JP5065694B2 - Method and system for evaluating genotyping results - Google Patents

Description

  The present invention relates to an evaluation method and an evaluation system for genotype determination results for an analysis operation for determining genotypes that are considered to be involved in differences between individuals of organisms (differences in form, disease susceptibility, etc.) In particular, the present invention relates to a method and system for evaluating a result of determining a genotype signal from a noise signal when a DNA fragment containing a gene to be analyzed is amplified by PCR and detected by electrophoresis.

  Complete decoding of the entire genome of various organisms including humans is underway. With regard to living organisms such as humans that have already been decoded, gene analysis studies on the entire genome and a relatively wide area are being actively conducted. In particular, in medical research, a technique for automatically determining a large number of genotypes is attracting attention in searching for genes involved in the presence or absence of diseases, the effects on drugs, and the presence or absence of side effects. A genotype evaluation technique automatically determined for each individual is eagerly desired in order to further increase the determination accuracy.

Microsatellite Usually, genomes between individuals of the same species have the same base sequence in many parts, but it is known that they have different base sequences in several places. Such a difference in the base sequence on the genome between individuals is called polymorphism. It is also known that there are several types of polymorphisms, and single nucleotide polymorphisms such as SNPs (Single Nucleotide Polymorphisms) and microsatellite have attracted particular attention for use in analytical research.

  A microsatellite is a sequence in which a short sequence pattern of 2 to 6 bases is repeated several to several tens of times. In the case of the human genome, there are tens of thousands or more. An example of microsatellite appearing on the genome is shown in FIG. The repeating unit in microsatellite is called unit, and the base number of unit is called unit length. For example, in the microsatellite “ATATATAT ...” shown in FIG. 18, the unit is “AT” and the unit length is 2 bases. As shown in FIG. 18, even if the microsatellite has the same unit and unit length, a difference (polymorphism) is observed in the number of repetitions.

  As described above, since SNPs and microsatellite have polymorphisms, they are easily distinguishable from other base sequences on the genome, and can be easily detected experimentally. In addition, depending on the species, the approximate position where the SNPs and microsatellite exist on the genome is known, and therefore it can be used as an index indicating the position on the genome. Because of these properties, SNPs and microsatellite are called DNA markers. In particular, since microsatellite includes a plurality of bases, it has a larger amount of information than SNPs and is frequently used as a DNA marker for genome-wide analysis studies.

  By the way, as shown in FIG. 18, many organism individuals have a pair of genomes (homologous chromosomes) derived from a female gamete and a male gamete. Genes present at mutually corresponding sites on a pair of genomes are called alleles, and combinations thereof are called genotypes. As described above, since SNPs and microsatellite on the genome are portions where the nucleotide sequences can differ between individuals, generally, there are two or three alleles in SNPs, and several types of microsatellite There are ~ 20 or more alleles.

  In the example shown in FIG. 18, the individual A has a unit “AT” repeated 3 times and 5 times, and the individual B has a unit “AT” repeated 6 times and 3 Have repeated times. The individual C has two units obtained by repeating the unit “AT” four times. A state having one different type of allele, such as individual A or individual B, is called heterozygous, and a state having two alleles of the same type, such as individual C, is called homozygous.

PCR and Electrophoresis Experiments When microsatellite is used as a DNA marker, an experiment such as PCR (Polymerase Chain Reaction) or electrophoresis is performed as an experiment for extracting and detecting a portion where microsatellite appears on the genome. In PCR, a DNA fragment containing a microsatellite portion sandwiched between a pair of primer sequences is repeatedly replicated by reacting with a pair of base sequences called primer sequences at both ends of the microsatellite using a DNA replication enzyme. It is an experimental technique to obtain a sample with a constant yield. Electrophoresis includes methods such as gel electrophoresis and capillary electrophoresis, where amplified DNA fragments are run on a charged migration path, and DNA fragments of different lengths are used for differences in the degree of migration due to molecular weight and chargeability. This is an experimental technique for separation. FIG. 19 is a schematic diagram showing an experimental procedure for amplifying a DNA fragment of a microsatellite portion by PCR and gel electrophoresis. First, a pair of primer sequences 1900 and 1901 sandwiching a target microsatellite is specified, and a genomic region 1902 including the microsatellite and the primer sequence is amplified by a PCR experiment. In the example shown in FIG. 19, two types of PCR amplification products having different lengths from each other are heterozygotes having different numbers of microsatellite repeats on two homologous chromosomes, and the lengths of the microsatellite portions are different from each other. That is, DNA fragments (52 bases and 48 bases) are obtained. When these are electrophoresed on a gel for a certain period of time, the two kinds of PCR amplification products are separated by the difference in length of their DNA fragments. Each DNA fragment is labeled with a fluorescent dye in advance, and the intensity and position of the fluorescence signal from each DNA fragment is detected after electrophoresis, so that the length of the DNA fragment (fragment Size) and a graph in which the vertical axis represents the fluorescence signal intensity (ie, the abundance of DNA fragments) is obtained. In addition, the length of each PCR amplification product can be determined based on the detection position of the size marker by simultaneously performing electrophoresis of a DNA fragment (size marker) whose length is known in advance together with the PCR amplification product and detecting a fluorescent signal. I can know.

