Information processing device and information processing method

Abstract

To perform labeled positions alignment capable of dealing with apparent expansion and contraction of a target nucleic acid sequence.SOLUTION: An information processing device calculates a plurality of first ratios of intervals between partial sequences in a referenced nucleic acid sequence, constructs an index indicating a combination of the first ratios and information indicating a position of a partial sequence in the nucleic acid sequence corresponding to the combination of the first ratios, calculates a plurality of second ratios of intervals between partial sequences in a target nucleic acid sequence, extracts a combination of the first ratios corresponding to a combination of the second ratios based on a comparison result between the combination of the second ratios and the combination of first ratios indicated by the index, and outputs information indicating a position of a partial sequence corresponding to the extracted combination of the first ratios in the referenced nucleic acid sequence.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an information processing apparatus and an information processing method.

Background Art

[0002] With the progress of DNA (DeoxyriboNucleic Acid) sequencing technology, many personal genomes have been revealed. A personal genome contains many differences from the reference genome. The majority of these are SNVs (Single Nucleotide Variants) where only one base in the surrounding base sequence is different from the reference genome, but structural variants (SVs, Structural Variants) where large base sequences of several thousand bases or more change at once are also included, although the number is small compared to single-base mutations.

[0003] SNVs and SVs include not only germline mutations that cause individual differences but also acquired mutations called somatic mutations, and some of the acquired mutations cause canceration. Accurately detecting such mutations and elucidating their biological and clinical significance is an important issue in cancer treatment and biological research.

[0004] In order to clarify structural variants, it is necessary to capture changes in genomic regions larger than several thousand bases. However, the length of the base sequence that can be read at one time by current DNA sequencing technology is limited. The length of the base sequence that can be read at one time is limited to about 1000 bases at most by the Sanger method used when initially determining the human standard genomic sequence, and to about several hundred bases by the currently mainstream NGS (Next Generation Sequencing).

[0005] In NGS, it is also possible to obtain pairs of two base sequences separated by several hundred bases, called paired-end sequences. However, even when using paired-end sequences, only sequences in a narrow region of about 1000 bases can be obtained. However, the human genome contains many repetitive sequences such as SINE (Short Interspersed Nuclear Element) and LINE (Long Interspersed Nuclear Element), and repetitive sequences also exist in regions called centromeres and telomeres.

[0006] If at most only 1000 bases of base sequences can be seen at a time, these repetitive sequences cannot be distinguished, so even if the obtained base sequences are joined together, the base sequence of the entire genome cannot be estimated. In long-read sequencing technology that has become popular in recent years, base sequences spanning tens of thousands of bases can be obtained at a time, but it is insufficient to identify the positions of all repetitive sequences. Therefore, a technology for analyzing a wider region on the genome is needed.

[0007] A technology called genome mapping, which can be used for such applications, is to label specific short base sequences on the genome with fluorescence or the like and identify the positions on the genome from the pattern of the label intervals. In genome mapping, the DNA constituting the genome is amplified and cut to generate a large number of DNA fragments consisting of hundreds of thousands of bases.

[0008] In genome mapping, specific base sequences on each of the many generated DNA fragments are labeled, and the label position indicating at approximately which base position from the beginning each label sequence (hereinafter also simply referred to as a label) appears on the DNA fragment is measured. Furthermore, for each DNA fragment, by arranging the label positions in ascending order, each DNA fragment can be converted into an ascending numerical sequence. This numerical sequence is also referred to as measurement data below.

[0009] One document disclosing such genome mapping technology is Japanese Patent Publication No. 2009-022274 (Patent Document 1). This publication describes "a method for mapping positions on chromosomal DNA, comprising hybridizing nucleic acid with a single repeat base sequence in unfolded or extended chromosomal DNA, measuring the distance between multiple sets of said repeat base sequences on chromosomal DNA using a label introduced into the hybridized nucleic acid, and then determining the region or position on the chromosome of the set and the repeat base sequences contained in the set based on the characteristics of the measured distances" (see abstract).

[0010] The process of comparing measurement data obtained through genome mapping with label locations obtained from a reference genome sequence to identify common and non-common regions is called alignment. If there are no structural mutations in the DNA fragment from which the measurement data is based, or if there are no measurement errors in the measurement data, each label location indicated by the measurement data corresponds to one of the label locations on the reference genome. On the other hand, if structural mutations occur, the corresponding locations of the labels on the measurement data and the labels on the reference genome become discontinuous. Structural mutations can be detected by capturing such abnormalities in label locations. [Prior art documents] [Patent Documents]

[0011] [Patent Document 1] Japanese Patent Publication No. 2009-022274 [Overview of the project] [Problems that the invention aims to solve]

[0012] To detect structural variations, accurate alignment is necessary. If there are many errors during alignment, there is a risk that abnormalities in the labeling position caused by the errors will be mistakenly identified as structural variations.

[0013] To align measurement data, it is necessary to address errors contained in the measurement data. One such error is that the overall length of each DNA fragment may appear stretched or contracted in the measurement data. This is due to the uneven movement speed of molecules during measurement.

[0014] While Patent Document 1 discloses labeling repetitive sequences on the genome and identifying their location within the genome, it does not disclose a method for matching the measured labeling interval with the labeling location on a reference genome.

[0015] On the other hand, one aspect of the present invention performs alignment of the labeling position that can accommodate apparent stretching and shrinking of the target nucleic acid sequence. [Means for solving the problem]

[0016] To solve the above problems, the following configuration is adopted in one embodiment of the present invention. The information processing device comprises a processor and a memory, the memory holding a first numerical sequence indicating the position of a subsequence in a referenced nucleic acid sequence and a second numerical sequence indicating the measurement position of the subsequence in a target nucleic acid sequence, the processor calculating a plurality of first ratios of the intervals of the subsequences in the referenced nucleic acid sequence based on the first numerical sequence, constructing an index showing combinations of the first ratios and information indicating the position of the subsequence in the referenced nucleic acid sequence corresponding to the combinations of the first ratios, calculating a plurality of second ratios of the intervals of the subsequences in the target nucleic acid sequence based on the second numerical sequence, extracting a first ratio combination corresponding to the second ratio combination based on the comparison result of the second ratio combination and the first ratio combination shown by the index, and outputting information indicating the position of the subsequence in the referenced nucleic acid sequence corresponding to the extracted first ratio combination. [Effects of the Invention]

[0017] According to one aspect of the present invention, it is possible to perform alignment of the label position that can accommodate apparent stretching and shrinking of the target nucleic acid sequence.

