A DNA sequence splitting method and device

Abstract

The present application provides a method and device for splitting a DNA sequence, the method comprising: determining a search region in a DNA sequence according to a preset oligonucleotide length; intercepting a DNA sequence in the search region according to a preset sequence length to obtain a DNA subsequence group; scoring the DNA subsequence group through a deep learning model, and determining the DNA subsequence with the highest score from the DNA subsequence group as a target sequence according to the scoring result; adjusting the number of bases of the target sequence according to the principle of similar DNA denaturation temperature to obtain a base complementary pairing region between single-stranded oligonucleotides when the DNA sequence is split; splitting the DNA sequence according to the base complementary pairing region to obtain a single-stranded oligonucleotide. Through the method provided by the present application, the base complementary pairing region with the highest score can be obtained from the DNA sequence, and the DNA sequence can be split, thereby improving the assembly efficiency of PCA assembly.

Description

Technical Field

[0001] The present application relates to the field of DNA splitting, and in particular to a method and device for splitting a DNA sequence. Background Art

[0002] In the process of DNA synthesis, chemically synthesized oligonucleotide monomers need to be connected into single-stranded oligonucleotides (oligonucleotides, oligo), and then the single-stranded oligonucleotides are assembled into double-stranded DNA through polymerase cycling assembly (PCA) based on the principle of base complementary pairing. During PCA assembly, there are base complementary pairing regions between adjacent oligos. During the PCA reaction, the thermal motion of nucleotide molecules will produce random collisions between base complementary pairing regions. When the base complementary pairing regions of complementary pairing are combined, a DNA double-stranded structure is formed. Then, DNA polymerase uses the single-stranded region of the non-base complementary pairing region as a template to catalyze the oligonucleotide monomers to complete the double-stranded DNA to form a complete double-stranded DNA. In the PCA assembly process, whether the base complementary pairing region can be correctly combined is the basis for whether the double-stranded DNA can be correctly formed. Therefore, once the base complementary pairing region of the DNA sequence is determined, the splitting scheme of PCA assembly can be designed according to the base complementary pairing region.

[0003] The DNA splitting methods currently used include DNAworks, GeneDesign, FastPCR, etc., all of which consider the effects of biological parameters such as DNA denaturation temperature, repetitive sequence, and secondary structure in the base complementary pairing region on the PCA reaction process. For example, the commonly used GeneDesign algorithm sets the length of the oligo to be approximately 60 nucleotides (nt), the length of the base complementary pairing region to be 20nt, and the DNA denaturation temperature of the base complementary pairing region to be 56°C. Then, according to the set biological parameters, the splitting design is completed by adjusting the position of the base complementary pairing region. However, this type of splitting method is highly dependent on the selection of empirical parameter thresholds. Different parameter threshold selections will result in different splitting schemes and different assembly efficiencies. There is a lack of data-driven model learning and regularity discovery, so the assembly efficiency of PCA cannot be guaranteed. Summary of the invention

[0004] In view of this, the present application provides a method and device for splitting a DNA sequence, the purpose of which is to screen out the optimal base complementary pairing region of the DNA sequence, thereby improving the PCA assembly efficiency.

[0005] The first aspect of the present application provides a method for splitting a DNA sequence, the method comprising:

[0006] Determine the search region in the DNA sequence according to the preset single-stranded oligonucleotide length;

[0007] According to a preset sequence length, the DNA sequence is intercepted in the search region to obtain a DNA subsequence group;

[0008] Scoring the DNA subsequence group by using a deep learning model, and determining the DNA subsequence with the highest score from the DNA subsequence group as the target sequence according to the scoring result;

[0009] The base number of the target sequence is adjusted according to the principle of similar DNA denaturation temperature to obtain a base complementary pairing region of the DNA sequence;

[0010] The DNA sequence is split according to the base complementary pairing regions to obtain oligonucleotide fragments.

[0011] Optionally, scoring the DNA subsequence group by a deep learning model comprises:

[0012] Encoding the DNA subsequence group using a preset encoding method to obtain a DNA sequence matrix, wherein the DNA sequence matrix includes a plurality of DNA subsequences;

[0013] Inputting the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the multiple DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly (PCA) experiment;

[0014] The probability that the multiple DNA subsequences are the base complementary pairing regions is used as the scores of the multiple DNA subsequences.

[0015] Optionally, the step of inputting the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the multiple DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly (PCA) experiment comprises:

[0016] Inputting the multiple DNA subsequences into the preset base complementary pairing region classification model, wherein the preset base complementary pairing region classification model includes a convolutional layer, a maximum pooling layer and a fully connected layer;

[0017] Extracting sequence features of the multiple DNA subsequences through the convolution layer according to a preset size standard to obtain first sequence features of the multiple DNA subsequences;

[0018] Performing feature dimension reduction processing on the first sequence features of the multiple DNA subsequences through the maximum pooling layer to obtain multiple second sequence features;

[0019] The multiple second sequence features are combined through the fully connected layer, and the probability that the multiple DNA subsequences are base complementary pairing regions of the PCA experiment is calculated based on the combined second sequence features.