  In addition, although the experimental method using gel electrophoresis was described above, the same thing can be performed also by capillary electrophoresis. In capillary electrophoresis, the sample is run through a thin tube packed with gel, and the time taken for each sample to finish moving a certain distance (usually up to the end of the capillary) is measured. It is a technique to check the thickness. In capillary electrophoresis, it is common to detect a sample with a fluorescence signal detector provided at the end of the capillary rather than scanning the fluorescence signal from the sample in the gel.

Noise generated in PCR and electrophoresis experiments The peak of the PCR and electrophoresis results shown in FIG. 19 above are obtained in an ideal process. A peak often occurs. There are Stutter peak and + A peak as main noise peaks in interpreting the results.

  As shown in FIG. 20, the Stutter peak forms a complementary strand at the position where the template sequence strand to be replicated is shifted from the continuous repeating sequence of microsatellite during PCR, and the template strand has a hairpin loop shape. Due to the phenomenon of slipping-strand mispairing, the number of repeats of the microsatellite portion of the DNA fragment to be replicated increases or decreases, and the DNA fragment whose number of repeats increases or decreases in the fluorescence signal is a noise peak. As observed. It is also known that it tends to occur particularly when a microsatellite with a short unit length is amplified. The Stutter peak is observed as a DNA fragment peak whose length is increased or decreased by an integral multiple of the microsatellite unit length in addition to the DNA fragment having the same length as the original DNA fragment.

  The + A peak indicates that when a DNA fragment is replicated by PCR, one extra base (usually A) is added to the DNA fragment due to the action of the replication enzyme. It is observed as a noise peak of the length of the DNA fragment. Such addition of one base occurs not only with respect to the DNA fragment of the replication source but also with respect to each DNA fragment that is the origin of the above Stutter peak. A peak is observed.

  FIG. 21 shows a schematic diagram in which the above Stutter peak and + A peak are observed. FIG. 21 shows the waveform of a heterojunction in which two alleles are present. It includes two peaks (hereinafter referred to as “true peaks”) corresponding to the allyl size of the same length as the DNA fragment length of the replication source, and consists of a group of two peaks centered on each. In the first peak cluster, there are Stutter peaks at the position of 2 units left of the true peak, the position of 1 unit of left, and the position of 1 unit of right, corresponding to each of the true peak and the Stutter peak + A peak is present. In the second peak cluster, there is a Stutter peak at the position of 1 unit left and 1 unit of right of the true peak, and + A peak corresponding to each of the true peak and Stutter peak. . In the following, for a certain + A peak, a true peak or a Stutter peak corresponding to a DNA fragment to which the + A peak is generated and from which one base has not been added is referred to as an “original peak”.

  Regarding the waveform of the fluorescence signal obtained through the PCR and electrophoresis experimental process, several methods have already been reported for each individual to determine the true peak from multiple peaks that contain noise. It is disclosed in Document 1.

Several methods for evaluating the genotype determination results have already been reported.
Methods for evaluating genotype determination results are disclosed in Patent Literature 1, Non-Patent Literature 1, and the like. As software having a function of evaluating the genotype determination result, software “TrueAllele” of Cybergenetics, software “GeneMapperID” of ABI, etc. are known.

JP 2006-17461 A Matsumoto T et al., "Novel algorithm for automated genotyping of microsatellites", Nucleic Acids Research, Vol.32, No.20 (2004) p6069-6077

  In a method for automatically determining a genotype, a method for further evaluating the automatic determination result is required. This is because when a researcher actually interprets the results of automatic determination, it is only necessary to check the results of automatic determination together with the determination accuracy obtained by evaluating them. This is because it is possible to determine whether or not the result of determining the genotype is effective.

  Further, Patent Document 1 has already devised a method of using information obtained by calculating the appearance and characteristics of Stutter peaks and + A peaks in a population used for the same marker in determining a true peak. However, there is a concern that the determination accuracy may be lowered when a sufficient number of individuals are not used in one process. In the method of using the information obtained by calculating the appearance and characteristics of Stutter peak and + A peak in the population included in the same marker, specifically, the unit is centered on the true peak included in the waveform of each individual Each linear regression line of the height ratio of the original peak and the + A peak at a position that is a multiple of the length is calculated in the same manner as described in Non-Patent Document 1. Using the linear regression line, it is determined whether each peak included in a certain observed waveform is a true peak, a Stutter peak, or a + A peak. However, if the number of individuals used to calculate the linear regression line is not sufficient, the influence of the waveform fluctuation of some individuals will increase, and a linear regression line representing the individual group cannot be calculated. There is a concern that the peak determination result of the observed waveform using the linear regression line will also be incorrect. However, since the number of individuals used in one treatment depends on the number of samples used in one experiment, it is difficult to control the number of individuals used in one treatment.

  The present invention has been made in view of such circumstances, and even when there are few individuals that can be used in one processing of a marker, sufficient information on the characteristics of the Stutter peak and + A peak of the marker is available. And to provide a method and system for evaluating the result of automatic genotype determination.

Based on the technical idea of the invention disclosed in Patent Document 1, the present inventor has further considered the following and arrived at the means for solving the above problems.
First, we focused on the following characteristics regarding the height ratio and fragment length of the Stutter peak and + A peak obtained for a marker.