[0018] Other issues, configurations, and effects not mentioned above will be clarified by the following description of the embodiments. [Brief explanation of the drawing]

[0019] [Figure 1] This is a block diagram showing an example configuration of the genome labeling position alignment device in Example 1. [Figure 2] This figure shows an example of the data structure of the measurement data in Example 1. [Figure 3] This flowchart shows an example of the index construction process in Example 1. [Figure 4] This is an explanatory diagram showing an overview of the index construction process in Example 1. [Figure 5] This figure shows an example of the index data structure in Example 1. [Figure 6] This flowchart shows an example of the index search process in Example 1. [Figure 7] This flowchart shows an example of the process of searching for an index while thinning out some of the markers in Example 1. [Figure 8] This is an explanatory diagram showing an example of a DNA fragment from which the label was thinned in the measurement data of Example 1. [Figure 9] This flowchart shows an example of the alignment probability calculation process in Example 1. [Figure 10] This sequence diagram shows an example of the overall processing using a genome labeling site alignment device. [Figure 11] This is an explanatory diagram illustrating an example of a tree structure constructed by extending the k-tuple indicated by the index in Example 2. [Figure 12] This flowchart shows an example of alignment processing using a tree structure in Example 2. [Figure 13] This flowchart shows an example of the update process for sets S and T performed in the alignment process using a tree structure in Example 2. [Figure 14] This figure shows an example of the user interface displayed on the input / output device in Example 3. [Figure 15] This is an explanatory diagram showing an example of the structural variation detection process in Example 4. [Modes for carrying out the invention]

[0020] The following describes a genome labeling position alignment device according to one embodiment of the present invention. In the following figures, components common to all figures are denoted by the same reference numerals, and redundant explanations are omitted.

[0021] Errors in the measurement data include not only the apparent elongation or contraction of the overall length of the DNA fragment as described above, but also the failure to detect the label on the DNA fragment, false detection at locations where there is no label on the DNA fragment, and cutting of the DNA fragment during sample preparation. The genome labeling alignment device of this embodiment realizes alignment processing that can address these errors. [Examples]

[0022] <Device configuration> Figure 1 is a block diagram showing an example configuration of a genome labeling site alignment device. Mu sign The recognition and position alignment device is comprised of a computer 300, which includes, for example, a CPU (Central Processing Unit) 310, memory 311, auxiliary storage device 312, and interfaces 313 to 315. The hardware included in the computer 300 is electrically connected, for example, via internal communication lines such as a bus.

[0023] The CPU 310 reads programs and data stored in memory 311 and executes programs stored in memory 311. The CPU 310 includes a processor. The CPU 310 includes, for example, an index building unit 321, an index search unit 322, and an alignment probability calculation unit 323, all of which are functional units. The computer 300 functions as a genome labeling position alignment device through the processing performed by the CPU 310.

[0024] Memory 311 temporarily stores programs executed by the CPU 310 and data used during program execution. Memory 311 includes non-volatile memory elements such as ROM (Read Only Memory) and volatile memory elements such as RAM (Random Access Memory). ROM stores immutable programs (e.g., BIOS (Basic Input / Output System)). RAM is a high-speed, volatile memory element such as DRAM (Dynamic Random Access Memory) and temporarily stores programs executed by the CPU 310 and data used during program execution.

[0025] Memory 311 stores, for example, programs that implement the index construction unit 321, the index search unit 322, and the alignment probability calculation unit 323, as well as the reference genome marker position 330 and measurement data 340.

[0026] For example, the CPU 310 functions as an index construction unit 321 by operating according to an index construction program loaded into memory 311, as an index search unit 322 by operating according to an index search program loaded into memory 311, and as an alignment probability calculation unit 323 by operating according to an alignment probability calculation loaded into memory 311.

[0027] The auxiliary storage device 312 non-volatilely stores the program executed by the CPU 310 and the data used during program execution. In other words, the program is read from the auxiliary storage device 312, loaded into memory 311, and executed by the CPU 310.

[0028] The auxiliary storage device 312 is a high-capacity, non-volatile storage device such as an HDD (Hard Disk Drive) or SSD (Solid State Drive). The auxiliary storage device 312 stores the programs that implement the functions of the index construction unit 321, the index search unit 322, and the alignment probability calculation unit 323, as well as the reference genome labeling position 330 and measurement data 340.

[0029] Interfaces 313 to 315 each act as a medium for transmitting and receiving signals and converting protocols, and are connected to external devices. Interface 313 is an I / O interface connected to the input / output device 302 via a wired or wireless line. The input / output device 302 includes input devices such as a keyboard and mouse, and output devices such as a display device and a printer. Interface 313 acquires input information from the operator received by the input / output device 302. Interface 313 also outputs the program execution results to the input / output device 302 in a format that can be viewed by the operator.

[0030] Interface 315 is a network interface connected to the external storage device 301 via network 305. Interface 315 controls communication with other devices according to a predetermined protocol.

[0031] The external storage device 301 is a non-temporary storage device that stores data handled by the computer 300. The external storage device 301 includes, for example, storage devices such as HDDs and SSDs. The external storage device 301 can store reference genome labeling locations 330 and measurement data 340.

[0032] Data transmission and reception between the external storage device 301 and the computer 300 are performed via the network 305. The network 305 includes, for example, a LAN (Local Area Network) and the Internet. However, the type of network 305 is not limited to those described above. The network 305 may be wired or wireless.

[0033] Interface 314 is connected to a drive device that reads and writes to the removable media 303. Interface 314 includes, for example, a serial interface such as USB (Universal Serial Bus).

[0034] The removable media 303 is a non-temporary storage medium for storing data handled by the computer 300. The removable media 303 includes optical discs such as CDs and DVDs, magnetic discs, and semiconductor memory. The removable media 303 can store reference genome labeling locations 330 and measurement data 340.

[0035] Furthermore, some or all of the programs executed by the CPU 310 may be provided to the computer 300 via the interface 314 from the removable media 303, which is a non-temporary storage medium, or via the network 305 from the external storage device 301, which is a non-temporary storage device, or from an external computer equipped with the external storage device 301, and stored in the non-volatile auxiliary storage device 312, which is a non-temporary storage medium.

[0036] In Figure 1, the computer 300, which constitutes the genome labeling position alignment apparatus, is connected to an external storage device 301 and removable media 303. However, these external devices can be omitted if they are not needed. Furthermore, an input / output device 302 equipped with the external storage device 301 may be connected to the computer 300 via a network 305. Alternatively, the computer 300 may have a built-in device with input / output functionality instead of the input / output device 302 being connected.

[0037] The index construction unit 321 constructs an index based on k-tuples. The index construction unit 321 constructs an index based on the ratio of the indicator intervals, rather than an index using k-tuples that represent the interval between indicator positions (hereinafter simply referred to as the indicator interval). Furthermore, the index constructed by the index construction unit 321 is designed to handle cases where some indicators are missing, in order to address the possibility of missed indicator detection.

[0038] A k-tuple represents a combination of k numbers (where k is a predefined parameter). Assuming, as in conventional techniques, each k-tuple is a combination of the k marker intervals on the reference genome, and an index is constructed that is a correspondence table between each k-tuple and the identifier of the corresponding marker on the reference genome.

[0039] In this case, simply generating k-tuples in the measurement data 340 and performing alignment by comparing the generated k-tuples with the k-tuples indicated by the index will not be able to address the apparent expansion and contraction of the measurement data 340.

[0040] Furthermore, even if we were to generate k-tuples by seemingly stretching or shrinking the entire measurement data 340 at various stretching / shrinking rates in order to address the apparent stretching and shrinking of the measurement data 340, and then compare these k-tuples corresponding to the stretching / shrinking with the k-tuple indicated by the index, the processing time would be increased because it would be necessary to try a large number of potentially correct stretching / shrinking rates and adopt the optimal one in order to improve the accuracy of the alignment. Moreover, even if a vast number of stretching / shrinking rates were adopted, there is a risk that all of the adopted stretching / shrinking rates would deviate from the optimal stretching / shrinking rate for the measurement data 340, and consequently, accurate alignment may not be possible.