[0020] Optionally, the method for constructing the preset base complementary pairing region classification model includes:

[0021] Obtain a pre-designed DNA sample sequence;

[0022] Extracting base complementary pairing regions and non-base complementary pairing regions from the DNA sample sequence;

[0023] Cutting the DNA sample sequence in the base complementary pairing region of the DNA sample sequence to obtain a first sample;

[0024] According to a preset sequence length, the DNA sample sequence is intercepted in a non-base complementary pairing region of the DNA sample sequence to obtain a second sample;

[0025] The base complementary pairing region classification model is constructed according to the first sample and the second sample.

[0026] Optionally, the preset base complementary pairing region classification model includes at least one of the following: a convolutional neural network model, a recurrent neural network model, and a multi-layer perceptron model.

[0027] Optionally, the search region is multiple, and the base number of the target sequence is adjusted according to the principle of similar DNA denaturation temperature to obtain the base complementary pairing region of the DNA sequence, including:

[0028] Calculating the DNA denaturation temperatures of the plurality of target sequences corresponding to the plurality of search regions to obtain a first DNA denaturation temperature array;

[0029] If the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is smaller than a preset threshold, the target sequence is determined as the base complementary pairing region of the DNA sequence.

[0030] Optionally, the method further includes:

[0031] If the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is greater than or equal to a preset threshold, then calculating the absolute value of the difference between each DNA denaturation temperature in the first DNA denaturation temperature array and the median to obtain an absolute value set, wherein the median is the median of the first DNA denaturation temperature array;

[0032] Adjusting the number of bases of the target sequence corresponding to the maximum absolute value in the absolute value set to obtain an adjusted target sequence;

[0033] Calculating the adjusted DNA denaturation temperature of the target sequence to obtain a second DNA denaturation temperature array;

[0034] If the difference between the maximum value and the minimum value in the second DNA denaturation temperature array is smaller than a preset threshold, the adjusted target sequence is determined as the base complementary pairing region of the DNA sequence.

[0035] Optionally, adjusting the number of bases of the target sequence corresponding to the maximum absolute value in the absolute value set to obtain the adjusted target sequence includes:

[0036] If the DNA denaturation temperature of the target sequence is greater than the median, the number of bases of the target sequence is reduced according to a preset base number adjustment value to obtain the adjusted target sequence;

[0037] If the DNA denaturation temperature of the target sequence is less than the median, the number of bases of the target sequence is increased according to the base number adjustment value to obtain the adjusted target sequence.

[0038] A second aspect of the present application provides a DNA sequence splitting device, the device comprising: a sequence processing module, a scoring module, an adjustment module and a splitting module;

[0039] The sequence processing module is used to determine the search region in the DNA sequence according to the preset oligonucleotide length;

[0040] The sequence processing module is used to intercept the DNA sequence in the search area according to a preset sequence length to obtain a DNA subsequence group;

[0041] The scoring module is used to score the DNA subsequence group through a deep learning model, and determine the DNA subsequence with the highest score from the DNA subsequence group as the target sequence according to the scoring result;

[0042] The adjustment module is used to adjust the number of bases in the target sequence according to the principle of similar DNA denaturation temperature to obtain the base complementary pairing region of the DNA sequence;

[0043] The splitting module is used to split the DNA sequence according to the base complementary pairing region to obtain oligonucleotide fragments.

[0044] Optionally, the scoring module is specifically used to encode the DNA subsequence group in a preset encoding manner to obtain a DNA sequence matrix, wherein the DNA sequence matrix includes a plurality of DNA subsequences;

[0045] The scoring module is further used to input the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the multiple DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly (PCA) experiment;

[0046] The scoring module is further configured to use the probability that the multiple DNA subsequences are the base complementary pairing regions as the scores of the multiple DNA subsequences.

[0047] Optionally, the scoring module is further used to input the multiple DNA subsequences into the preset base complementary pairing region classification model, wherein the preset base complementary pairing region classification model includes a convolutional layer, a maximum pooling layer and a fully connected layer;

[0048] The scoring module is further used to extract sequence features of the multiple DNA subsequences through the convolution layer according to a preset size standard to obtain first sequence features of the multiple DNA subsequences;

[0049] The scoring module is further used to perform feature dimension reduction processing on the first sequence features of the multiple DNA subsequences through the maximum pooling layer to obtain multiple second sequence features;

[0050] The scoring module is further used to combine the multiple second sequence features through the fully connected layer, and calculate the probability that the multiple DNA subsequences are base complementary pairing regions of the PCA experiment based on the combined second sequence features.

[0051] Optionally, the scoring module is also used to obtain a pre-designed DNA sample sequence;

[0052] The scoring module is also used to extract base complementary pairing regions and non-base complementary pairing regions from the DNA sample sequence;

[0053] The scoring module is further used to intercept the DNA sample sequence in the base complementary pairing region of the DNA sample sequence to obtain a first sample;

[0054] The scoring module is further used to intercept the DNA sample sequence in the non-base complementary pairing region of the DNA sample sequence according to a preset sequence length to obtain a second sample;

[0055] The scoring module is further used to construct the base complementary pairing region classification model according to the first sample and the second sample.

[0056] The adjustment module is further used to calculate the DNA denaturation temperatures of the plurality of target sequences corresponding to the plurality of search regions to obtain a first DNA denaturation temperature array;

[0057] If the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is smaller than a preset threshold, the adjustment module is further configured to determine the target sequence as a base complementary pairing region of the DNA sequence.

