JP2003079366A - Information processing system for assisting primer walking - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a means for performing information processing necessary for the determination of a base sequence by primer walking method on multiple DNA specimens in high throughput at the same time. SOLUTION: The waveform data obtained from multiple DNA specimens as an output of a DNA sequencer are inputted and subjected to base calling, the accuracy of the total trace data is estimated, the processing steps of the sequence assembly and the primer design are collectively and automatically performed on the trace data judged to have high accuracy by the accuracy estimation and the reason of the failure in getting a high-accuracy data is estimated on a trace data which is not judged to have high accuracy by accuracy estimation. The data processing on multiple DNA specimens can be collectively carried out by the system in high efficiency.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a primer designing process in a primer walking method, and relates to base calling, trace data accuracy evaluation, sequence assembly, primer design, and trace data analysis.

[0002]

2. Description of the Related Art In June 2000, the international joint project and the American venture company declared the completion of the sequencing of the human genome. With the advance of sequencing technology such as the spread of DNA sequencers using 4-color fluorescent dyes and capillaries, not only human genome but also genome sequences of various species from microorganism to mammal mouse have been determined. Sequencing of gene transcript mRNAs is also widely practiced.

The shotgun method and the primer walking method are known as methods for determining the DNA base sequence. Among them, the base walking method by the primer walking method is a useful method for determining the sequence of several thousand bases such as the work of determining the base sequence of cDNA or filling the gap sequence of the genome sequence, and is widely used. . In the primer walking method, sequencing is performed sequentially from both ends of the sequence.
From the sequence near the 3'end of the already determined sequence, a new primer is designed and synthesized, DNA amplification and extension reaction are performed, a sequence of about 500 bases is obtained using a DNA sequencer, and the determined sequence is extended. Go. By repeating this, the nucleotide sequence of several thousand bases is determined.

[0004] In the primer walking method, since the base sequence is reliably determined from the end of the sequence, there are few uncertainties in the sequence due to repetitive sequences present in the sequence, and the base sequence determination of the genome sequence is possible. The feature is that the determined sequence is more accurate than the shotgun method used. In addition, a primer walking method capable of position-specific sequence determination is indispensable in the work of filling a gap sequence in a genome sequence.

[0005]

[Problems to be Solved by the Invention] In the primer walking method, DNA amplification / elongation reaction and electrophoresis by a DNA sequencer and information processing necessary for the next experimental operation are performed.
You need to alternate. This information processing includes base calling, sequence assembly, vector sequence removal, and primer design. These information processing operations are sequentially performed based on the trace data obtained by the experimental operation, and cannot be performed all at once. On the other hand, there is a problem that a computer system that collectively carries out these information processings has not become widespread, and a large number of people need to work for a long time.

In addition, simply designing a primer in some cases may not be useful for nucleotide sequence determination. If the accuracy of the entire trace data obtained in the experiment is low, it is necessary to redo the experiment again, taking into consideration the possibility of incorrect experimental operation and equipment failure. For individual samples,
It is known that when GC clusters are present in the nucleotide sequence, the signal value of the trace data drops sharply, making subsequent sequencing impossible. When the cluster of A (or the cluster of T) is present, the background noise becomes high and it becomes difficult to obtain a highly accurate array.

[0007] The problem to be solved by the present invention is to provide a means for simultaneously performing, with high throughput, a plurality of DNA samples for information processing required for nucleotide sequencing by the primer walking method, and with low accuracy. When the trace data is obtained, it is to automatically determine whether the cause is an experiment or a characteristic peculiar to the base sequence of the trace data.