-Feature 1 The ratio of Stutter peak height to true peak is reproducible.
For the Stutter peak of the fluorescence signal, the absolute value of the height of each peak fluctuates from one experimental plate to another and every experimental opportunity, but there is no reproducibility, but when considering the same allele of the same marker as described below, it is true The ratio of Stutter peak height to peak is reproducible. The Stutter peak is relative to the true allele peak when amplifying DNA fragments of the same allele fragment length with the same marker because the mechanism by which it occurs is due to a phenomenon relative to the true allele peak. Stutter peaks occur to the same extent (at the same height ratio). For example, in FIG. 1, the height ratio of the Stutter peak at the position of 1 unit to the left of the true peak in the lump of the second peak in the waveform of both the first and second individuals (the height of 101 to 100) The height ratio and the height ratio of 103 to 102 are almost equal.

Feature 2: The ratio of the height of the + A peak to the true peak is reproducible when the experimental protocol including the action time of the replicative enzyme is the same.
Similarly for the + A peak, when considering the same marker, the ratio of the height of the + A peak to the true peak is reproducible. The + A peak is similar to the Stutter peak in that it occurs relative to the original peak (true peak or Stutter peak), but the length of time that the replication enzyme acts on the degree to which the + A peak is generated. Is known to have a strong influence. In the case of amplifying DNA fragments related to the same allele of the same marker, the experimental protocol is generally fixed, and the time for the enzyme to act (the time until inactivation) is considered to be constant. Again, reproducibility can be expected.

  For example, in FIG. 1, in the waveform of both the first and second individuals, the height ratio of the + A peak to the true peak (104 high to 100 high) in both the first and second peak masses. The height ratio, the height ratio of 106 to 105, the height ratio of 108 to 107, and the height ratio of 109 to 102) are approximately equal.

Feature 3 Fragment lengths that can be true peaks, Stutter peaks, and + A peaks are often known.
When genotyping a certain marker, the allyl type that can be the marker is often well examined and known. Therefore, the fragment length that is an integral multiple of the unit length on the left and right is the fragment length that can be used as the Stutter peak, based on the fragment length that can be used as the allyl type (true peak), and the original peak (true peak or Stutter peak) A fragment length obtained by adding one base to the fragment length is a fragment length that can be a + A peak. For example, in a marker whose unit length is 2 bases, there can be 44 bases as a true peak, 44-2 = 42 bases, 44 + 2 = 46 bases, etc. are fragment lengths that can be Stutter peaks, 42 + 1 = 43 bases, 44 + 1 = 45 bases, 46 + 1 = 47 bases, and the like are possible fragment lengths as + A peaks.

  Therefore, the present inventor pays attention to the characteristics of the above three points, and is sufficient for the characteristics of the Stutter peak and the + A peak of the marker even if there is a small number of individuals used in the single processing. The inventors have come up with the idea of realizing a method and system for obtaining information with the following functions. The user / operator of this system is hereinafter referred to as a “user”. The genotype is determined by calculating the ratio of the height to the true peak and the height ratio of the Stutter peak and the + A peak, and the tendency of the appearance of the Stutter peak and + A peak to the true peak. Adopt the method used for

Function 1-1: Database expansion function regarding the ratio of Stutter peak height to true peak Considering the reproducibility of Stutter peak height ratio to true peak, the waveform information of each individual population is added to the database. By doing so, the more the number of individuals that are processed, the more abundant number of individuals can be used as statistically stable information on the appearance of Stutter peaks and the characteristic height ratio information. System. However, instead of registering all height ratios of all populations every time as they are, detection of outliers in the population used for the processing at that time and all data stored in the database In order to construct a database that accumulates statistically more stable data, it is necessary to detect outliers and filter the data to be additionally registered. Filter by two kinds of verification.

  The first filtering is performed by verifying the variance value of the entire height ratio of each individual group. It is assumed that the user can define a threshold value for the variance value of the entire height ratio of the individual group. Using this threshold value, it is verified whether the variance of the height ratio of the individual group is equal to or less than the threshold value. When the variance value is less than or equal to the threshold value, all the height ratios are additionally registered. When the variance value is greater than the threshold value, this is displayed (step 603 in FIG. 6 described later), and the height ratio is set. No additional registration shall be made. (As will be described later, FIG. 14 is a display example indicating that the variance value of all the height ratios of the population is larger than the threshold value.)

  The second filtering focuses on the average value of the individual height ratios of each individual group, and in what range the average value of the individual height ratios is relative to the standard deviation of all data. This is done by verifying. The total data here refers to all data stored in the database and all the individual populations at this time, and the average value and standard deviation of the height ratio of all the data are obtained (the figure to be described later). 6 step 602). Only when the average value of the height ratio of the observed waveform of each individual is in the range of (average value of all data) ± 2 × (standard deviation of all data), the height ratio of that individual is additionally registered. If a certain height ratio is outside the range, no additional registration is made in the database.

  In the filtering based on the above two verifications, those that are validated are added to the database. With this function, it is possible to expand the height ratio database using only valid determination results.

  Here, a 95% confidence interval based on the average value of the total height ratio and the standard deviation is used to determine whether or not the observed waveform is an outlier, and to perform filtering. This is not the only way to choose.

Function 1-2: Outlier detection function regarding the ratio of the height of the Stutter peak to the true peak In the function 1-1, there are two types that were not validated in the first or second filtering. A warning is displayed as an outlier with the verification result. This function makes it possible to check whether the genotype determination result of each population is valid.