[0058] Optionally, if the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is greater than or equal to a preset threshold, the adjustment module is further used to calculate the absolute value of the difference between each DNA denaturation temperature in the first DNA denaturation temperature array and the median to obtain an absolute value set, wherein the median is the median of the first DNA denaturation temperature array;

[0059] The adjustment module is further used to adjust the number of bases in the target sequence corresponding to the maximum absolute value in the absolute value set to obtain an adjusted target sequence;

[0060] The adjustment module is further used to calculate the adjusted DNA denaturation temperature of the target sequence to obtain a second DNA denaturation temperature array;

[0061] If the difference between the maximum value and the minimum value in the second DNA denaturation temperature array is smaller than a preset threshold, the adjustment module is further configured to determine the adjusted target sequence as the base complementary pairing region of the DNA sequence.

[0062] Optionally, if the DNA denaturation temperature of the target sequence is greater than the median, the adjustment module is further configured to reduce the number of bases of the target sequence according to a preset base number adjustment value to obtain the adjusted target sequence;

[0063] If the DNA denaturation temperature of the target sequence is less than the median, the adjustment module is further configured to increase the number of bases of the target sequence according to the base number adjustment value to obtain the adjusted target sequence.

[0064] The present application provides a method and device for splitting a DNA sequence, the method comprising: determining a search region in a DNA sequence according to a preset oligonucleotide length; intercepting the DNA sequence in the search region according to the preset sequence length to obtain a DNA subsequence group; scoring the DNA subsequence group through a deep learning model, and determining the DNA subsequence with the highest score from the DNA subsequence group as a target sequence according to the scoring result; adjusting the number of bases of the target sequence according to the principle of similar DNA denaturation temperature to obtain a base complementary pairing region of the DNA sequence; splitting the DNA sequence according to the base complementary pairing region to obtain oligonucleotide fragments. Through the method provided by the present application, the base complementary pairing region with the highest score can be obtained from the DNA sequence, and the DNA sequence can be split according to the base complementary pairing region with the highest score, thereby improving the assembly efficiency of PCA assembly. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A schematic diagram of a process for splitting a DNA sequence provided in an embodiment of the present application;

[0066] Figure 2 A schematic diagram of another method for splitting a DNA sequence provided in an embodiment of the present application;

[0067] Figure 3 A schematic diagram of another method for splitting a DNA sequence provided in an embodiment of the present application;

[0068] Figure 4 An example diagram of a DNA sequence splitting process provided in an embodiment of the present application;

[0069] Figure 5 A schematic diagram of the architecture of a deep learning model provided in an embodiment of the present application.

[0070] Figure 6 A schematic diagram of the structure of a DNA sequence splitting device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0072] like Figure 1As shown in the schematic diagram of the splitting process, the currently used DNA sequence synthesis method is to split the target sequence to be synthesized into single-stranded oligonucleotides with base complementary pairing regions through computer-assisted splitting, and then synthesize oligonucleotides based on the designed oligonucleotide single strands. The synthesized oligonucleotides are assembled by polymerase chain reaction through a thermal cycler to obtain the target sequence. The target design of the single-stranded oligonucleotide with a base complementary pairing region is based on empirical identification, taking into account empirical difficult synthesis factors, such as artificially extracted biological empirical parameters such as CG content, and then further trial and error in the experiment to obtain the target sequence. Since the DNA sequence is divided into base complementary pairing regions and non-base complementary pairing regions in the directional oligonucleotide design, the assembly efficiency of the base complementary pairing region needs to be considered when performing polymerase chain reaction assembly.

[0073] This application adopts a deep learning modeling approach to encode the DNA sequence to be split, extract sequence features of the encoded DNA sequence, input the extracted sequence features into a fully connected layer, predict the assembly efficiency of the base complementary pairing region of the DNA sequence, and then optimize the synthesis design and improve the synthesis assembly efficiency.

[0074] Combine the following Figure 1 , introducing a DNA sequence splitting method provided in the present application, which can be implemented through steps S101-S105.

[0075] S101: Determine a search region in a DNA sequence according to a preset oligonucleotide length.

[0076] Specifically, for a DNA sequence to be split, multiple search regions are determined in the DNA sequence according to a preset upper limit and lower limit of the length of the oligonucleotide oligo.

[0077] After determining the multiple search regions, the DNA sequence to be split is split into the multiple search regions, which can ensure that the target sequences corresponding to each search region are independent of each other.

[0078] S102: Cutting off a DNA sequence in the search region according to a preset sequence length to obtain a DNA subsequence group.

[0079] Specifically, according to a preset sequence length, DNA sequences are intercepted in multiple search regions to obtain a DNA subsequence group corresponding to each search region.

[0080] In the embodiment of the present application, a DNA subsequence is truncated with 18 nucleotides (nt). In actual application scenarios, the length of the truncated DNA subsequence can be adjusted according to actual conditions, and the preset sequence length is not limited here.

[0081] S103: Score the DNA subsequence group using a deep learning model, and determine the DNA subsequence with the highest score from the DNA subsequence group as the target sequence according to the scoring result.

[0082] Specifically, the extracted DNA subsequence group is input into the deep learning model for scoring. The deep learning model scores the DNA subsequence group based on the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment, and obtains the scoring result of each DNA subsequence in the DNA subsequence group. The DNA subsequence with the highest score in the DNA subsequence group is selected as the target sequence. Among them, the score of the DNA subsequence is used to identify the assembly efficiency of the DNA subsequence.