[0008]

[Means for Solving the Problems] In order to process information involved in the sequencing work by the primer walking method, a system for performing the following steps is introduced. (1) A step of inputting the trace data output by the DNA sequencer for a plurality of DNA samples. (2) Base calling step for obtaining high-precision base sequences from trace data. The base calling is a process of analyzing a peak of a fluorescent signal recorded in trace data to obtain a base sequence. In addition, in the present specification, the high-precision base sequence is the longest partial sequence without N extracted from the sequence obtained as a result of the base cola of the present invention. (3) Input the high-precision base sequence length of the trace data and use the accuracy evaluation of the entire trace data obtained in the same experiment to determine whether the proportion of sequences whose sequence length is less than the user-specified parameter exceeds another user-specified parameter. And a step of performing accuracy evaluation of each trace data depending on whether or not the array length and its deviation value are below different user-specified parameters. (4) Sequence assembly step for each DNA sample. (5) Primer design step for each DNA sample. (6) In step (3), it was determined that the accuracy was not high as a result of accuracy evaluation for each trace data (that is, low accuracy).
The step of inferring the reason why the accuracy of each trace data was not high.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described with reference to FIG. The system of the present invention takes as input, for a plurality of samples, all trace data that is waveform data obtained by using a DNA sequencer for each sample. In Figure 2 at 202, DNA using 4-color fluorescent dye
An example of trace data obtained using a sequencer is shown. Trace data is the number of data points according to the number of samplings during electrophoresis with a DNA sequencer, A,
The signal values corresponding to each base of T, G, and C are recorded. Hereinafter, the value of the trace data at base b and data point x is represented by f (b, x).

In the system of the present invention, after the input trace data 101 is read, a base calling step 102 extracts a highly accurate base sequence from the trace data. Then, the trace data accuracy evaluation step 103 is executed. If it is judged in the accuracy evaluation step 103 that there are many trace data with low accuracy, it is judged that there is a problem in the experimental operation, and the processing is stopped. In other cases, the accuracy of each trace data is evaluated, and a base sequence that can be determined by that time is created in the sequence assembly step 104 for the DNA sample having the trace data that is determined to have high accuracy. In the primer designing process step 105, a primer for extending the determined sequence is designed. For a DNA sample having trace data that is determined to be not highly accurate, in step 106 of inferring the cause, the reason why the highly accurate sequence was not obtained is predicted.

The final output is, for each sample, the determined sequence and new primers up to that point, or
One of the reasons why the trace data was not accurate will be output.

Hereinafter, the base calling step 102,
Trace data accuracy evaluation step 103, sequence assembly step 104, primer design processing step 1
The embodiment of step 106 of predicting the reason why the high-accuracy trace data was not obtained will be described.

First, the base calling step 102 will be described with reference to FIG. First, trace data
f (b, x) First derivative f '(b, x) obtained by differentiating 202 with respect to x
203, the second derivative f ″ (b, x) 204 is calculated. Then, as in 206, the value of the trace data exceeds the value of the trace data corresponding to another base, and the second derivative
The position where f ″ (b, x) 204 takes a minimum value is considered as a candidate for a peak position representing a base. Such a position can be calculated from the position where the third derivative 205 of the trace data crosses zero. Widely used base cola phred (Ew
ing, B. et al., Genome Research, 8: 175-185, 199.
In 8), the position where f (b, x) is the maximum is a peak position candidate, but in the present invention, the position where the second derivative has a minimum value is a peak position candidate, as indicated by 301 in FIG. f (b, x)
This is because it is possible to detect not only a clearly shaped peak having a maximum value of, but also a peak buried in another peak such as 302 and having no maximum value.

A method of analyzing the peak position candidate 401 is shown in FIG.
Will be explained. max {f (b, x)} (b ∈ {A, T, G, C})
Let b be B and the second derivative f ″ (B, x) 204 has a minimum value at x = X. At this position 401, it is calculated whether the following conditions are satisfied. Where P1, P2, P3, P4
Is a parameter given by the user and has a different value depending on the condition of the required array precision. 1. The value of f ″ (B, X) 402 is less than or equal to the threshold P1. 2. f (B, X) 403 is greater than or equal to the threshold P2. 3.f (B, X) / max {f (b, x)} (where b ∈ {A, T, G,
C} − {B}) is greater than or equal to the threshold P3.