[0041] During DNA fragment measurement, the label spacing in the measurement data 340 expands and contracts significantly because the movement speed of each DNA fragment molecule is not uniform during measurement. However, the error in the ratio of the label spacings in the measurement data 340 is small. Therefore, in this embodiment, as described above, the index construction unit 321 constructs an index using k-tuples based on the ratio of the label spacings, thereby achieving highly accurate alignment that does not require expansion or contraction of the measurement data 340.

[0042] The index search unit 322 uses an index based on k-tuples constructed by the index construction unit 321 to identify locations on the reference genome that are highly likely to correspond to the measurement data. During the search, the index search unit 322 performs processing to address false detections of labels.

[0043] The alignment probability calculation unit 323 calculates the probability that the generated alignment will occur. When there are multiple candidate positions for the alignment of the DNA fragment indicated by the measurement data 340, if this probability is high, there is a high probability that it is a correct alignment, so the alignment probability calculation unit 323 adopts the alignment with the highest probability. Even if only a portion of the DNA fragment can be aligned, this probability is used to determine whether that alignment is optimal or not.

[0044] Before the genome labeling position alignment process begins, the computer 300 has the reference genome labeling position 330 and measurement data 340 input and stored in it. The CPU 310 may, for example, read the reference genome labeling position 330 and measurement data 340 when the computer 300 is started up or when processing is executed, and load them into memory 311.

[0045] The reference genome marker locations 330 and measurement data 340 may be stored in all of the auxiliary storage device 312, the external storage device 301, and the removable media 303, or they may be stored in only some of them. When the computer 300 is stopped or when there is insufficient free space in the auxiliary storage device 312, this data may be moved or copied to the external storage device 301 or the removable media 303. Therefore, it is desirable that the reference genome marker locations 330 stored in different storage devices all contain the same information. The same applies to the measurement data 340.

[0046] The reference genome marker position 330 includes numerical data representing the location of the marker present on each of the multiple reference genomes. The measurement data 340 includes data obtained by measuring the marker position on each of the numerous DNA fragments.

[0047] Figure 2 shows an example of the data structure of measurement data 340. Measurement data 340 shows the ID that identifies each DNA fragment (an example of the target nucleic acid sequence), the molecular length (length of the base sequence) of each DNA fragment, and the position of the marker (an example of a subsequence) on each DNA fragment. Blank spaces in measurement data 340 indicate that no marker was observed.

[0048] For example, the molecular length of a DNA fragment with ID "2" is 44,951 bases, and four markers are measured in this DNA fragment. The positions of these four markers are, respectively, the 10,844th base, 19,749th base, 23,353rd base, and 35,735th base from the beginning of the DNA fragment. Note that the position of the markers in the DNA fragment indicated by measurement data 340 may, for example, indicate the position at the beginning or the position at the end of the marker.

[0049] In this embodiment, an example of a short base sequence to be labeled on a DNA fragment is GCTCTTC, which is recognized by an enzyme called Nt.BspQI. As described above, measurement data 340 shows the labeling positions of GCTCTTC on each DNA fragment in ascending order. In other words, measurement data 340 shows information recorded as a numerical sequence (an example of a second numerical sequence) in ascending order of the labeling positions on each DNA fragment.

[0050] <Index Construction> Figure 3 shows the index structure construction place A flowchart shows an example of the process. The index construction unit 321 constructs k-tuples for any label location on the reference genome based on the ratio of label intervals, and constructs an index that shows the correspondence between the constructed k-tuples and the label locations on the reference genome.

[0051] Step S401: The index building unit 321 acquires genome sequence data. The genome sequence data indicates the number of each of the multiple chromosomes that make up the genome, and the base sequence of each of the multiple chromosomes (an example of a reference nucleic acid sequence). The base sequence of each of the multiple chromosomes is indicated by a string of characters consisting of letters representing the four types of bases A, T, G, and C, and N representing an unknown base. The genome sequence data is pre-stored in at least one of the following: memory 311, auxiliary storage device 312, removable media 303, and external storage device 301.

[0052] Step S402: If the index building unit 321 determines that all chromosomes have been selected, it completes the index building process. If it determines that there are unselected chromosomes, it proceeds to step S403.

[0053] Step S403: The index building unit 321 selects one unprocessed chromosome. The index building unit 321 then constructs a k-tuple and registers it in the index using the following procedure for the selected chromosome.

[0054] Step S404: The index building unit 321 calculates a numerical sequence (an example of a first numerical sequence) indicating the location of a marker (an example of a partial sequence) on the chromosome selected in the most recent step S403, and stores the numerical sequence in association with the chromosome number at the reference genome marker location 330. For example, the aforementioned GCTCTTC is used as the marker sequence. It should be noted that, since genomic DNA is a double helix, locations where the complementary sequence (a sequence obtained by substituting A, T, G, C with T, A, C, G respectively and reversing the order) matches the marker sequence are also labeled in genome mapping. Therefore, when the index building unit 321 calculates a numerical sequence representing the marker location, it is necessary to add the location of the marker sequence or its complementary sequence to the numerical sequence without distinction.

[0055] Step S405: The index building unit 321 calculates the ratio of the sign spacing and the ratio of the sign spacing when some signs are thinned out, based on the numerical sequence indicating the sign positions calculated in step S404. It then registers a k-tuple based on the calculated ratios in the index and returns to step S402.

[0056] A specific example of the process in step S405 will be explained using Figure 4. Figure 4 is an explanatory diagram showing an overview example of the index construction process. In the example in Figure 4, it is assumed that there are 10 markers, consisting of marker 1 to marker 10, in the reference genome, and that the positions of these markers have been obtained. Also, the interval between marker i and marker j is denoted by d(i,j).

[0057] In step S405, the index building unit 321 first uses the numerical sequence indicating the sign positions calculated in step S404 to calculate the sign spacing between adjacent signs and the sign spacing between signs that become adjacent after some signs are thinned out according to a predetermined rule. The predetermined rule here is, for example, a rule for thinning out signs when generating a k-tuple using k ratios of sign spacings, and in the case of k+3 consecutive signs, one of the signs from the 2nd to the k+2nd sign is thinned out. In other words, the signs to be thinned out are all signs except the 1st and k+3rd signs at both ends. Below, an example of calculating a k-tuple while sequentially thinning out one of the signs from sign 1 to sign 10 will be explained.

[0058] In this case, the index building unit 321 calculates the label interval d(i,i+1) for i=1,...,9. Here, the label interval d(x,y) represents the distance between label x and label y, and its unit is the number of bases. Furthermore, the index building unit 321 calculates the label intervals of adjacent labels by sequentially thinning out labels according to the predetermined rule, for example, by further thinning out label 2 so that label 1 and label 3 become adjacent, and so on, by further thinning out the labels.

[0059] The index building unit 321 calculates the ratio of adjacent marker intervals and the ratio of marker intervals that become adjacent by thinning out some markers according to the predetermined rule. That is, the index building unit 321 calculates d(i,i+1) / d(i+1,i+2) for i=1,...,8. Furthermore, the index building unit 321 calculates the ratio of marker intervals that become adjacent by sequentially thinning out markers according to the predetermined rule, for example, by further thinning out marker 2 so that marker intervals d(1,3) and d(3,4) become adjacent.