[0083] like Figure 4 As shown, the four DNA subsequences in the search area in 401 are input into the deep learning model for scoring, and the scoring results of the four DNA subsequences are output, which are 0.31, 0.2, 0.79, and 0.43 respectively. The DNA subsequence corresponding to 0.79 is selected as the target sequence in the second search area.

[0084] Through steps S102-S103, it is possible to determine the DNA subsequence with the highest score in the search area, and use the DNA subsequence as the target sequence in the search area.

[0085] S104: adjusting the number of bases in the target sequence according to the principle of similar DNA denaturation temperature to obtain a base complementary pairing region of the DNA sequence.

[0086] Specifically, according to the principle of similar DNA denaturation temperature (Tm), the base numbers of multiple target sequences are adjusted, and the adjusted target sequences are the base complementary pairing regions of the DNA sequences to be separated.

[0087] like Figure 4 As shown, multiple target sequences in 402 are adjusted according to the Tm similarity principle to obtain multiple base complementary pairing regions in 403.

[0088] After adjusting the target sequence according to the Tm similarity principle, the Tm of the obtained base complementary pairing region can be similar. In the PCA experiment, the similar Tm of the DNA sequence can increase the success rate of the PCA experiment and improve the assembly efficiency of the split DNA sequence.

[0089] S105: Splitting the DNA sequence according to the base complementary pairing regions to obtain oligonucleotide fragments.

[0090] Specifically, after the base complementary pairing region of the DNA sequence is determined, the DNA sequence is split into oligonucleotide fragments having single-stranded base complementary pairing regions according to the base complementary pairing regions.

[0091] like Figure 4 As shown, the fragment is split according to the multiple base complementary pairing regions in 403 to obtain four oligonucleotide fragments.

[0092] The base complementary pairing region obtained by adjusting the target sequence with the highest score is used to split the DNA sequence to obtain oligonucleotide fragments. Since the selected target sequences are all target sequences with the highest assembly efficiency, the assembly efficiency of the PCA assembly can be improved when the oligonucleotide fragments are used for PCA assembly.

[0093] Through the method provided in the embodiment of the present application, the base complementary pairing region with the highest score can be obtained from the DNA sequence, and the DNA sequence can be split according to the base complementary pairing region with the highest score, thereby improving the assembly efficiency of PCA assembly.

[0094] Combine the following Figure 2 , introducing another DNA sequence splitting method provided in the present application, which is used to implement the scoring of DNA subsequence groups through a deep learning model in step S103, which can be implemented through steps S201-S203.

[0095] S201: Encode the DNA subsequence group using a preset encoding method to obtain a DNA sequence matrix, where the DNA sequence matrix includes a plurality of DNA subsequences.

[0096] Specifically, after the DNA subsequence group is input into the deep learning model, the deep learning model encodes the DNA subsequence group using a preset encoding method to obtain a DNA sequence matrix, wherein the DNA sequence matrix includes multiple DNA subsequences.

[0097] like Figure 5 As shown in the coding sequence part of , in the embodiment of the present application, a hot encoding method is used to fill the DNA subsequence group into a 4x18 dimensional matrix to obtain a DNA sequence matrix.

[0098] It can be understood that, in this application, hot encoding is used as an example, and the encoding method and matrix dimension used in actual situations can be adjusted according to actual conditions, and this is not limited here.

[0099] By encoding DNA subsequence groups and generating DNA sequence matrices, the budget efficiency of deep learning models can be improved.

[0100] S202: Inputting the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the multiple DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly (PCA) experiment.

[0101] Specifically, the DNA sequence matrix is ​​input into a preset base complementary pairing region classification model through a deep learning model, and the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment is calculated through the base complementary pairing region classification model.

[0102] The base complementary pairing region classification model calculates the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment through the following steps S2021-S2024.

[0103] S2021: Input multiple DNA subsequences into a preset base complementary pairing region classification model.

[0104] Specifically, multiple DNA subsequences in the DNA sequence matrix are input into a preset base complementary pairing region classification model, wherein the preset base complementary pairing region classification model includes a convolutional layer, a maximum pooling layer and a fully connected layer.

[0105] S2022: Extracting sequence features of the multiple DNA subsequences through a convolutional layer according to a preset size standard to obtain first sequence features of the multiple DNA subsequences.

[0106] Specifically, the first sequence feature of the DNA subsequence is extracted by using a convolution kernel with a preset standard size built into the convolution layer.

[0107] like Figure 5 As shown in the convolution layer and splicing part, in the embodiment of the present application, three convolution kernels of sizes 3, 4, and 5 are used to extract the sequence features of the DNA subsequence, and the results obtained by the three convolution kernels of different sizes are spliced ​​together to obtain the first sequence feature of the DNA subsequence.

[0108] S2023: Perform feature dimension reduction processing on the first sequence features of the plurality of DNA subsequences through a maximum pooling layer to obtain a plurality of second sequence features.

[0109] Specifically, the first sequence features obtained by the convolution layer are input into the maximum pooling layer, which performs feature dimension reduction processing on the first sequence features to obtain multiple second sequence features.

[0110] like Figure 5 As shown in the concatenated part and the maximum pooling layer, the dimension of the first sequence feature is 18x300, and the dimension of the second sequence feature processed by the maximum pooling layer is 9x300.