Function 2-1: Database expansion function for the ratio of the height of + A peak to the true peak Considering the reproducibility of the height ratio of the + A peak to the true peak, the waveform information of each individual population is database By adding to the above, the more the number of individuals is processed, the more stable the statistical information about how the + A peak appears and the characteristic height ratio. The system can be used as information. However, instead of registering all height ratios of all populations every time as they are, detection of outliers in the population used for the processing at that time and all data stored in the database In order to construct a database that accumulates statistically more stable data, it is necessary to detect outliers and filter the data to be additionally registered. Filter by two kinds of verification.

  The first filtering is performed by verifying the variance value of the entire height ratio of each individual group. It is assumed that the user can define a threshold value for the variance value of the entire height ratio of the individual group. Using this threshold value, it is verified whether the variance of the height ratio of the individual group is equal to or less than the threshold value. If the variance value is less than or equal to the threshold value, all of those height ratios are additionally registered. If the variance value is greater than the threshold value, a message to that effect is displayed (step 703 in FIG. 7 described later), and the height ratios are displayed. No additional registration is made. (As will be described later, FIG. 15 is a display example showing that the variance value of all the height ratios of the population is larger than the threshold value.)

  The second filtering focuses on the average value of the height ratio of each individual in each population, and in what range the average value of the height ratio of each individual observation waveform is relative to the standard deviation of all data. This is done by verifying that it exists. The total data here refers to all data stored in the database and all the individual populations at this time, and the average value and standard deviation of the height ratio of all the data are obtained (the figure to be described later). 7 702). Only when the average value of the height ratio of each individual is in the range of (average value of all data) ± 2 × (standard deviation of all data), the height ratio of that individual is additionally registered. If a certain height ratio is outside the range, no additional registration is made in the database.

  In the filtering based on the above two verifications, those that are validated are added to the database. With this function, it is possible to expand the height ratio database using only valid determination results.

  Here, a 95% confidence interval based on the average value of the total height ratio and the standard deviation is used to determine whether or not the observed waveform is an outlier, and to perform filtering. This is not the only way to choose.

Function 2-2: Outlier detection function regarding the ratio of the height of the + A peak to the true peak In the function 2-1, there are two types that were not validated in the first or second filtering. A warning is displayed as an outlier along with the verification result. This function makes it possible to check whether the genotype determination result of each population is valid.

Function 3-1: Expanded database function by adding fragment length value information of each time population. Considering that fragment lengths that can be true peaks, Stutter peaks, and + A peaks are often known. Of the fragment length values of the peaks detected in the individual group processed for the marker, those determined to be valid are stored in the database. By doing so, by examining whether the fragment length value of the peak detected in an individual in each round is valid, whether it is within the range of the peak fragment length value that can be detected in the same marker stored in the database Verification is possible (step 806 in FIG. 8 described later).

  The comparison between the peak information obtained for each individual (individual to be verified) and the peak information stored in the database is stored in the database with the same allele data more than the number of records predefined by the user. Shall be done if

  Start with the true peak. If true peaks match, verify the left and right Stutter peaks and + A peaks. If fragment length information of a certain Stutter peak or + A peak is stored in the database but not detected in the individual to be verified, conversely, the fragment length information is not stored in the database, but the individual to be verified If detected, the system displays a warning to the user. Especially when the fragment length information of a certain Stutter peak or + A peak is not stored in the database but is detected in the individual to be verified, not only a warning that it is not stored in the database is displayed, Compared with the unit length of the marker and the fragment length of the true peak, as described in Feature 3, whether it is the fragment length value of the “possible” peak compared to “knowledge of the cause of the peak” The verification result information is also displayed along with the warning display. The “knowledge of the cause of the peak” here means that the Stutter peak occurs at a fragment length value that is an integral multiple of the unit length to the left and right based on the fragment length value of the true peak, and the + A peak is the original peak It refers to the finding that it occurs in the fragment length value obtained by adding one base to the fragment length value of (true peak or Stutter peak).

  In the above verification, those validated are added to the database. With this function, it is possible to expand a fragment length database using only valid determination results.

Function 3-2: Detection of outliers by addition of fragment length value information for each time population In the case of function 3-1 which is not valid, a warning is displayed as an outlier along with the verification result. This function makes it possible to check whether the genotype determination result of each population is valid.

  The present invention provides the following evaluation system for genotype determination results as a form of realizing the various functions as described above.

  A system for displaying the result of analyzing the length of a PCR amplification product of a DNA fragment containing microsatellite, and displaying the detection signal of the PCR amplification product in a graph with the detection signal intensity and the fragment length as axes. And a first determination process for determining a + A peak corresponding to the detection signal of the PCR amplification product in which one adenine is added to the end of the DNA fragment and a peak other than the + A peak in the detection signal of the PCR amplification product In the detection signal of the PCR amplification product, the true peak corresponding to the detection signal of the PCR amplification product of the DNA fragment, and the detection signal of the PCR amplification product in which the repetitive sequence of the microsatellite is increased or decreased by 1 unit or more. A second determination processing unit for determining the corresponding Stutter peak, the + A peak and the + A peak And a determination result display processing unit for displaying the determination result of the true peak and the Stutter peak together with the graph, and further, for a plurality of individuals, a DNA fragment containing microsatellite It has a database that accumulates the results of analyzing the length of PCR amplification products, and the determination result by the first determination processing unit and the second determination processing unit is based on at least one of the following criteria A system characterized in that a determination result is evaluated based on the evaluation result.