In all of the above judgments, it is judged that a base exists when the condition as a peak corresponding to the base is satisfied. N is inserted into the sequence at peak positions that do not satisfy the above conditions. The bases obtained by this treatment are linked to form a base sequence. However, when the position of the immediately preceding base is Y, it is determined whether the distance X-Y, that is, 404 in FIG. 4 is appropriate, and the average value of the peak-to-peak distances of the preceding n bases (n is a user parameter) is d. When it does, it judges by whether or not (X−Y) / d ≦ P4 is satisfied, and if the interval is too wide, N is inserted. Here, the reason why d is calculated and the ratio to XY is calculated is to consider that the peak-to-peak distance in the trace data is not constant. The longest sequence without N is a high-precision base sequence. In the trace data with low precision, N is frequently entered, so the length of the longest array without N is short. Therefore, it can be said that the length of the array without N output by the base calling step 102 reflects the accuracy of the trace data. As an example of the base calling result, f '(b, x) = f (x + 2) + f (x + 1)) − (f (x−1) + f (x−2)) f''(b, x) = f (x + 4) + f (x + 3) + f (x + 2) + f (x−2) + f (x−3) + f
(x−4) − (f (x−1) + f (x) + f (x + 1)) P1 = 100, P2 = 50, P3 = 0.5, P4 = 1.5 20 for the array
Shown in 1. This array 201 is an array containing N immediately before extracting the longest array without N.

Next, the accuracy evaluation step 103 will be described. In step 103, the accuracy evaluation is performed on the entire trace data obtained in the latest experiment in the process of alternating information processing and experiment in primer walking. The accuracy of individual trace data is evaluated using all the trace data obtained in the same latest experiment. The length of the high precision base sequence obtained in the base calling step 102 reflects the precision of each trace data. In the accuracy evaluation step 103, the accuracy of the latest entire experiment is evaluated using the parameters P5 and P6 provided by the user. It is determined whether or not the length of the high-precision base sequence derived from each trace data is shorter than P5. If the proportion of sequences whose length is less than P5 exceeds P6, it is judged that there is a small proportion of highly accurate trace data in the entire experiment,
To that effect, the user is notified and the processing is terminated.

If not, the accuracy of each trace data is evaluated. First, the average value m and the standard deviation σ of the lengths of all high-precision base sequences obtained in the latest experiment are obtained.
Then, when the length of the high-precision base sequence obtained from certain trace data is L, and when any of the following conditions is satisfied, it is determined that this trace data has high precision. 1. L> P7 2. (L−m) / σ> P8 where P7 and P8 are parameters given by the user.

When each trace data is judged to be highly accurate in the accuracy evaluation step 103, the sequence fragment obtained from the trace calling data in the base calling unit is applied to the same DNA sample in the sequence assembling step. The consensus sequence is generated by assembling from the trace data obtained up to now with the sequence obtained by base calling. This is the partial sequence (801, 802) sequenced up to that point. The assembling method will be described with reference to FIG. Arbitrary sequence fragment 501
And an arbitrary sequence fragment 502 are combined when the tail sequence 503 of the sequence fragment 501 and the head sequence 504 of the sequence fragment 502 are sufficiently similar. Similarly, the arbitrary sequence fragment 505 and the arbitrary sequence fragment 506 are combined when they match a part of the sequence fragment 505 over the entire length of the sequence fragment 506.