[0111] The maximum pooling layer is used to reduce the feature dimension of the sequence features of the DNA subsequence, so that the subsequent calculation of the probability that the DNA subsequence is a base complementary pairing region in the PCA experiment based on the sequence features of the DNA subsequence is more accurate.

[0112] S2024: Combining the multiple second sequence features through a fully connected layer, and calculating the probability that the multiple DNA subsequences are base complementary pairing regions of the PCA experiment based on the combined second sequence features.

[0113] Specifically, multiple second sequence features obtained by the maximum pooling layer are input into the fully connected layer. The fully connected layer combines multiple second sequence features of the DNA subsequence, calculates the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment based on the combined second sequence features, and feeds back the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment to the deep learning model.

[0114] like Figure 5 As shown in the fully connected layer, multiple sequence features with dimensions of 9x300 output by the maximum pooling layer are assembled into a sequence feature with a dimension of 2700, and the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment is output based on the sequence feature. The probability of the embodiment in the figure is 0.79.

[0115] In a possible embodiment, when the number of DNA subsequences is small or the probability that all DNA subsequences are base complementary pairing regions of the PCA experiment is too low, it is determined whether the probability that the DNA subsequences are base complementary pairing regions of the PCA experiment is greater than a preset threshold.

[0116] If the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment is greater than a preset threshold, the DNA subsequence is identified as a base complementary pairing region of the PCA experiment.

[0117] If the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment is less than or equal to a preset threshold, it is determined that the DNA subsequence is not a base complementary pairing region of the PCA experiment.

[0118] By setting a threshold to determine whether a DNA subsequence is a base complementary pairing region, it is ensured that when the number of DNA subsequences is too small or the probability that all DNA subsequences are base complementary pairing regions of the PCA experiment is too low, the DNA subsequences that are base complementary pairing regions of the PCA experiment that do not meet the standards are screened out.

[0119] Through the above steps S2021-S2024, the base complementary pairing region classification model can be implemented to calculate the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment.

[0120] In a possible embodiment, a base complementary pairing region classification model is constructed through the following steps S2025-S2020.

[0121] S2025: Obtain a pre-designed DNA sample sequence.

[0122] Specifically, 34,000 DNA sample sequences with completed splitting designs were obtained from the DNA database.

[0123] S2026: Extract base complementary pairing regions and non-base complementary pairing regions from the DNA sample sequence.

[0124] Specifically, base complementary pairing regions and non-base complementary pairing regions are extracted from the obtained multiple DNA sample sequences and used as a standard for the base complementary pairing region classification model to judge whether a DNA subsequence is a base complementary pairing region of a PCA experiment.

[0125] S2027: Cutting the DNA sample sequence in the base complementary pairing region of the DNA sample sequence to obtain a first sample.

[0126] Specifically, a DNA sample sequence in the base complementary pairing region of the DNA sample sequence is intercepted as the first sample, which is used as a base complementary pairing region classification model to identify whether the DNA subsequence is a positive sample of the base complementary pairing region. The more similar the DNA subsequence is to the positive sample, the higher the probability that the DNA subsequence is the base complementary pairing region of the PCA experiment.

[0127] S2028: Cutting off the DNA sample sequence in the non-base complementary pairing region of the DNA sample sequence according to a preset sequence length to obtain a second sample.

[0128] Specifically, a DNA sample sequence in a non-base complementary pairing region of the DNA sample sequence is cut off according to a preset sequence length as a second sample, which is used as a base complementary pairing region classification model to identify whether the DNA subsequence is a negative sample of the base complementary pairing region. The more similar the DNA subsequence is to the negative sample, the lower the probability that the DNA subsequence is the base complementary pairing region of the PCA experiment.

[0129] In general, the portion of a DNA sequence that is a non-base complementary pairing region is much larger than the base complementary pairing region of the DNA sequence. In order to shorten the learning time of the base complementary pairing region classification model, the non-base complementary pairing region of the DNA sample sequence is intercepted according to a preset sequence length. In actual situations, the preset sequence length can be adjusted according to actual conditions.

[0130] S2029: Construct a base complementary pairing region classification model based on the first sample and the second sample.

[0131] Specifically, the first sample is used as a positive sample and the second sample is used as a negative sample to construct a base complementary pairing region classification model. Through the base complementary pairing region classification model, the probability that the DNA subsequence is a base complementary pairing region of the PCA experiment can be calculated.

[0132] In a possible embodiment, the base complementary pairing region classification model includes at least one of the following: a convolutional neural network model, a recurrent neural network model, and a multi-layer perceptron model.

[0133] S203: Using the probability that the multiple DNA subsequences are base complementary pairing regions as the scores of the multiple DNA subsequences.

[0134] Specifically, the probability that the DNA subsequence is a base complementary pairing region is obtained, and the probability that the DNA subsequence is a base complementary pairing region is used as the score of the multiple DNA subsequences.

[0135] Through the method provided in the embodiment of the present application, it is possible to score DNA subsequences when only the DNA sequence is input, thereby obtaining the DNA subsequence with the highest assembly efficiency as the base complementary pairing region of the PCA experiment, thereby improving the assembly efficiency of the PCA assembly. It is also possible to determine the DNA subsequence with the highest assembly efficiency without inputting artificially extracted features such as the CG content and Tm of the DNA.