(1) Is the height ratio between the determined true peak and the Stutter peak not significantly different from the same ratio for a plurality of individuals accumulated in the database?
(2) Is the height ratio between the determined true peak and + A peak not significantly different from the same ratio for a plurality of individuals accumulated in the database?
(3) Are the determined true peak, Stutter peak, and + A peak fragment lengths not significantly different from the same length for multiple individuals accumulated in the database?

  In the evaluation system for genotyping results of the present invention, the database further stores an experimental protocol when the analysis is performed together with the analysis results for each individual. In the evaluation of the determination results, the database Of the data stored in the database, only data for which the determination target and the experimental protocol match to a certain extent are used as the determination criterion.

  In the evaluation system for genotype determination results of the present invention, when the determination result is evaluated to be valid in the evaluation of the determination result, the analysis result of the determination target is stored in the database.

  As described above, according to the evaluation method and evaluation system of the genotyping result of the present invention, the true peak is determined from the noise peak of the Stutter peak or + A peak for the graph showing the fluorescence analysis result of the PCR amplification product. If there is enough data processed with the same allele of the same marker in the past, even if the number of individuals used in a single process is not large enough, high-quality information on noise peak feature information It is possible to obtain information on whether the determination result of the population and the genotype used in the processing at that time are valid or invalid (whether they are outliers). As a result, the genotype determination process for each individual group can be performed with high accuracy without adding extra experiment and processing costs even if the number of individuals is small.

  Hereinafter, the best mode for carrying out the genotyping result evaluation method and evaluation system of the present invention will be described in detail with reference to the accompanying drawings. 2 to 17 are diagrams illustrating embodiments of the present invention. In these drawings, the same reference numerals denote the same components, and the basic configuration and operation are the same. To do.

System Configuration FIG. 2 is a functional block diagram schematically showing the internal configuration of the genotype determination result evaluation system constructed as an embodiment of the present invention. This evaluation system for genotyping results includes a waveform data DB 200 storing waveform data (waveform data of a target population) obtained as a result of fluorescence analysis of PCR amplification products after each PCR and electrophoresis experiment, Display device 201 for displaying genotyping results, keyboard 202 for performing operations such as selecting individuals and peaks for the displayed waveform data and genotyping results, and pointing device 203 such as a mouse, necessary A central processing unit 204 that performs various arithmetic processing, control processing, and the like, and a DB 205 that stores height ratios of waveform data used in processing performed in the past.

  The central processing unit 204 groups a peak appearing in the waveform data into a pair of the original peak and the + A peak in the genotyping process, and a + A peak separation processing unit 206, and the original peak in the genotyping process. A true peak separation processing unit 207 for determining whether a peak is a true peak or a Stutter peak, and an individual validated in the above function 1, function 2 or function 3 is added to the database and processed A warning display processing unit 208 is displayed for displaying that the individual group or each individual is an outlier for all data. The waveform data DB 200 and the DB 205 storing the height ratio of the waveform data used in the processing performed in the past are the individual waveform data 209 that holds the waveform data for each individual, and the peak data for each waveform data. Peak data 210 and experimental protocol input data 211 are included.

  FIG. 3 is a diagram showing a waveform data structure group for each individual included in the waveform data DB 200 and the DB 205 storing the height ratio of the waveform data used in the processing performed in the past. This waveform data structure group WaveFormData [] includes, for each of j individual groups, an individual ID 300 for identifying each individual, waveform data 301 (corresponding to the data shown in FIG. 4), a true peak, and its + A peak. Ratio data 302 and experimental protocol information 303 are included. The value stored in the data 302 is a null value when the calculation related to the waveform data has not yet been performed.

  FIG. 4 is a diagram showing a peak data structure group of waveform data included in the waveform data DB 200 and the DB 205 storing the height ratio of the waveform data used in the processing performed in the past. This data structure group PeakData [] includes, for k peaks, a fragment length 400 of each peak, a height 401 of each peak, whether each peak is a true peak or a true peak + A peak, or other Stutter peaks Or the data of the label 402 representing the + A peak. The data 402 includes “Selected” for a true peak, “Selected + A” for a true peak + A peak, and “Stutter” for a stutter peak that is not a true peak. “+ A” is stored in the case of the + A peak for the non-Stutter peak.

Processing Procedure by System Next, the flow of processing performed in this genotype determination result evaluation system will be described with reference to the flowcharts shown in FIG. 5, FIG. 6, FIG. 7, and FIG.

  First, the waveform data of each individual is read from the waveform data DB 200 (step 500). Here, the waveform data of all individuals for the target microsatellite marker stored in the waveform data DB 200 is read, and in the waveform data DB 200 and the DB 205 storing the height ratio in the waveform data used in the processing performed in the past. Individual waveform data 209 and peak data 210 are held. The experimental protocol input data is read and held as the experimental protocol input data 211 in the waveform data DB 200 or the DB 205 storing the height ratio of the waveform data used in the processing performed in the past. Next, the + A peak and the original peak are grouped for each individual (step 501). This process is executed by the + A peak separation processing unit 206 of the central processing unit 204, and the peak determination method uses a conventional technique. For a peak determined to be a + A peak, a value indicating that it is a + A peak is written in a peak label 402 of the peak data 210, and for a peak determined to be an original peak, a peak label 402 of the peak data 210 is written. Is written with a value indicating that it is a true peak or a Stutter peak. Further, the ratio of the height of the original peak of each group and the + A peak is written in the data 302 of the peak data 210. Further, experimental protocol input data is written to data 303.