[0060] The index construction unit 321 registers in the index a k-tuple, which is a combination of k consecutive ratio values, and the corresponding location on the reference genome, based on the calculated ratio of the label spacing. Furthermore, the index construction unit 321 registers in the index a k-tuple, which is a combination of k consecutive ratio values, resulting from thinning out some labels according to a predetermined rule, based on the calculated ratio of the label spacing, and the corresponding location on the reference genome, based on the k-tuple, and registers in the index a k-tuple, which is a combination of k consecutive ratio values.

[0061] For example, the chromosome number from which the k-tuple was derived, and an integer value representing the position of the k+2 markers from the beginning of that chromosome, which were used to calculate the ratio values ​​contained in the k-tuple, are registered as the index of their positions on the reference genome.

[0062] For example, if k=3, then for each of i=1,...,6, the k-tuples obtained without thinning out the labels are d(i,i+1) / d(i+1,i+2), d(i+1,i+2) / d(i+2,i+3), and d(i+2,i+3) / d(i+3,i+4), which represent the chromosome number being selected and the position on the reference genome, and labels i to i+4.

[0063] Furthermore, the index building unit 321 registers in the index the ratio of the intervals between k adjacent markers that become adjacent by sequentially thinning out the markers according to the predetermined rules. For example, for the five markers 1, 3, 4, 5, and 6 that become consecutive by thinning out marker 2 of the selected chromosome, k tuples consisting of d(1,3) / d(3,4), d(3,4) / d(4,5), and d(4,5) / d(5,6) are registered in the index.

[0064] Furthermore, when registering ratio values ​​in the index, the index building unit 321 may perform an operation to treat similar values ​​as the same value by, for example, using a method such as binning to forcibly consider the third decimal place and beyond as zero, in order to absorb the errors in the measurement data 340, which inevitably contain experimental errors.

[0065] Furthermore, when the index building unit 321 calculates the ratio between adjacent first and second marker intervals, if the first marker interval is significantly larger than the second marker interval, the ratio value will be greatly greater than 1.0, while if it is smaller, the ratio value will be less than 1.0, resulting in an asymmetry dependent on the relative sizes. To prevent this, the index building unit 321 may use the logarithm of the ratio (for example, a logarithm with a predetermined base, such as a common logarithm or a natural logarithm) instead of the ratio value.

[0066] Furthermore, since a k-tuple represents the ratio of k marker intervals, the index construction unit 321 may not directly calculate the ratio of adjacent marker intervals, but rather represent it with integer values ​​such that the sum of the ratio values ​​matches the integer value given as a parameter. For example, if the integer value given as a parameter is 100 and k=3, and the ratios of three adjacent marker intervals are 3000, 3000, and 4000 respectively, this k-tuple may be represented by integers such as 30:30:40, whose sum is 100.

[0067] Figure 5 shows an example of the data structure of the index constructed in step S405. Index 701 shows, for example, the correspondence between a k-tuple and the corresponding location on the reference genome. Index 701 is stored, for example, in the same storage device as the reference genome label location 330.

[0068] For example, record 7011 shows that the k-tuples for "markers 986-990" of "chromosome 9" and the k-tuples for "markers 1532-1536" of "chromosome 10" are both "1.12, 0.98, 0.88". Also, for example, record 7012 shows that the k-tuples for "markers 312-317" (i.e., markers 312, 313, 314, 316, 317) of "chromosome 15" after "marker 315" has been thinned out are "0.54, 0.99, 1.21".

[0069] <Index Search> Figure 6 is a flowchart showing an example of the index search process. The index search unit 322 constructs k-tuples based on the ratio of label intervals for any position of the DNA fragment indicated by the measurement data 340, and uses the index 701 created by the index construction unit 321 to identify the position on the reference genome corresponding to the DNA fragment indicated by the measurement data 340.

[0070] Step S501: The index search unit 322 receives the measurement data 340.

[0071] Step S502: If the index search unit 322 determines that all DNA fragments included in the input measurement data 340 have been selected, it terminates the index search process. If it determines that there are unselected DNA fragments, it proceeds to step S503.

[0072] Step S503: The index search unit 322 selects one unselected DNA fragment from the measurement data 340. The index search unit 322 may select DNA fragments in any order; for example, it may simply select DNA fragments in the order in which the information was entered into the measurement data 340.

[0073] Step S504: The index search unit 322 obtains the detected label positions of the DNA fragment selected in the most recent step S503 from the measurement data 340.

[0074] Step S505: The index search unit 322 calculates each label interval in the DNA fragment based on the label positions obtained in step S504, and calculates the ratio of adjacent label intervals. The index search unit 322 then compares each k-tuple, which is the ratio value of k consecutive elements in the DNA fragment, with each k-tuple in the index constructed in step S321, and obtains candidate positions on the corresponding reference genome.

[0075] Furthermore, in step S505, the index search unit 322, in order to address the false detection of labels in the measurement data 340, further thins out the labels from the selected DNA according to the predetermined rules, compares each of the k-tuples generated when the thinned-out labels are not present with each of the k-tuples in index 701, and obtains candidate positions on the corresponding reference genome.

[0076] Therefore, in step S505, a correspondence is generated between the labeling position indicated by the k-tuple in the selected DNA fragment and candidate positions on the reference genome.

[0077] The details of the index reference process in step S505 will be described later with reference to Figure 7. Furthermore, in step S505, the index search unit 322 checks the variable p max Initialize to 0. Since there can be multiple candidate locations on the reference genome corresponding to each k-tuple in the DNA fragment, process them sequentially using the following procedure.

[0078] Step S506: If the index search unit 322 determines that all candidate locations on the corresponding reference genome obtained in step S505 have been processed, it proceeds to step S511; if it determines that there are unprocessed candidate locations, it proceeds to step S507.

[0079] Step S507: The alignment probability calculation unit 323 calculates the probability p that the labeling positions indicated by each k-tuple of DNA fragments corresponding to each candidate position on the reference genome are aligned. Details of the calculation procedure for the alignment probability p will be described later with reference to Figure 8.

[0080] Step S508: The alignment probability calculation unit 323 calculates p > p max If it is determined that p ≤ p, the process proceeds to step S509. max If it is determined that this is the case, the process returns to step S506.

[0081] Step S509: The alignment probability calculation unit 323 records the candidate positions being selected, p max Substitute p into the equation and return to step S506. If a candidate position has already been recorded, the alignment probability calculation unit 323 will overwrite the candidate position.

[0082] Step S510: The index search unit 322 outputs candidate positions from the recorded data and terminates the index search process.

[0083] Although steps S506 to S508 describe a method for outputting the smallest reference genome location with the highest probability, it is also possible to add a procedure for outputting multiple candidates, not just the single location with the highest probability.

[0084] Figure 7 is a flowchart illustrating an example of the process of searching for index 701 while thinning out some of the labels in step S505. In the example in Figure 7, the predetermined rule for thinning out some of the labels is to sequentially thin out any one of the labels contained in the selected DNA fragment.

[0085] Step S1301: The index search unit 322 sets the variable n to the number of labels observed in the selected DNA fragment of the measurement data 340.

[0086] Step S1302: If n < k + 2, the interval between labels in the DNA is k or less, and the ratio of the label intervals becomes k - 1 or less. Therefore, since the index search unit 322 cannot refer to the index 701, the process of FIG. 7 ends. If n ≥ k + 2, the index search unit 322 transitions to step S1303.