[0136] Combine the following Figure 3 , introduces a DNA sequence splitting method provided in an embodiment of the present application, which is used to implement the step S104 of adjusting the number of bases of the target sequence according to the principle of similar DNA denaturation temperature, which can be achieved through steps S301-S306.

[0137] S301: Calculate DNA denaturation temperatures of multiple target sequences corresponding to multiple search regions to obtain a first DNA denaturation temperature array.

[0138] Specifically, the DNA denaturation temperature of the target sequence corresponding to each search region is calculated and stored in the form of an array to obtain a first DNA denaturation temperature array.

[0139] S302: Determine whether the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is less than a preset threshold.

[0140] Specifically, the minimum value and the maximum value in the first DNA denaturation temperature array are extracted, and the difference between the minimum value and the maximum value is calculated, and it is determined whether the difference is less than a preset threshold value.

[0141] If the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is smaller than a preset threshold, the target sequence is determined as a base complementary pairing region of the DNA sequence.

[0142] If the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is greater than or equal to the preset threshold, the process proceeds to step S303.

[0143] S303: Calculate the absolute value of the difference between each DNA denaturation temperature in the first DNA denaturation temperature array and the median to obtain an absolute value set.

[0144] Specifically, the median of the first DNA denaturation temperature array is calculated, and the absolute value of the difference between each DNA denaturation temperature in the first DNA denaturation temperature array and the median is calculated to obtain an absolute value set.

[0145] By calculating the absolute value of the difference between each DNA denaturation temperature and the median, the DNA denaturation temperature with the largest deviation value can be found, and the corresponding target sequence can be obtained based on the DNA denaturation temperature.

[0146] S304: Adjust the number of bases of the target sequence corresponding to the maximum absolute value in the absolute value set to obtain an adjusted target sequence.

[0147] Specifically, a target sequence corresponding to the maximum absolute value in the absolute value set is obtained, and the target sequence is adjusted to obtain an adjusted target sequence.

[0148] If the DNA denaturation temperature of the target sequence is greater than the median, the base number of the target sequence is reduced according to a preset base number adjustment value to obtain an adjusted target sequence.

[0149] If the DNA denaturation temperature of the target sequence is less than the median, the number of bases of the target sequence is increased according to the base number adjustment value to obtain an adjusted target sequence.

[0150] The target sequence corresponding to the DNA denaturation temperature with the largest deviation value is adjusted so that the DNA denaturation temperature corresponding to the target sequence is close to the remaining DNA denaturation temperatures in the first DNA denaturation temperature array.

[0151] S305: Calculate the adjusted DNA denaturation temperature of the target sequence to obtain a second DNA denaturation temperature array.

[0152] Specifically, the adjusted DNA denaturation temperature of the target sequence is recalculated, and the DNA denaturation temperature of the target sequence in the first DNA denaturation temperature array is replaced to obtain a second DNA denaturation temperature.

[0153] S306: If the difference between the maximum value and the minimum value in the second DNA denaturation temperature array is smaller than a preset threshold, the adjusted target sequence is determined as a base complementary pairing region of the DNA sequence.

[0154] Specifically, the minimum value and the maximum value in the second DNA denaturation temperature array are re-extracted, and the difference between the minimum value and the maximum value is calculated, and it is determined whether the difference is less than a preset threshold.

[0155] If the difference between the maximum value and the minimum value in the second DNA denaturation temperature array is smaller than a preset threshold, the adjusted target sequence is determined as a base complementary pairing region of the DNA sequence.

[0156] If the difference between the maximum value and the minimum value in the second DNA denaturation temperature array is greater than the preset threshold, the process returns to step S303.

[0157] Through the method provided in the embodiment of the present application, the target sequence can be adjusted according to the Tm similarity principle, so that the Tm of the obtained base complementary pairing region can be similar, thereby improving the success rate of the PCA test and improving the assembly efficiency of the split DNA sequence.

[0158] Combine the following Figure 6 , introducing a DNA sequence splitting device provided in the present application, the device includes: a sequence processing module 601, a scoring module 602, an adjustment module 603 and a splitting module 604.

[0159] The sequence processing module 601 is used to determine a search region in a DNA sequence according to a preset oligonucleotide length.

[0160] The sequence processing module 601 is used to intercept the DNA sequence in the search area according to a preset sequence length to obtain a DNA subsequence group.

[0161] The scoring module 602 is used to score the DNA subsequence group through a deep learning model, and determine the DNA subsequence with the highest score from the DNA subsequence group as the target sequence according to the scoring result.

[0162] The adjustment module 603 is used to adjust the base number of the target sequence according to the principle of similar DNA denaturation temperature to obtain the base complementary pairing region of the DNA sequence.

[0163] The splitting module 604 is used to split the DNA sequence according to the base complementary pairing regions to obtain oligonucleotide fragments.

[0164] Specifically, the scoring module 602 is specifically used to encode the DNA subsequence group using a preset encoding method to obtain a DNA sequence matrix, and the DNA sequence matrix includes multiple DNA subsequences.

[0165] The scoring module 602 is also used to input the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the multiple DNA subsequences are base complementary pairing regions of the polymerase chain reaction assembly (PCA) experiment.

[0166] The scoring module 602 is further configured to use the probability that the multiple DNA subsequences are base complementary pairing regions as the scores of the multiple DNA subsequences.

