the standard waveform data; and the sample waveform data. And a second calculating means for calculating a correlation between a matching part of the standard waveform data and a matching part of the standard waveform data, and performing collation based on the correlation by the first calculating means and the correlation by the second calculating means. Characteristic microorganism identification device. 6. An apparatus for identifying a microorganism by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting means for extracting a portion of the sample waveform data that matches the standard waveform data; first calculating means for calculating a correlation between the sample waveform data and the standard waveform data; and the sample waveform data. A second calculating means for calculating a correlation between a matching part of the sample waveform data and the standard waveform data; and a third calculating means for calculating a correlation between a matching part of the sample waveform data and a matching part of the standard waveform data. And checking based on the correlation by the first calculating means, the correlation by the second calculating means, and the correlation by the third calculating means. Microbial identification device, characterized in that. 7. The apparatus according to claim 1, wherein the extraction unit normalizes at least one of the sample waveform data and the standard waveform data with a predetermined width and a predetermined height. Means for determining a match based on a result of multiplication of the standardized waveform data and other waveform data. 8. The apparatus according to claim 1, wherein at least one of the sample waveform data and the standard waveform data is shaped by a predetermined normalization function. Microorganism identification device. 9. A method for identifying microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting a portion of the sample waveform data that matches the standard waveform data, and calculating a similarity between the sample waveform data and the standard waveform data using the portion of the sample waveform data that matches. And B. a microorganism identification method, comprising: 10. The method according to claim 9, wherein, in the calculating step, a similarity between a portion where the sample waveform data matches and the standard waveform data is calculated. 11. The method according to claim 9, wherein in the calculating step, a similarity between a matching part of the sample waveform data and a matching part of the standard waveform data is calculated. . 12. A method for identifying microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting a portion of the sample waveform data that matches the standard waveform data; calculating a similarity between the sample waveform data and the standard waveform data; and matching the sample waveform data. Calculating a similarity between the portion and the standard waveform data, and performing collation based on a plurality of similarities. 13. A method for identifying a microorganism by collating sample waveform data showing a concentration distribution depending on the size of a sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting a portion of the sample waveform data that matches the standard waveform data; calculating a similarity between the sample waveform data and the standard waveform data; and matching the sample waveform data. Calculating a similarity between a portion and a portion where the standard waveform data matches, and performing a collation based on a plurality of similarities. 14. A method for identifying microorganisms by comparing sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database. Extracting a portion of the sample waveform data that matches the standard waveform data; calculating a similarity between the sample waveform data and the standard waveform data; and matching the sample waveform data. Calculating a similarity between a portion and the standard waveform data; and calculating a similarity between a matching portion of the sample waveform data and a matching portion of the standard waveform data. A microorganism identification method characterized in that collation is performed based on the information. 15. The method according to claim 9, wherein the extracting comprises normalizing at least one of the sample waveform data and the standard waveform data with a predetermined width and a predetermined height. And determining a match based on a result of multiplication of the normalized waveform data and other waveform data. 16. The method according to claim 9, wherein at least one of the sample waveform data and the standard waveform data is shaped by a predetermined normalization function. Microorganism identification method. 17. A method for comparing microorganisms by collating sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers with standard waveform data stored in a database using a computer. A program for causing the computer to determine at least a match between the sample waveform data and the standard waveform data, and A program causing the processor to execute a correlation operation for a part of the correlation operation, and causing the processor to compare a result of the correlation operation with a predetermined threshold value stored in a memory. 18. A method for comparing microorganisms by comparing, using a computer, sample waveform data showing a concentration distribution depending on the size of sample DNA amplified using each of a plurality of primers and standard waveform data stored in a database. A program for causing the computer to determine at least a match between the sample waveform data and the standard waveform data, and causing the processor to perform a correlation operation between the sample waveform and the standard waveform data. Causing the processor to execute a correlation operation of all or part of the matching portion of the sample waveform data and the standard waveform data and storing the result in the memory, and storing a plurality of correlation operation results and the memory Causing the processor to compare with a predetermined threshold stored in the processor Program which is characterized. 19. The apparatus according to claim 1, wherein said calculating means calculates said correlation using an area of a portion where said sample waveform data and said standard waveform data coincide with each other. Microorganism identification device. 20. The apparatus according to claim 19, wherein the calculating means calculates, as the correlation, an area of a peak that matches a peak of the sample waveform data among all peaks of the standard waveform data, and A microorganism identification device for calculating the ratio of the area of all peaks. 21. The apparatus according to claim 1, wherein the calculating means calculates the correlation using orthogonalized standard waveform data obtained by removing a coincident portion between the standard waveform data. A microorganism identification device characterized by performing: 22. The apparatus according to claim 1, wherein the calculating unit converts the sampled waveform data into a predetermined DN.
A microorganism identification apparatus, wherein the correlation is calculated using data quantized with an A size width and data obtained by quantizing the standard waveform data with a predetermined DNA size width. 23. The method according to claim 9, wherein in the step of calculating the similarity, the similarity is calculated using an area of a portion where the sample waveform data and the standard waveform data match. A method for identifying microorganisms, comprising calculating. 24. The method according to claim 23, wherein, in the step of calculating the similarity, an area of a peak corresponding to a peak of the sample waveform data among all peaks of the standard waveform data, A method for identifying microorganisms, comprising calculating the ratio of the area of all peaks. 25. The method according to claim 9, wherein in the step of calculating the similarity, orthogonalized standard waveform data obtained by removing a matching portion between the standard waveform data is used. A method for identifying microorganisms, comprising calculating. 26. The method according to claim 9, wherein in the step of calculating the similarity, data obtained by quantizing the sample waveform data with a predetermined DNA size width and the standard waveform data are used. A method for identifying a microorganism, wherein the operation is performed using data quantized by the predetermined DNA size width. 27. The program according to claim 17, wherein the correlation operation is executed using an area of a portion where the sample waveform data and the standard waveform data match. 28. The program according to claim 27, wherein, in the correlation calculation, an area of a peak that matches a peak of the sample waveform data among all peaks of the standard waveform data, and an area of all peaks of the standard waveform data A program executed by calculating the ratio of 29. The program according to claim 17, wherein the correlation operation is performed using orthogonalized standard waveform data obtained by removing a coincident portion between the standard waveform data. 30. The program according to claim 17, wherein the correlation calculation includes converting the sampled waveform data to a predetermined DN.
A program which is executed using data quantized by an A size width and data obtained by quantizing the standard waveform data by a predetermined DNA size width.