[0087] Step S1303: The index search unit 322 initializes the variable i to 1.

[0088] Step S1304: The index search unit 322 calculates a total of k ratios from a total of k + 1 label intervals at labels i to i + k + 1 without thinning out the labels from the selected DNA fragment to generate a k-tuple, and refers to the index 701. For example, in the process of referring to the index 701, the index search unit 322 acquires a k-tuple that matches the generated k-tuple from the index 701, and acquires the position on the reference genome corresponding to the acquired k-tuple in the index 701 as the candidate.

[0089] Step S1305: If i ≥ n - k - 1, since the label interval becomes k + 1 or less when the labels are thinned out, the index search unit 322 ends the process of FIG. 7. If i < n - k - 1, the index search unit 322 transitions to step S1306.

[0090] Step S1306: The index search unit 322 initializes the variable j to 2. The reason for not initializing the variable j to 1 is that the leftmost label does not need to be thinned out (even if the leftmost label is thinned out, no new k-tuple is generated).

[0091] Step S1307: If j < k + 2, the index search unit 322 transitions to step S1310. If j ≥ k + 2, the index search unit 322 transitions to step S1308.

[0092] Step S1308: The index search unit 322 calculates a total of k ratios from a total of k+1 label intervals obtained when label i~i+k+2 of the selected DNA fragment is considered to have no label i+j-1, generates a k tuple, and references index 701. That is, the index search unit 322 references index 701 with the k tuple obtained by thinning out label i+j-1.

[0093] Step S1309: The index search unit 322 updates the indicator to be thinned out to the next indicator by adding 1 to the variable j, and returns to step S1307.

[0094] Step S1310: The index search unit 322 updates the position of the k-tuple of the selected DNA fragment that it wants to reference at index 701 by adding 1 to the variable i, and returns to step S1304.

[0095] Furthermore, the index registration process for the reference genome in step S405 can also be implemented using the same method as shown in Figure 7. Specifically, one can replace "DNA fragment" with "chromosome" in the process in Figure 7, and replace the process of referencing index 701 in Figure 7 with a process of registering the generated k-tuple with its position on the reference genome in index 701.

[0096] Figure 8 is an explanatory diagram showing an example of a DNA fragment from which labels have been thinned in measurement data 340. In the example in Figure 8, it is assumed that five labels, labels 1 to 5, were observed in the DNA fragment from measurement data 340. If k=2, the processing in Figure 7 generates a k-tuple for the case where no labels are thinned from the DNA fragment, a k-tuple for the case where label 2 is thinned from the DNA fragment, a k-tuple for the case where label 3 is thinned from the DNA fragment, and a k-tuple for the case where label 4 is thinned from the DNA fragment.

[0097] <Calculation of alignment probability> FIG. 9 is a flowchart showing an example of the alignment probability calculation process in step S507. Hereinafter, let m be the aligned part of the DNA fragment for which probability calculation is to be performed, and P(m) be the probability that the alignment occurs (p in steps S507 to S509).

[0098] In this embodiment, the alignment probability calculation unit 323 calculates P(m), for example, by using a probability model represented by P(m)=P scale w1 ·P pos w2 ·P ins w3 ·P del w4 P scale 、P pos 、P ins 、and P del are the probability of the stretch ratio indicating the ratio between the observed molecular length of the DNA fragment and the molecular length on the reference genome, the probability of the deviation of the observed label interval (from the interval on the genome), the probability of false detection of the label, and the probability of missed detection of the label, respectively. Also, w1, w2, w3, and w4 are weights, and w1 = w2 = w3 = w4 = 1 may be used unless otherwise specified by the user.

[0099] Note that in the above example, P(m)=P scale w1 ·P pos <00​​​​​​​​​​​​​​​​​​​​​​​​​​​​It is calculated using (|extension ratio - 1|). For example, f scale (x) has a mean of 0 and a variance of σ. scale 2 The probability density function of the normal distribution is σ scale This is determined from real data given in advance.

[0101] The alignment probability calculation unit 323 is P pos For example, P pos =f pos (|x1-y1|)f pos (|x2-y2|)···f pos (|x n -y n It is calculated by |). For example, f pos (x) is a normal distribution N(0,σ pos 2 ) is the probability density function, and σ pos 2 x is determined from pre-given real data. i y is the i-th label interval in the DNA fragment. i This is the marker interval of the i-th chromosome as indicated by the corresponding reference genome.

[0102] The alignment probability calculation unit 323 is P ins For example, the number of false detections of a sign n ins Based on n ins The calculation is performed using a function that decreases monotonically as x increases. Specifically, for example, the alignment probability calculation unit 323 uses the exponential function exp(x) to calculate P ins (n ins )=R ins exp(-n ins ) by P ins It is possible to calculate this. For example, R ins This is a constant coefficient, determined from pre-given real data.

[0103] Similarly, the alignment probability calculation unit 323 calculates P del For example, the number of times the label was not detected n del Using P del (n del )=Rdel exp(-n del It can be calculated using the following method. The calculation process for the probability of alignment occurring based on the above definition will now be explained.

[0104] Step S601: The alignment probability calculation unit 323 obtains the correspondence between the candidate alignment destination obtained as a result of the reference processing in step S505, i.e., the labeling position of the selected DNA fragment and the labeling position on the reference genome. In other words, for example, the alignment probability calculation unit 323 identifies the region where the k-tuple matches between the selected DNA fragment and the reference genome indicated by index 701.

[0105] Step S602: The alignment probability calculation unit 323 calculates the stretching rate of the entire alignment. Specifically, for example, the alignment probability calculation unit 323 calculates the stretching rate of the entire alignment by calculating the distance between the label positions at both ends for each of the DNA fragment selected in step S503 and the reference genome (chromosome) indicated by the correspondence obtained in step S601, and then finding the ratio of the calculated distances between the label positions at both ends.

[0106] Step S603: The alignment probability calculation unit 323, in a loop with step S604, calculates the difference between all label intervals of the DNA fragments shown by the correspondence obtained in step S601 and the label intervals of the reference genome |x i -y i The | is calculated for each. In step S603, the alignment probability calculation unit 323 determines whether all marker intervals have been processed. If they have been processed, the unit proceeds to S605; if any have not been processed, the unit proceeds to step S604.

[0107] Step S604: The alignment probability calculation unit 323 calculates the corresponding set of marker intervals x i , y i Regarding the difference in the spacing between signs |x i -y i Calculate | and return to step S603.

[0108] Step S605: The alignment probability calculation unit 323, in a loop with step S606, counts the number of false detections of labels on the DNA fragments, assuming that the label positions on the DNA fragments indicated by the correspondence obtained in step S601 match the positions on the reference genome. Here, a false detection refers to a label position on the DNA fragment that does not correspond to a position on the reference genome. If the alignment probability calculation unit 323 has processed all labels for measurement data not present in the reference genome, it proceeds to step S607; if there are any unprocessed labels for measurement data not present in the reference genome, it proceeds to step S606.

[0109] Step S606: The alignment probability calculation unit 323 increments the number of false positives by 1 and returns to step S605. Note that the number of false positives is initially initialized to 0.