For comparison between sequences, a dynamic programming algorithm capable of searching for similar sequences even in the presence of mismatches and gaps (“Bioinformatics”, Chapter 2, Dur
bin et al., translated by Akutsu et al., medical publication). The scores in the leftmost column and the uppermost row, which are the calculation start positions of the DP matrix, are 0 according to the requirement of the algorithm that allows duplicate matching described in Chapter 2 of “Bioinformatics”.
And

Sequence comparison by dynamic programming is known as a time-consuming algorithm because the calculation precision is high but the calculation time is proportional to the product of the two sequence lengths to be compared. In this system, when comparing two sequences A and B,
Fixed-length array A'601 from the beginning of each array,
B ′ 602 is taken out and A and B ′ 602 and B and A ′ 601 are compared by dynamic programming. If B'602 was found to be similar to the sequence starting at the p-th base of A, then A and B as a whole,
The search range is compared by a dynamic programming method that is limited to a certain range around the diagonal line that includes the p-th base of A in the DP matrix. Similarly, if it is found that A ′ 601 is similar to the sequence starting from the qth base of B, the entire search range of A and B is set as a constant around the diagonal line including the qth base 603 of B in the DP matrix. The comparison is made by dynamic programming limited to the range. FIG. 6 shows the area on the DP matrix that needs to be searched for the latter case. With the method of the present invention, the calculation amount is suppressed to a time proportional to each sequence length.

The array obtained as a result of the assembly is a partial array whose sequence determination has been completed up to that point, and this array data is output. When the sequencing by the primer walking method is completed, the sequence output here becomes the base sequence itself of the DNA sample for which the sequencing was attempted.

The primer used to obtain the nucleotide sequence 702 extending the sequence 701 obtained as a result of the assembly is designed near the end of the sequence 701 at a position 703 having a good property as a primer. However, referring to the position information 101 of the existing primer, if there is one of the existing primers in which the distance to the end of the sequence is shorter than the expected sequence length obtained by one-time electrophoresis, next time, It is judged that the primer necessary for the migration of DNA is already present, and no new primer is designed. In designing the primer, the possibility of dimer formation, the possibility of secondary structure formation such as hairpins, described in "Forefront of PCR Method" (edited by Takeo Sekiya and Kaoru Fujinaga, Kyoritsu Shuppan) near the end of the sequence, melting
Design primers considering temperature and false positive priming site. Melting temperature
The nearest neighbor method (Bre
slauer et al., Proc. Natl, Acad. Sci. USA, vol.
38, pp.3746-3750, 1986) and parameters of SantaLucia (SantaLucia, J. Jr., Proc. Natl. Acad.
Sci. USA, vol 95, pp.1460-1465, 1998).

False positive priming site in sample DNA
If present, the primer will anneal to the originally unintended region of the sample DNA, which is a major factor causing noise in the data. In order to eliminate the possibility of the presence of false positive priming sites to the maximum extent, this system uses the forward-side sequenced region 801 and the reverse-side sequenced region 802, the vector site sequence, and their entire complementary sequences when designing the primer. Is searched for a partial sequence similar to the candidate primer sequence. When a similar partial sequence is found, the candidate primer sequence is rejected. This search process is performed by suffix trie or suffix tree
High-speed processing using (Inenaga, S. et al., Proc. 1
2 ^th Ann. Symp. On Combinatorial Pattern Matching,
LectureNote in Computer Science 2089, p.169, 20
01, Hass, S. et al., Nucleic Acid Research, Vol
26, No. 12, pp. 3006-3012).

The designed primer sequence is used for the next DNA
Primer 1 used in amplification / extension reaction and electrophoresis
It is output as 07. The primer sequences are output in tab-delimited text format that can be read by general-purpose spreadsheet software, making it easy to maintain and manage primer information.

Next, in the trace data accuracy evaluation step 103, with respect to the trace data which is judged not to be highly accurate, the reason why it is not highly accurate is estimated step 106.
Will be described. First, it is determined whether or not a region having an extremely low signal value is found in the trace data, which makes it difficult to extract a base sequence based on a waveform. If the data points of the trace data are from 1 to N (N is the number of data points recorded in the trace data), for each base b of A, T, G, C, the signal value of the trace data The average value M can be calculated by Equation 1.