  The result of grouping the peaks included in the waveform data of each individual in this way into the original peak and its + A peak is displayed as a waveform as shown in FIG. 9 (step 502). The display screen shown in FIG. 9 shows a result 900 obtained by grouping each peak into the original peak and + A peak for the waveform data of a certain individual, and a table 901 showing the fragment length and height ratio of each peak group. And the calculation result 902 of the dispersion value of each height ratio according to each grouping method when the highest peak is not the + A peak is displayed.

  Subsequently, for each peak determined to be a peak other than the + A peak (original peak) in Step 501, it is determined whether it is a true peak or a Stutter peak (Step 503). This processing is executed by the true peak separation processing unit 207 of the central processing unit 204, and the peak determination method uses a conventional technique. The determination result of each peak is written in the peak label 402 of the peak data 210. In addition, the height ratio between the true peak and the + A peak is calculated for each individual, and sequentially added as each element value of the data 302 of the waveform data 209 of the individual.

  The result of determining the original peak included in the waveform data of each individual as the true peak and the Stutter peak in this way is displayed in a graph as shown in FIG. The display screen shown in FIG. 10 shows the result 1000 of determining each peak as + A peak, true peak and Stutter peak for the waveform data of a certain individual, and the fragment length and height ratio of each peak set. Table 1001, calculation result 1002 of the dispersion value of each height ratio according to each grouping method when the highest peak is not + A peak, and the height ratio between the true peak and each + A peak 1003 is displayed.

  In the subsequent confirmation processing of whether or not the ratio of the true peak to the height of each Stutter peak in step 504 (the process described later shown in FIG. 6), the ratio of the height of the true peak to each Stutter peak is DB205. 6, the warning display processing unit 208 of the central processing unit 204 displays a predetermined warning (see FIG. 6). Process in step 610). An example of the warning display screen in this case is shown in FIG. The warning display screen shown in FIG. 11 shows the result of determining each peak as + A peak, true peak, and Stutter peak for the waveform data of a certain individual, and the fragment length and height ratio of each peak set. Table 1101, the calculation result 1102 of the dispersion value of each height ratio by each grouping method when the highest peak is not + A peak, and the height of the true peak and each + A peak A ratio 1103, a histogram of the height ratio between the true peak and each Stutter peak in this individual and other individuals, and a predetermined warning display 1104 are displayed.

  In the subsequent confirmation processing (step 505 shown in FIG. 7, which will be described later) for checking whether the ratio of the height of the true peak to each + A peak is appropriate in step 505, the ratio between the height of the true peak and each + A peak Is determined to be significantly different (invalid) from the corresponding value stored in the DB 205, the warning display processing unit 208 of the central processing unit 204 displays a predetermined warning (FIG. 7 in step 710). An example of the warning display screen in this case is shown in FIG. The warning display screen shown in FIG. 12 shows the result 1200 of determining each peak as + A peak, true peak and Stutter peak for the waveform data of a certain individual, and the fragment length and height ratio of each peak set. Table 1201, the calculation result 1202 of the dispersion value of each height ratio by each grouping method when the highest peak is not + A peak, and the height of the true peak and each + A peak A ratio 1203, a histogram of the height ratio between the true peak and each + A peak in this individual and other individuals, and a predetermined warning display 1204 are displayed.

  As a result of the process of confirming whether or not the fragment length value of the original peak and the + A peak in the last step 506 is valid (the process described later in FIG. 8), the fragment length value of the original peak or + A peak is DB205. 8, a predetermined warning is displayed by the warning display processing unit 208 of the central processing unit 204 (see FIG. 8). Step 810). An example of the warning display screen in this case is shown in FIG. The warning display screen shown in FIG. 13 shows the result 1300 of determining each peak as + A peak, true peak and Stutter peak for the waveform data of an individual, and the fragment length and height ratio of each peak set. Table 1301, the calculation result 1302 of the dispersion value of each height ratio according to each grouping method when the highest peak is not + A peak, and the height of the true peak and each + A peak A ratio 1303, a histogram of fragment length values of a certain peak in this individual and other individuals (a fragment length value of a true peak in this example), and a predetermined warning display 1304 are displayed.

  FIG. 6 is a flowchart showing in detail the validity confirmation processing of the height ratio between the true peak and each Stutter peak in step 504 of FIG. This flowchart shows processing for all individuals. First, the variance Vall of all height ratios included in all individuals is calculated (step 600). Then, it is determined whether or not the variance Vall of all height ratios is equal to or less than a value Vdef defined by the user (step 601). If the determination result in step 601 is No, it is displayed that the variance values of all the height ratios of the population are outliers (step 603). FIG. 14 shows an example of the dialog displayed in step 603, which consists of 1400 including a warning message and an OK button. If the determination result in step 601 is Yes, the average value Aall and the standard deviation Sall of all the height ratios stored in the population and the DB 205 are calculated (step 602).

  Subsequently, the processing described below is performed in a loop for all individuals (processing included between step 604 and step 609). First, the average value Athis of the height ratio in the waveform is calculated for each individual (step 605). Then, it is determined whether or not Athis is within the range of Aall ± 2 × Sall for Aall and Sall calculated in step 602 (step 606). If the determination result is No, it is stored as waveform data of an outlier (Step 608), and if the determination result is Yes, it is stored as valid waveform data to be additionally registered in the DB 205 (Step 607). The above is performed in a loop for all individuals, and a waveform data group of outliers and an appropriate waveform data group to be additionally registered in the DB 205 are accumulated. Finally, information on the waveform data of the outliers is displayed (step 610), and the waveform data group determined to be valid is additionally registered in the DB 205 (step 611). The screen displayed in step 610 is the screen shown in FIG. 11 as described in step 504, and this processing corresponds to function 1-2. Step 611 corresponds to function 1-1.