[0110] Step S607: The alignment probability calculation unit 323, in a loop with step S608, counts the number of undetected labels on the DNA fragments, assuming that the label positions on the DNA fragments indicated by the correspondence obtained in step S601 match the positions on the reference genome. However, undetected labels here refer to label positions on the reference genome that cannot be associated with the labels on the DNA fragments. If all labels on the reference genome that are not in the measurement data have been processed, the alignment probability calculation unit 323 proceeds to step S609; if there are any unprocessed labels on the reference genome that are not in the measurement data, it proceeds to step S608.

[0111] Step S608: The alignment probability calculation unit 323 increments the number of undetected objects by 1 and returns to step S607. Note that the number of undetected objects is initialized to 0 beforehand.

[0112] Step S609: The alignment probability calculation unit 323 calculates P(m) by substituting the values ​​obtained in the above processes (the expansion / contraction ratio obtained in step S602, the deviation in the marking interval obtained in step S604, the number of false detections obtained in step S606, and the number of undetected items obtained in step S608) into the definition formula for P(m), and then terminates the alignment probability calculation process.

[0113] Furthermore, it is preferable that the parameters determined from the actual data are not common values ​​across the entire genome, but rather set individually for each region of each chromosome. Through the above processing, the genome labeling position alignment device of this embodiment becomes capable of aligning k-tuples based on the ratio of labeling intervals in each DNA shown in the measurement data 340 with k-tuples based on the ratio of labeling intervals of the reference genome.

[0114] In summary, the genome labeling position alignment device of this embodiment can perform alignment that can handle the stretching and shrinking of DNA fragments during measurement by performing alignment using k-tuples based on the ratio of labeling intervals.

[0115] Furthermore, the genome label position alignment device registers k-tuples constructed by thinning out some of the labels on the chromosomes of the reference genome to index 701. By comparing the k-tuples generated by thinning out the label positions of the DNA fragments indicated by the measurement data 340 with the k-tuples indicated by index 701, the device can achieve alignment that can address false detections of labels on DNA fragments and missed detections of labels on DNA fragments.

[0116] Furthermore, in the above process, the genome labeling position alignment device is P scale , P pos , P ins , and P delBy determining the corresponding position on the reference genome based on the alignment probability using each method, it is possible to achieve alignment that can address stretching and shrinking of DNA fragments during measurement, displacement of label positions on DNA fragments, false detection of labels on DNA fragments, and missed detection of labels on DNA fragments. Furthermore, the genome label position alignment device can provide a means to evaluate the reliability of partial alignment caused by DNA molecule cleavage or structural mutations during sample preparation and then perform the alignment. [Examples]

[0117] In Example 1, the genome labeling position alignment device can align k-tuples based on the labeling interval ratio between the DNA fragments of measurement data 340 and the reference genome. However, in this example, the genome labeling position alignment device extends the alignment by sequentially associating labels around the k-tuples registered in index 701.

[0118] Figure 10 is a sequence diagram showing an example of the overall processing using a genome labeling position alignment device. In the processing shown in Figure 10, the process in which the index construction unit 321, the index search unit 322, and the alignment probability calculation unit 323 work in coordination while referring to the reference genome labeling position 330 and the measurement data 340 is explained. It is expected that, assuming there are no DNA molecule cleavages or structural mutations during sample preparation, the entire molecule indicated by the measurement data 340 can be aligned with the reference genome through the processing shown in Figure 10. Note that the processing up to step S1004 is the same as in Example 1.

[0119] Step S1001: The index building unit 321 receives the genome sequence as input.

[0120] Step S1002: The index construction unit 321 constructs an index 701 using k tuples based on the ratio of the indicator intervals in steps S401 to S405.

[0121] Step S1003: The index search unit 322 acquires the measurement data 340 and obtains the labeling positions of the DNA fragments contained in the acquired measurement data 340.

[0122] Step S1004: The index search unit 322 searches for a location on the reference genome where the k-tuple matches, by referring to index 701 while thinning out the labels of some of the DNA fragments in steps S501 to S505.

[0123] Step S1005: The index search unit 322 also compares the ratio of the label intervals between the DNA fragment and the reference genome for labels around the label positions where the k-tuple matched in step S1004.

[0124] Step S1006: The alignment probability calculation unit 323 calculates the alignment probability, taking into account the comparison result of the surrounding sign spacing in step S1005.

[0125] Step S1007: The alignment probability calculation unit 323 outputs the alignment probability calculated in step S1006 to the index search unit 322.

[0126] Step S1008: The index search unit 322 determines the optimal alignment based on the alignment probability and outputs the determined alignment to the input / output device 302.

[0127] The details of steps S1005 to S1008 will be described later using Figures 12 and 13.

[0128] Figure 11 is an explanatory diagram showing an example of a tree structure constructed by extending index 701 to compare the ratio of the marker spacing around k tuples. In this embodiment, index 701 is extended, and two tree structures are constructed for each k tuple by the index construction unit 321.

[0129] For each k-tuple, one tree structure represents the ratio of the label spacings upstream (closer to the beginning of the string representing the chromosome) of that k-tuple in the reference genome, and the other tree structure represents the ratio of the label spacings downstream (closer to the end of the string representing the chromosome).

[0130] In each tree structure, the node closest to the root represents the ratio of the marker spacing adjacent to the k-tuple, and the child nodes represent the ratio of the marker spacing adjacent to the parent node. That is, in a tree structure that shows the ratio of marker spacing upstream, it represents the ratio of adjacent marker spacing upstream, and in a tree structure that shows the ratio of marker spacing downstream, it represents the ratio of adjacent marker spacing downstream.

[0131] Note that the ratio values ​​shown by k-tuples and tree structures are considered identical if they are as close as possible using methods such as binning. Therefore, the same k-tuple may appear at multiple locations on the reference genome, and because the ratio of adjacent marker intervals differs at each location, a single parent node may have multiple child nodes in the tree structure. This is why the data structure showing the ratio of marker intervals around the marker interval shown by a k-tuple becomes a tree structure.

[0132] The index 701 and tree structures 1111 and 1112 constructed by the index construction unit 321 from the marker position 330 of the reference genome in the example shown in Figure 11 will be explained in detail. Record 1110 of index 701 indicates a k-tuple consisting of a combination of ratios of three adjacent marker intervals of "1.78", "1.34", and "0.97", and the corresponding position on the reference genome. Tree structure 1111 indicates the ratio of the marker intervals upstream of the k-tuple indicated by record 1110, and tree structure 1112 indicates the ratio of the marker intervals downstream of the k-tuple indicated by record 1110.

[0133] For example, tree structure 1111 is constructed by sequentially searching for the ratio of the marker interval adjacent to the upstream side of "1.78" for each position on the reference genome corresponding to the k-tuple indicated by record 1110, and tree structure 1112 is constructed by sequentially searching for the ratio of the marker interval adjacent to the downstream side of "0.97".

[0134] In tree structure 1111, the node closest to the root node is "0.87". This indicates that for all the reference genome locations corresponding to the k-tuple indicated by record 1110, the ratio of the marker spacing adjacent to "1.78" upstream is "0.87".

[0135] Furthermore, in tree structure 1111, there are two child nodes of "0.87": one indicating "1.03" and another indicating "1.04". Therefore, at the location on the reference genome corresponding to the k-tuple indicated by record 1110, the ratio of the two label intervals upstream of "1.78" (i.e., the ratio of the label intervals adjacent to the upstream side of "0.87") is either "1.03" or "1.04".