[0026]

[Equation 1] In the present invention, in the trace data, the trace data corresponding to any of A, T, G, and C is less than or equal to M in the window of the width P9 (P9 is a parameter given by the user) whose left end is a position x. If there is, the signal value is judged to be low at the position x and the user is notified. To prevent false detection of waveform distortion at the beginning of trace data, consider only the range where x is P10 (P10 is a user parameter) or higher. Strictly speaking, if there exists x that satisfies the following equation, it is determined that there is a signal decrease at position x. x ≧ P10, x ≦ ∀y <x + P9, ∀b∈ {A, T, G, C}, f (b, y)
≦ M In the present invention, when a signal decrease is observed, the characteristic of the sequence considered to be the cause is searched for from the sequence relating to the DNA sample from which the trace data was obtained. For the sequence to be searched, the base calling step 1 is performed from the trace data.
Starting with the base sequence extracted in No. 02, other base sequences obtained from the DNA sample are included. An example of the feature to be searched is a GC cluster. In the present invention, in the sequence to be searched, a position where 10 out of 12 consecutive bases are G or C is detected as a GC cluster, and there is a GC cluster that may cause a signal decrease to the user. Notify to that effect.

Next, it is determined whether the background noise is high. In the present invention, in the trace data,
Of the signal values irrelevant to the base sequence obtained in the base calling unit 102, the maximum value at each data point is selected, and the sum thereof is defined as background noise. To calculate the background noise, the present invention uses the set B at any data point x.
Define [x]. The element of B [x] is a base corresponding to a non-noise peak including the data point x as a peak range. FIG. 9 shows an example of B [x]. In the present invention, the region where the signal value continues to increase monotonically, reaches a maximum, and then continues to decrease monotonously is considered to be the peak range. Therefore, the left part of the peak range is a section where f '(b, x) is positive, and the right part is a section where f' (b, x) is negative. Strictly speaking, B [x] is defined by Equation 2. This equation 2 is an expression representing the definition of the set B [x] of bases having a signal that is not noise at the data point x in the trace data.

[0028]

[Equation 2] In Equation 2, s [i] is the base calling step 102.
In the base sequence obtained in step i, the i-th base, p [i] is the position of the data point where the i-th base s [i] was detected. B
Using [x], the background noise e of the trace data is defined by Equation 3.

[0029]

[Equation 3] The values of e are calculated for the trace data derived from a plurality of samples that were simultaneously processed by the experimental operation, and the average value m (e) and the variance σ (e) are obtained. Deviation value (e−m (e)) /
σ (e) is obtained, and when the value is P11 (P11 is a parameter given by the user) or more, it is determined that the background noise is high and the user is notified.

When it is determined that the background noise is high, the characteristic of the sequence considered to be the cause is searched for from the sequence relating to the DNA sample for which the trace data was obtained. The sequence to be searched includes the base sequence extracted in the base calling step 102 from the trace data,
Other base sequences obtained from the DNA sample are included. Examples of features to be searched include Poly-A and Poly-T. In the present invention, the region where A continues for 7 bases or more is Poly-A, T
The area where 7 or more are continuous is defined as Poly-T. Poly-A,
The presence of Poly-T may be causing the background noise to the user.
Notify that Poly-A and Poly-T exist.

[0031]

According to the present invention, it becomes possible to collectively execute data processing for a plurality of DNA samples. If parameters are given in advance, there is no step that requires human intervention during the process, and the time required for data processing of multiple DNA samples, which conventionally took several days, is greatly improved to about 30 minutes. can do. Furthermore, false positi
The ve priming site is searched for from the forward side determinant sequence, the reverse side determinant sequence, the vector sequence, and the complementary sequences of all of them, and the possibility of false positive priming site is reduced to the maximum, so that not only simple efficiency improvement , The accuracy of the decision sequence is improved. Since the accuracy of the entire trace data obtained in the experiment is automatically evaluated, there is no need to manually inspect the quality of the trace data. Users will also be alerted early on the presence of obfuscated clones as they will be warned about problematic samples.
Can take measures.