[0167] Specifically, the scoring module 602 is further used to input the multiple DNA subsequences into a preset base complementary pairing region classification model, wherein the preset base complementary pairing region classification model includes a convolutional layer, a maximum pooling layer and a fully connected layer.

[0168] The scoring module 602 is further configured to extract sequence features of the multiple DNA subsequences through a convolution layer according to a preset size standard to obtain first sequence features of the multiple DNA subsequences.

[0169] The scoring module 602 is further used to perform feature dimension reduction processing on the first sequence features of the plurality of DNA subsequences through a maximum pooling layer to obtain a plurality of second sequence features.

[0170] The scoring module 602 is further used to combine the multiple second sequence features through the fully connected layer, and calculate the probability that the multiple DNA subsequences are base complementary pairing regions of the PCA experiment based on the combined second sequence features.

[0171] Specifically, the scoring module 602 is also used to obtain a pre-designed DNA sample sequence.

[0172] The scoring module 602 is also used to extract base complementary pairing regions and non-base complementary pairing regions from the DNA sample sequence.

[0173] The scoring module 602 is further used to intercept the DNA sample sequence in the base complementary pairing region of the DNA sample sequence to obtain a first sample.

[0174] The scoring module 602 is further used to intercept the DNA sample sequence in the non-base complementary pairing region of the DNA sample sequence according to a preset sequence length to obtain a second sample.

[0175] The scoring module 602 is further used to construct a base complementary pairing region classification model according to the first sample and the second sample.

[0176] The adjustment module 603 is further used to calculate the DNA denaturation temperatures of multiple target sequences corresponding to the multiple search regions to obtain a first DNA denaturation temperature array.

[0177] If the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is less than a preset threshold, the adjustment module 603 is further configured to determine the target sequence as a base complementary pairing region of the DNA sequence.

[0178] Specifically, if the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is greater than or equal to a preset threshold, the adjustment module 603 is also used to calculate the absolute value of the difference between each DNA denaturation temperature in the first DNA denaturation temperature array and the median to obtain an absolute value set, where the median is the median of the first DNA denaturation temperature array.

[0179] The adjustment module 603 is further configured to adjust the number of bases of the target sequence corresponding to the maximum absolute value in the absolute value set to obtain an adjusted target sequence.

[0180] The adjustment module 603 is further used to calculate the adjusted DNA denaturation temperature of the target sequence to obtain a second DNA denaturation temperature array.

[0181] If the difference between the maximum value and the minimum value in the second DNA denaturation temperature array is smaller than a preset threshold, the adjustment module 603 is further configured to determine the adjusted target sequence as a base complementary pairing region of the DNA sequence.

[0182] Specifically, if the DNA denaturation temperature of the target sequence is greater than the median, the adjustment module 603 is further configured to reduce the number of bases of the target sequence according to a preset base number adjustment value to obtain an adjusted target sequence.

[0183] If the DNA denaturation temperature of the target sequence is less than the median, the adjustment module 603 is further configured to increase the number of bases of the target sequence according to the base number adjustment value to obtain an adjusted target sequence.

[0184] Through the device provided in the embodiment of the present application, it is possible to obtain the base complementary pairing region with the highest score from the DNA sequence, and split the DNA sequence according to the base complementary pairing region with the highest score, thereby improving the assembly efficiency of PCA assembly.

[0185] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0186] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0187] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical business division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0188] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0189] In addition, each business unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software business units.

[0190] If the integrated unit is implemented in the form of a software business unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0191] Those skilled in the art will appreciate that in one or more of the above examples, the services described in the present invention may be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, the services may be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any media that facilitates the transmission of computer programs from one place to another. Storage media may be any available media that can be accessed by a general-purpose or special-purpose computer.

[0192] The above specific implementation modes further describe the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific implementation modes of the present invention.

[0193] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person of ordinary skill in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for splitting a DNA sequence, characterized in that: The method comprises: Determine the search region in the DNA sequence according to the preset single-stranded oligonucleotide length; According to a preset sequence length, the DNA sequence is intercepted in the search region to obtain a DNA subsequence group; Scoring the DNA subsequence group by using a deep learning model, and determining the DNA subsequence with the highest score from the DNA subsequence group as the target sequence according to the scoring result; The number of bases of the target sequence is adjusted according to the principle of similar DNA denaturation temperatures to obtain a base complementary pairing region of the DNA sequence; wherein, there are multiple search regions, and the number of bases of the target sequence is adjusted according to the principle of similar DNA denaturation temperatures to obtain a base complementary pairing region of the DNA sequence, including: calculating the DNA denaturation temperatures of multiple target sequences corresponding to the multiple search regions to obtain a first DNA denaturation temperature array; if the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is less than a preset threshold, the target sequence is determined as the base complementary pairing region of the DNA sequence; Splitting the DNA sequence according to the base complementary pairing region to obtain single-stranded oligonucleotide fragments; The scoring of the DNA subsequence group by the deep learning model includes: encoding the DNA subsequence group by a preset encoding method to obtain a DNA sequence matrix, wherein the DNA sequence matrix includes a plurality of DNA subsequences; inputting the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the plurality of DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly (PCA) experiment; and using the probability that the plurality of DNA subsequences are the base complementary pairing regions as the scores of the plurality of DNA subsequences; The step of inputting the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the multiple DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly experiment comprises: inputting the multiple DNA subsequences into the preset base complementary pairing region classification model, wherein the preset base complementary pairing region classification model comprises a convolutional layer, a maximum pooling layer and a fully connected layer; extracting sequence features of the multiple DNA subsequences through the convolutional layer according to a preset size standard to obtain first sequence features of the multiple DNA subsequences; performing feature dimension reduction processing on the first sequence features of the multiple DNA subsequences through the maximum pooling layer to obtain multiple second sequence features; combining the multiple second sequence features through the fully connected layer, and calculating the probability that the multiple DNA subsequences are base complementary pairing regions of a PCA experiment according to the combined second sequence features; Among them, the method for constructing the preset base complementary pairing region classification model includes: obtaining a pre-designed DNA sample sequence; intercepting base complementary pairing regions and non-base complementary pairing regions from the DNA sample sequence; intercepting the DNA sample sequence in the base complementary pairing region of the DNA sample sequence to obtain a first sample; intercepting the DNA sample sequence in the non-base complementary pairing region of the DNA sample sequence according to a preset sequence length to obtain a second sample; and constructing the base complementary pairing region classification model based on the first sample and the second sample.