[0136] Similarly, by calculating the ratio of the label intervals upstream of the location on the reference genome, we obtain nodes indicating "0.61" and "0.94" as child nodes of "1.03", and a node indicating "1.78" as a child node of "1.04" (assuming that no node indicating "0.94" as a child node of "1.04" is obtained).

[0137] Here, if the tree structure 1111 includes a group of nodes that share the same parent node and have similar values, such as "1.03" and "1.04" (for example, the difference is within a predetermined value), then, for example, an edge 1113 may be added from one node in that group to a child node of another node in that group, or the search process shown in Figure 12, described later, may be performed assuming that such an edge virtually exists. By adding edge 1113, even if there is a small error in the ratio of the calculated indicator intervals, the likelihood of obtaining a correct alignment by searching along edge 1113 increases.

[0138] The genome labeling position alignment device of this embodiment uses the tree structure shown in Figure 11 described above to compare the ratio of labeling intervals of DNA fragments indicated by measurement data 340 with the ratio of labeling intervals on the reference genome around the k-tuple aligned in the procedure of Example 1, and generates the optimal alignment. The method is described below.

[0139] To ensure that the alignment is optimal, the genome labeling position alignment device sequentially calculates the probability of generating the alignment using the alignment probability calculation unit 323. By sequentially examining the node that can generate the optimal alignment at that point, i.e., the alignment that maximizes the probability p calculated by the alignment probability calculation unit 323, the optimal alignment can ultimately be obtained. More precisely, the calculation is performed according to the procedure shown in Figures 12 and 13.

[0140] Figure 12 is a flowchart showing an example of alignment processing using a tree structure.

[0141] Step S1401: The index search unit 322 initializes the set S with {(root node of the tree, 0)}. The index search unit 322 also initializes the set T with an empty set. The first value (left element) contained in each element of set S is called a node, and the second value (right element) is called a score. The calculation proceeds so that set S becomes the set of nodes in the search in progress, and T becomes the set of nodes corresponding to indicators that indicate the end of the search.

[0142] Step S1402: The index search unit 322 proceeds to step S1410 if set S is an empty set, and to step S1403 if set S is not an empty set.

[0143] Step S1403: The index search unit 322 selects the element with the highest score from the set S and removes that selected element from the set S. Let the selected element be (v,p).

[0144] Step S1404: The index search unit 322 returns to step S1402 if it has processed all of v's child nodes, and proceeds to step S1405 if there are any unprocessed child nodes of v.

[0145] Step S1405: The index search unit 322 selects one child node of v and denotes the selected node as u.

[0146] Step S1406: The index search unit 322 selects a marker adjacent to the processed marker (the marker corresponding to v) in the measurement data 340 as the marker to be associated with u.

[0147] Step S1407: The index search unit 322 and the alignment probability calculation unit 323 perform an update process on sets S and T for u and the marker selected in step S1406. Details of the update process for sets S and T will be described later with reference to Figure 13.

[0148] Step S1408: The index search unit 322 selects a label further adjacent to the label corresponding to u in order to address the case where the label corresponding to u is not detected in the DNA fragment being processed.

[0149] Step S1409: The index search unit 322 and the alignment probability calculation unit 323 consider the marker selected in step S1408 as a new u and perform the update process for sets S and T described later using Figure 13, and after execution, return to step S1404.

[0150] Step S1410: The index search unit 322 identifies the node with the highest score in the set T as u. The index search unit 322 also outputs the alignment corresponding to node u, that is, the alignment that associates the markers on the genome and the markers on the DNA fragments corresponding to each node selected by traversing the tree structure to finally reach u, as the optimal alignment.

[0151] Figure 13 is a flowchart showing an example of the update process for sets S and T performed in alignment processing using a tree structure.

[0152] Step S1501: The alignment probability calculation unit 323 calculates the probability of the alignment corresponding to the selected node u in the same manner as in step S507, as shown in Figure 9, and denotes the calculated probability as q.

[0153] Step S1502: The index search unit 322 determines whether the labeled position corresponding to node u is at the end of the DNA fragment molecule. If the index search unit 322 determines that the labeled position is at the end, it proceeds to step S1503; otherwise, it proceeds to step S1504.

[0154] Step S1503: The index search unit 322 adds (u,q) to the set T.

[0155] Step S1504: The index search unit 322 adds (u,q) to the set S. [Examples]

[0156] Example 3 provides a user interface for setting parameters to adjust the processing performed by the genome labeling site alignment device. This user interface also allows for the visualization of the alignment results.

[0157] Figure 14 shows an example of a user interface displayed on the input / output device 302. The user interface 200 includes, for example, an input data setting area 210, a parameter setting area 220, and an alignment result display area 230.

[0158] The input data setting area 210 is an area for setting the source of the reference genome labeling position 330 and the measurement data 340. In the example in Figure 14, the source is specified by a file name, but it may also be specified by a URL (Uniform Resource Locator) or the like, as needed.

[0159] The parameter setting region 220 is a region for setting parameters used in processing by the genome labeling position alignment device. In the parameter setting region 220, for example, the number of ratio values ​​(k) in a k-tuple can be set.

[0160] Furthermore, in the parameter setting area 220, it is possible to set, for example, the number of markers to be thinned out by the index construction unit 321 and the index search unit 322 in order to deal with detection omissions and false detections. In the example described above, the case where the number of markers to be thinned out is 1 was explained, but it is also possible to set the number of markers to be thinned out to 2 or more. Note that increasing the number of markers to be thinned out will increase the data size of the index 701, but will also increase its error tolerance.

[0161] In the parameter setting area 220, for example, parameters w1, w2, w3, and w4 can be set for weighting errors, which are used by the alignment probability calculation unit 323 when calculating the probability P(m). By being able to set these weights in the parameter setting area 220, it becomes possible to tune the importance of various errors.

[0162] The alignment result display area 230 displays information that visualizes the alignment obtained as a calculation result. In the alignment result display area 230, the corresponding position on the reference genome for the DNA fragment of the measurement data 340 is displayed, and the corresponding markers between the two can be confirmed.

[0163] Furthermore, in the alignment result display area 230, it is possible to distinguish between labels that correspond to the DNA fragment on the reference genome, specifically those mapped using k-tuples (an example of a non-extended region, represented by a solid line in Figure 14) and surrounding regions mapped using a tree structure (an example of an extended region, represented by a dotted line in Figure 14). In addition, the alignment result display area 230 can assist users in verifying the calculation results by showing the ratio of the label spacing and the overall molecular stretching ratio of the DNA fragment. [Examples]

[0164] In this embodiment, the genome labeling site alignment system detects structural mutations present in the subject's genome. The genome labeling site alignment system includes a genome mapping device and the genome labeling site alignment device described in Examples 1 to 3.

[0165] Figure 15 is an explanatory diagram showing an example of a structural mutation detection process. The genome mapping device 1200 acquires genomic DNA collected from a subject. The genome mapping device 1200 performs genome amplification and fragmentation on the acquired subject's genomic DNA. Furthermore, the genome mapping device 1200 obtains measurement data 340 by measuring the position of the label in each DNA fragment.

[0166] The genome labeling position alignment device can identify the position on the reference genome corresponding to each label shown in the measurement data 340 by aligning the DNA fragments shown in the measurement data 340 with the reference genome labeling position 330 through k-tuple-based alignment processing using the labeling interval ratio described in Examples 1 and 2.