  Here, a 95% confidence interval based on the average value of the total height ratio and the standard deviation is used to determine whether the observed waveform is an outlier. This is not the case.

  FIG. 7 is a flowchart showing in detail the validity confirmation process of the height ratio between the true peak and each + A peak in step 505 of FIG. This flowchart shows processing for all individuals. First, the variance Vall of all height ratios included in all individuals is calculated (step 700). Then, it is determined whether or not the variance Vall of all height ratios is equal to or less than a value Vdef defined by the user (step 701). If the determination result in step 701 is No, it is displayed that the variance values of all the height ratios of the population are outliers (step 703). FIG. 15 shows an example of a dialog displayed in step 703, which is composed of 1500 including a warning message and an OK button. If the determination result in Step 701 is Yes, the average value Aall and the standard deviation Sall of all the height ratios stored in the population and the DB 205 are calculated (Step 702).

  Subsequently, the processing described below is performed in a loop for all individuals (processing included between step 704 and step 709). First, for each individual, an average value Athis of the height ratio in the waveform is calculated (step 705). Then, it is determined whether or not Athis is within the range of Aall ± 2 × Sall for Aall and Sall calculated in step 702 (step 706). If the determination result in step 706 is No, it is stored as outlier waveform data (step 708), and if the determination result in step 706 is Yes, it is stored as valid waveform data to be additionally registered in the DB 205 (step 707). ). The above is performed in a loop for all individuals, and appropriate waveform data groups for additionally registering waveform data groups of outliers in the DB 205 are accumulated. Finally, information on the waveform data of the outliers is displayed (step 710), and the waveform data group determined to be valid is additionally registered in the DB 205 (step 711). The screen displayed in step 710 is the screen shown in FIG. 12 as described in step 505, and this processing corresponds to function 2-2. Step 711 corresponds to function 2-1.

  FIG. 8 is a flowchart showing in detail the validity confirmation processing of the fragment length value of each original peak or + A peak in step 506 of FIG. This flowchart shows processing for all individuals. The processing described below is performed in a loop (processing included between step 800 and step 809). First, it is determined whether the number of waveform data (number of fragment length data) of the corresponding marker and allele registered in the DB 205 is equal to or greater than the user-defined value Ndef (step 801). If the determination result in step 801 is No, a message and a waveform indicating that the number of corresponding waveform data registered in the DB 205 is less than the defined value Ndef are displayed (step 802). FIG. 16 shows an example of the display screen in step 802, which includes a waveform data display unit 1600 and a warning message display unit 1601. At the same time, a confirmation dialog to the user as to whether to additionally register the observation waveform data in the DB 205 is displayed (step 803). FIG. 17 shows an example of a confirmation dialog displayed in step 802, which consists of 1700 including a confirmation message, a Yes button, and a No button. If the user selects additional registration (Yes) in the confirmation dialog displayed in step 803, the observed waveform is accumulated as waveform data to be stored in the database (step 804).

  On the other hand, if the determination result in Step 801 is Yes, each peak of the observed waveform is compared with each peak of the corresponding waveform data registered in the DB 205 (Step 805). Then, it is determined whether or not the observed waveform matches the fragment length value of each peak of the corresponding waveform data registered in the DB 205. In this case, there may be a peak that exists only on one side. In this case, as described in Function 3-1, the peak generation cause is checked according to whether the peak is a Stutter peak or a + A peak. The determination is made based on the validity of the peak (step 806). If the determination result is No, the waveform data of the observed waveform is accumulated as the outlier value waveform data together with the determination content information (step 808). If the determination result is Yes, the observed waveform data is accumulated as waveform data to be additionally registered in the DB 205 (step 807). The above is performed in a loop for all individuals, and a waveform data group of outliers and an appropriate waveform data group to be additionally registered in the DB 205 are accumulated. Finally, information on the waveform data of the outliers is displayed (step 810), and the waveform data group determined to be valid is additionally registered in the DB 205 (step 811). Step 810 corresponds to function 3-2. Step 811 corresponds to function 3-1.

  As mentioned above, although the specific embodiment was shown and demonstrated about the evaluation method and evaluation system of the genotype determination result of this invention, this invention is not limited to these. A person skilled in the art can make various changes and improvements to the configurations and functions of the invention according to the above-described embodiments or other embodiments without departing from the gist of the present invention.

  The genotyping result evaluation system of the present invention uses the height ratio of the true peak and the height ratio of the Stutter peak and + A peak, and the tendency of the appearance of the Stutter peak and + A peak to the true peak. The genotype determination system can be used by being mounted on, for example, a personal computer used as an experimental data analysis system.