[Brief description of drawings]

FIG. 1 is a diagram showing a system configuration of the present invention.

FIG. 2 is an explanatory diagram of base calling and differentiation calculated in the process in the present invention.

FIG. 3 is an explanatory diagram of a peak where a signal value in trace data has a maximum value and a peak which is buried in another peak and does not have a maximum value.

FIG. 4 is an explanatory diagram of an evaluation scale for determining whether a peak of a signal value in trace data represents a base.

FIG. 5 is an explanatory diagram showing a relationship between sequence fragments to be bound in a sequence assembly unit.

FIG. 6 is an explanatory diagram of a region in the array assemble unit that requires calculation on the DP matrix.

FIG. 7 is an explanatory diagram of a primer design position and a similar sequence search position during sequence assembly.

FIG. 8 is an explanatory diagram of a forward side sequence-determined region, a reverse side sequence-determined region, a sequence undetermined region, and a vector site.

FIG. 9 is an explanatory diagram of the definition of B [x].

[Explanation of symbols]

Reference numeral 101 is trace data of a plurality of samples, which is an input of the system of the present invention. At 102, base calling is performed. At 103, the accuracy of the trace data is evaluated. At 104, the base sequence fragments are assembled. At 105, a primer for determining the base sequence of the undetermined sequence portion is designed. At 106, the cause of the inaccurate trace data is analyzed. 107 is a newly designed primer or a prediction result of the reason why the accuracy of trace data is not high, and is the output of this program. Reference numeral 108 denotes a processing flow for the high precision trace data. Reference numeral 109 is a processing flow for trace data that is not highly accurate. 201 is an example of a base sequence obtained by base calling. 202 is an example of trace data obtained by a DNA sequencer. 203 is an example of the first derivative of trace data. Reference numeral 204 is an example of the second derivative of the trace data. 205 is an example of the third derivative of the trace data. 206 is the position where the second derivative has a minimum value and there is a peak candidate. 301 is a peak position where the trace data has a maximum value. 302 is a peak position of the trace data, but it is a peak position that is buried in other peaks and does not have a maximum value. Reference numeral 401 represents a peak position candidate of the trace data. Reference numeral 402 represents the value of the second derivative of the trace data. Reference numeral 403 represents the intensity of the trace data. 404 represents the distance between peaks derived from adjacent bases. 501 is a sequence fragment. 502 is a sequence fragment different from 501. 503 is a sequence fragment 50 similar to the first sequence of 502.
The partial array at the end of 1. 504 is a sequence fragment 50 similar to the sequence at the end of 501
The first partial array of 2. 505 is a sequence fragment. 506 is a sequence fragment similar to the partial sequence of 505. An array A ′ 601 is a partial array at the head of the array A. An array B ′ 602 is a partial array at the head of the array B. 603 is the position of the qth base of sequence B. 701 is a sequence fragment. 702 is a sequence fragment obtained by using a primer having 701 as a template sequence. 703 is a position serving as a template for the primer used to obtain 702. 704 is a partial array at the end of 601 which is similar to the beginning part of 702. 801 is a sequenced site on the forward side. 802 is the reverse side sequenced site. 803 is a vector site.

Continued front page    (72) Inventor Takashi Minowa             4-6 Kanda Surugadai, Chiyoda-ku, Tokyo             Hitachi, Ltd. Life Science Promotion             Within the business unit (72) Inventor Hiroshi Kondo             4-6 Kanda Surugadai, Chiyoda-ku, Tokyo             Hitachi, Ltd. Life Science Promotion             Within the business unit (72) Inventor Takako Furuya             4-6 Kanda Surugadai, Chiyoda-ku, Tokyo             Hitachi, Ltd. Life Science Promotion             Within the business unit (72) Inventor Emiko Yoshikawa             4-6 Kanda Surugadai, Chiyoda-ku, Tokyo             Hitachi, Ltd. Life Science Promotion             Within the business unit (72) Inventor Yuhi Komura             4-6 Kanda Surugadai, Chiyoda-ku, Tokyo             Hitachi, Ltd. Life Science Promotion             Within the business unit (72) Inventor Tetsuo Nishikawa             1-280, Higashi Koikekubo, Kokubunji, Tokyo             Central Research Laboratory, Hitachi, Ltd. F-term (reference) 4B024 AA11 AA20 HA08 HA19                 4B063 QA13 QQ42 QQ52 QS39                 5B075 ND20 UU18