[0167] The genome labeling alignment device determines whether there is a structural mutation in the subject's genome based on a comparison between the labeling position on the DNA fragment indicated by the measurement data 340 and the corresponding labeling position in the reference genome on the DNA fragment.

[0168] Specifically, for example, the genome labeling position alignment device can determine that there is a structural variation in the subject's genome if it determines that the labeling positions on the subject's genome are discontinuous (specifically, if the labels on the reference genome corresponding to the continuous labels on the DNA fragment are not continuous), or if it determines that the labeling interval is abnormally large or small (for example, if the difference between the labeling interval on the DNA fragment shown by measurement data 340 and the labeling interval on the reference genome corresponding to that labeling interval is greater than or less than a predetermined value). By outputting a list of the structural variations detected in this way to, for example, the input / output device 302, the user of the genome labeling position alignment device can comprehensively understand the structural variations present in the subject's genome.

[0169] The present invention is not limited to the embodiments described above, and includes various modifications. For example, the embodiments described above are described in detail to make the present invention easier to understand, and are not necessarily limited to those having all the configurations described. It is also possible to replace parts of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add configurations from other embodiments to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace parts of the configuration of each embodiment with other configurations.

[0170] Furthermore, each of the above configurations, functions, processing units, and processing means may be implemented in hardware, either partially or entirely, by designing them as integrated circuits, for example. Alternatively, each of the above configurations and functions may be implemented in software by having the processor interpret and execute programs that implement each function. Information such as programs, tables, and files that implement each function can be stored in memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0171] Furthermore, the control lines and information lines shown are those deemed necessary for explanatory purposes, and not all control lines and information lines are necessarily shown in the actual product. In reality, it is safe to assume that almost all components are interconnected. [Explanation of symbols]

[0172] 300 Computer, 302 Input / Output Device, 310 CPU, 311 Memory, 312 Auxiliary Storage Device, 313 Interface, 321 Index Construction Unit, 322 Index Search Unit, 323 Alignment Probability Calculation Unit, 330 Reference Genome Marking Position, 340 Measurement Data, 701 Index

Claims

1. An information processing device, Equipped with a processor and memory, The memory holds a first numerical sequence indicating the position of a subsequence in a referenced nucleic acid sequence, and a second numerical sequence indicating the measurement position of the subsequence in the target nucleic acid sequence. The aforementioned processor, Based on the first numerical sequence, a plurality of first ratios of the intervals between the subsequences in the referenced nucleic acid sequence are calculated. An index is constructed that shows the first combination of ratios and information indicating the position of a sub-sequence in the referenced nucleic acid sequence corresponding to the first combination of ratios. Based on the second numerical sequence, a number of second ratios of the intervals between the subsequences in the target nucleic acid sequence are calculated. Based on the comparison result between the second ratio combination and the first ratio combination indicated by the index, the first ratio combination corresponding to the second ratio combination is extracted. An information processing device that outputs information indicating the position of the subsequence corresponding to the extracted first ratio combination in the referenced nucleic acid sequence.

2. An information processing apparatus according to claim 1, The processor is an information processing device that calculates, based on the first numerical sequence, the ratio of the spacing between adjacent subsequences in the referenced nucleic acid sequence and the ratio of the spacing between subsequences that would become adjacent if some of the subsequences were thinned out from the referenced nucleic acid sequence according to a predetermined rule, as a plurality of first ratios.

3. An information processing apparatus according to claim 1, The processor is an information processing device that calculates, based on the second numerical sequence, the ratio of the spacing between adjacent subsequences in the target nucleic acid sequence and the ratio of the spacing between subsequences that would become adjacent if some of the subsequences were thinned out from the target nucleic acid sequence according to a predetermined rule, as a plurality of second ratios.

4. An information processing apparatus according to claim 1, The aforementioned processor, Identify the first ratio combination that matches the second ratio combination from the index, For each of the specified first ratio combinations, the probability that the specified first ratio combination corresponds to the second ratio combination is calculated based on a predetermined probability model. An information processing device that extracts a first ratio combination corresponding to the second ratio combination based on the calculated probability.

5. An information processing apparatus according to claim 4, An information processing device wherein the predetermined probability model is a model that reflects at least one of the following: the probability of the stretch ratio representing the ratio of the observed molecular length to the correct molecular length of the target nucleic acid sequence; the probability of the discrepancy between the measured position and the correct position of the subsequence in the target nucleic acid sequence; the probability of false detection of the subsequence in the target nucleic acid sequence; and the probability of failure to detect the subsequence in the target nucleic acid sequence.

6. An information processing apparatus according to claim 1, The aforementioned processor, With respect to the first combination of ratios indicated by the index, the first combination of ratios is extended based on the ratio of the intervals between subsequences adjacent to the subsequence in the referenced nucleic acid sequence corresponding to the index, The second combination of ratios is extended based on the ratio of the intervals between subsequences adjacent to a subsequence in the target nucleic acid sequence corresponding to the second combination of ratios, An information processing device that extracts a first ratio combination corresponding to the second ratio combination based on the comparison result between the expanded second ratio combination and the expanded first ratio combination.

7. An information processing apparatus according to claim 6, The aforementioned processor, Identify the extended first ratio combination that matches the extended second ratio combination, In the extended portion of the extended first ratio combination that matches the extended second ratio combination, information indicating the position of the subsequence of the referenced nucleic acid sequence indicated by the first ratio that matches the second ratio, An information processing device that outputs information indicating the position of a subsequence of the referenced nucleic acid sequence represented by the first ratio that matches the second ratio, in the unextended portion of the extended first ratio combination that matches the extended second ratio combination.

8. An information processing apparatus according to claim 1, Connected to the input device, The aforementioned processor, An information processing device that receives input via the input device the number of first ratios included in each of the first combinations of ratios, and the number of second ratios included in each of the second combinations of ratios.

9. An information processing apparatus according to claim 1, The aforementioned processor, Based on the comparison between the measurement position of the subsequence in the target nucleic acid sequence indicated by the second ratio combination and the position of the subsequence in the reference nucleic acid sequence indicated by the extracted first ratio combination, it is determined whether there is a structural mutation in the target nucleic acid sequence. An information processing device that, when it is determined that a structural mutation exists in the target nucleic acid sequence, outputs information indicating the structural mutation.

10. An information processing method using an information processing device, The aforementioned information processing device comprises a processor and memory, The memory holds a first numerical sequence indicating the position of a subsequence in a referenced nucleic acid sequence, and a second numerical sequence indicating the measurement position of the subsequence in the target nucleic acid sequence. The aforementioned information processing method is The processor calculates a plurality of first ratios of the intervals between the subsequences in the referenced nucleic acid sequence based on the first numerical sequence, The processor constructs an index that indicates the first combination of ratios and information indicating the position of a sub-sequence in the referenced nucleic acid sequence corresponding to the first combination of ratios. The processor calculates a plurality of second ratios of the intervals between the subsequences in the target nucleic acid sequence based on the second numerical sequence. The processor extracts a first ratio combination corresponding to the second ratio combination based on the comparison result between the second ratio combination and the first ratio combination indicated by the index. An information processing method comprising: the processor outputting information indicating the position of the sub-sequence corresponding to the extracted first ratio combination in the referenced nucleic acid sequence.

Images