It is a figure which shows that the ratio of the height of a Stutter peak with respect to a true peak and the ratio of the height of a + A peak with respect to a true peak are reproducible. It is a functional block diagram which shows roughly the internal structure of the evaluation system of the genotype determination result constructed | assembled as one Embodiment of this invention. It is a figure which shows the data structure of the waveform data 209 of the individual contained in DB205 which stored height ratio in the waveform data DB200 of the gene information display system shown in FIG. 2, and the waveform data used by the process performed in the past. It is a figure which shows the data structure of the peak data 210 contained in DB205 which stored the height ratio in the waveform data DB200 of the gene information display system shown in FIG.2, and the waveform data used by the process performed in the past. It is a flowchart which shows the flow of the process performed in the evaluation system of the genotype determination result shown in FIG. It is a flowchart which shows the validity confirmation process of the height ratio of a true peak and each Stutter peak in step 504 of FIG. 5 in detail. 6 is a flowchart showing in detail the validity confirmation processing of the height ratio between a true peak and each + A peak in step 505 of FIG. 5. 6 is a flowchart showing in detail a fragment length value validity confirmation process in step 506 of FIG. 5. It is a figure which shows the screen which carries out the graph display of the result which grouped each peak into + A peak and the original peak about the waveform data of each individual | organism | solid by the + A peak isolation | separation processing part. It is a figure which shows the screen which carries out the graph display of the result of having determined the original peak contained in the waveform data of each individual into the true peak and the Stutter peak by the true peak separation processing unit. When the height ratio between the true peak and each Stutter peak is significantly different from the corresponding value stored in the DB 205 (invalid), the warning display processing unit displays a predetermined warning. It is a figure which shows an example. A screen for displaying a predetermined warning by the warning display processing unit when the height ratio between the true peak and each + A peak is significantly different from the corresponding value stored in the DB 205 (invalid) It is a figure which shows the example of. Example of a screen for displaying a predetermined warning by the warning display processing unit when the fragment length value of the original peak or + A peak is significantly different (invalid) from the corresponding value stored in the DB 205 FIG. It is a figure which shows the example of the dialog which displays that the dispersion | distribution value of the ratio of the height of all the true peaks contained in a population and a Stutter peak is an outlier. It is a figure which shows the example of the dialog which displays that the dispersion | distribution value of the ratio of the height of all the true peaks contained in a population and + A peak is an outlier. It is a figure which shows the example of the screen which displays that the waveform data number in the database corresponding to an observation waveform is less than a user-defined value. It is a figure which shows the example of the dialog which confirms whether observation waveform data is registered into a database, when the number of waveform data in the database corresponding to an observation waveform is less than a user-defined value. It is a figure which shows that the polymorphism is seen in the repetition number of a microsatellite for every individual and every homologous chromosome. It is a figure which shows typically the experimental procedure which extracts and amplifies the DNA fragment of a microsatellite part by PCR and electrophoresis. It is a figure which shows the slippage phenomenon in PCR as a cause of generation of a Stutter peak. It is a figure which shows about a general waveform, about the true peak, the Stutter peak of the position left | separated by unit length to the right and left, and the + A peak of the true peak and the Stutter peak at the 1 base right position of each. This figure shows an example in which the waveform of an individual is observed as a cluster of two peaks for two true peaks.

Explanation of symbols

200 DB that stores waveform data for each target population
201 Display device 202 Keyboard 203 Pointing device 204 Central processing unit 205 DB storing waveform data processed in the past
206 + A Peak separation processing unit 207 True peak separation processing unit 208 Warning display processing unit 209 Individual waveform data 210 Peak data 211 Experimental protocol input data

Claims (3)

A system for displaying the result of analyzing the length of a PCR amplification product of a DNA fragment containing microsatellite,
A graph display processing unit for displaying the detection signal of the PCR amplification product in a graph with the detection signal intensity and the fragment length as axes;
A first determination processing unit for determining a + A peak corresponding to the detection signal of the PCR amplification product in which one adenine is added to the end of the DNA fragment and a peak other than the + A peak in the detection signal of the PCR amplification product; ,
In the detection signal of the PCR amplification product, a true peak corresponding to the detection signal of the PCR amplification product of the DNA fragment and a Stutter corresponding to the detection signal of the PCR amplification product in which the repetitive sequence of the microsatellite is increased or decreased by 1 unit or more. A second determination processing unit for determining a peak;
In the system having the determination result of the + A peak and the peak other than the + A peak, and the determination result display processing unit for displaying the determination result of the true peak and the Stutter peak together with the graph,
Furthermore, the results of analyzing the length of PCR amplification products of DNA fragments containing microsatellite for a plurality of individuals, including the true peak, + A peak, and Stutter peak of a plurality of individuals processed in the past, And a database that stores the fragment lengths of multiple individuals processed in
A system for evaluating a determination result based on at least one of the following determination results for the determination results by the first determination processing unit and the second determination processing unit.
(1) Is the height ratio between the determined true peak and the Stutter peak not significantly different from the same ratio for a plurality of individuals accumulated in the database?
(2) Is the height ratio between the determined true peak and + A peak not significantly different from the same ratio for a plurality of individuals accumulated in the database?
(3) Are the determined true peak, Stutter peak, and + A peak fragment lengths not significantly different from the same length for multiple individuals accumulated in the database? The database further stores an experimental protocol when the analysis is performed together with the analysis result for each individual,
2. The evaluation system according to claim 1, wherein, in the evaluation of the determination result, only data in which a determination target and an experimental protocol coincide to a predetermined extent are used as the determination criterion among data stored in the database.   3. The evaluation system according to claim 1, wherein when the determination result is evaluated to be valid in the evaluation of the determination result, the analysis result of the determination target is accumulated in the database.

Images