Claims (8)

[Claims] 1. With respect to a plurality of DNA samples, trace data, which is waveform data output by a DNA sequencer, is input,
Base calling, which is the process of extracting the base sequence from the trace data, is performed, and the accuracy of the entire trace data obtained in the same experiment is evaluated in the process of alternating the information processing and the experiment in primer walking. For the sample for which the accuracy evaluation was performed and the trace data that was determined to be highly accurate in the accuracy evaluation for each trace data was obtained, the sequence obtained by performing sequence assembly and the primer design were performed. An information processing system for primer walking, which is characterized by predicting the reason why highly accurate trace data was not obtained for a sample for which trace data that was determined to be not highly accurate was obtained by the accuracy evaluation of the above trace data. 2. The function of performing accuracy evaluation of the entire trace data and individual trace data, and the function of predicting the cause of the trace data which is not highly accurate in the accuracy evaluation of the individual trace data are deleted.
Information processing system for walking primer. 3. When performing base calling, a position at which the second derivative of the trace data has a minimum, not the maximum value of the trace data itself, is used as a candidate for a position where a peak corresponding to a base exists. The information processing system for primer walking according to claim 1. 4. A high-precision base sequence obtained by base calling, when performing accuracy evaluation of the entire trace data obtained in the same experiment in the process of alternating information processing and experiment in primer walking, The information processing system for primer walking according to claim 1, wherein the accuracy of the whole is evaluated by calculating the ratio of those having a short sequence length. 5. When performing accuracy evaluation of individual trace data, individual trace data are obtained by using a value calculated from an average value and standard deviation of sequence lengths obtained from trace data derived from a plurality of samples. 2. The primer walking information processing system according to claim 1, wherein the accuracy of the primer walking is evaluated. 6. When a sequence to be compared is set to A and B in an overlap search between two base sequences in a sequence assembly process, a fixed length head portion of a base sequence fragment of A and B are dynamically moved in advance. If a similar sequence is not found by performing the sequence comparison by the design method, the sequence comparison by the dynamic programming method is performed for the whole A and the fixed-length head portion of the base sequence fragment of B,
When a sequence similarity region is found in one of the two sequence comparisons, the overlap search is performed as the sum of the sequence lengths of both sequences by applying the sequence comparison algorithm by dynamic programming in which the number of gaps is limited based on the position. The information processing system for primer walking according to claim 1, wherein an assembling method that is performed in a time proportional to is implemented. 7. A method for predicting the cause of trace data that is not highly accurate as a result of an accuracy evaluation among individual trace data, for a plurality of samples,
The value of background noise obtained from the trace data is obtained, and the background noise of each trace data is evaluated by using the value calculated from the average value and standard deviation of those values. When it is determined that the value is high, the characteristic sequence that causes it is searched from the sequence obtained in the sequence determination process of the DNA sample from which the trace data was obtained. Information processing system for primer walking. 8. A decrease in signal when predicting the cause of trace data that has been determined to be not highly accurate as a result of accuracy evaluation among individual trace data, and a decrease in signal is indicated by an average value of signal intensities in the trace data. When a signal decrease is observed, the characteristic sequence that causes the decrease is searched for from the sequence obtained in the sequencing process of the DNA sample from which the trace data was obtained. The information processing system for primer walking according to claim 1.