2. The method according to claim 1, characterized in that The preset base complementary pairing region classification model includes at least one of the following: a convolutional neural network model, a recurrent neural network model and a multi-layer perceptron model.

3. The method according to claim 1, characterized in that The method further comprises: If the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is greater than or equal to a preset threshold, then calculating the absolute value of the difference between each DNA denaturation temperature in the first DNA denaturation temperature array and the median to obtain an absolute value set, wherein the median is the median of the first DNA denaturation temperature array; Adjusting the number of bases of the target sequence corresponding to the maximum absolute value in the absolute value set to obtain an adjusted target sequence; Calculating the adjusted DNA denaturation temperature of the target sequence to obtain a second DNA denaturation temperature array; If the difference between the maximum value and the minimum value in the second DNA denaturation temperature array is smaller than a preset threshold, the adjusted target sequence is determined as the base complementary pairing region of the DNA sequence.

4. The method according to claim 3, characterized in that The step of adjusting the number of bases of the target sequence corresponding to the maximum absolute value in the absolute value set to obtain the adjusted target sequence includes: If the DNA denaturation temperature of the target sequence is greater than the median, the number of bases of the target sequence is reduced according to a preset base number adjustment value to obtain the adjusted target sequence; If the DNA denaturation temperature of the target sequence is less than the median, the number of bases of the target sequence is increased according to the base number adjustment value to obtain the adjusted target sequence.

5. A DNA sequence splitting device, characterized in that: The device comprises: a sequence processing module, a scoring module, an adjustment module and a splitting module; The sequence processing module is used to determine the search region in the DNA sequence according to the preset single-stranded oligonucleotide length; The sequence processing module is used to intercept the DNA sequence in the search area according to a preset sequence length to obtain a DNA subsequence group; The scoring module is used to score the DNA subsequence group through a deep learning model, and determine the DNA subsequence with the highest score from the DNA subsequence group as the target sequence according to the scoring result; The adjustment module is used to adjust the number of bases of the target sequence according to the principle of similar DNA denaturation temperatures to obtain the base complementary pairing region of the DNA sequence; wherein the DNA denaturation temperatures of the multiple target sequences corresponding to the multiple search regions are calculated to obtain a first DNA denaturation temperature array; if the difference between the maximum value and the minimum value in the first DNA denaturation temperature array is less than a preset threshold, the target sequence is determined to be the base complementary pairing region of the DNA sequence; The splitting module is used to split the DNA sequence according to the base complementary pairing region to obtain single-stranded oligonucleotide fragments; The scoring of the DNA subsequence group by the deep learning model includes: encoding the DNA subsequence group by a preset encoding method to obtain a DNA sequence matrix, wherein the DNA sequence matrix includes a plurality of DNA subsequences; inputting the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the plurality of DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly (PCA) experiment; and using the probability that the plurality of DNA subsequences are the base complementary pairing regions as the scores of the plurality of DNA subsequences; The step of inputting the DNA sequence matrix into a preset base complementary pairing region classification model to obtain the probability that the multiple DNA subsequences are base complementary pairing regions of a polymerase chain reaction assembly experiment comprises: inputting the multiple DNA subsequences into the preset base complementary pairing region classification model, wherein the preset base complementary pairing region classification model comprises a convolutional layer, a maximum pooling layer and a fully connected layer; extracting sequence features of the multiple DNA subsequences through the convolutional layer according to a preset size standard to obtain first sequence features of the multiple DNA subsequences; performing feature dimension reduction processing on the first sequence features of the multiple DNA subsequences through the maximum pooling layer to obtain multiple second sequence features; combining the multiple second sequence features through the fully connected layer, and calculating the probability that the multiple DNA subsequences are base complementary pairing regions of a PCA experiment according to the combined second sequence features; Among them, the method for constructing the preset base complementary pairing region classification model includes: obtaining a pre-designed DNA sample sequence; intercepting base complementary pairing regions and non-base complementary pairing regions from the DNA sample sequence; intercepting the DNA sample sequence in the base complementary pairing region of the DNA sample sequence to obtain a first sample; intercepting the DNA sample sequence in the non-base complementary pairing region of the DNA sample sequence according to a preset sequence length to obtain a second sample; and constructing the base complementary pairing region classification model based on the first sample and the second sample.

Images