Method and device for judging heterozygous frame-shift mutation in gene editing target site sequencing result based on machine learning, electronic equipment and computer readable medium

Abstract

The invention discloses a method and a device for judging heterozygous frame-shift mutation in a gene editing target site sequencing result based on machine learning, electronic equipment and a computer readable medium. Whether heterozygous frame-shift mutation occurs or not is judged according to a gene editing target site region sequencing result, and labeling is performed; and extracting a main peak sequence, a secondary peak sequence and a consistency sequence from a sequencing result, determining a target site area, and extracting characteristic values of sequences around the target site area to form sample characteristics. Mapping from sample features to sample labels is constructed, and a model is trained based on a supervised classification algorithm in machine learning; and extracting feature information of a sequencing result of a to-be-predicted sample, and inputting the feature information into the model so as to obtain a judgment result about whether the to-be-predicted sample is subjected to frame-shift mutation or not. According to the method, the whole classification process does not need to depend on a network and manual operation, the heterozygous frame-shift mutation event can be rapidly and accurately judged, and compared with manual recognition, the efficiency is improved, and the labor cost is saved.

Description

Technical Field

[0001] The present invention relates to the field of detection of heterozygous frameshift mutations in gene editing, and in particular to a method, device, electronic device and computer-readable medium for determining heterozygous frameshift mutations in sequencing results of gene editing target sites based on machine learning. Background Art

[0002] Gene editing technology, especially the Crispr / Cas9 gene editing system, has been widely used in gene editing breeding of animals and plants. Detection of target site editing is essential in large-scale gene editing breeding. Technology based on PCR amplification and first-generation sequencing is one of the important methods to detect whether the target site is edited.

[0003] The existing sequencing result identification method mainly determines whether a sample has a heterozygous frameshift mutation by manually identifying the bimodal features in the target site region in the sequencing peak graph. This has a huge workload in practical applications. In addition, the manual identification of heterozygous frameshift mutations in sequencing peak graphs depends on personal experience and judgment. The results are less controllable and time-consuming and laborious. It can be seen that the existing classification method has many drawbacks. Therefore, it is urgent to provide a processing method that can accurately determine whether a frameshift mutation has occurred based on the peak graph in batches to improve recognition efficiency and reduce labor costs. Summary of the invention

[0004] In order to solve the technical problem that the judgment of heterozygous frameshift in the prior art relies on manual labor, resulting in low controllability and time-consuming and labor-intensive, the present invention provides a method, device, electronic device and computer-readable medium for judging heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning, which can improve recognition efficiency and reduce manual recognition costs.

[0005] In order to achieve the above technical purpose, the technical solution of the present invention is:

[0006] A method for determining heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning, comprising the following steps:

[0007] Step 1: extract the main peak sequence, the secondary peak sequence and the consensus sequence from the sequencing results that have been annotated with heterozygous frameshift mutation tags, and after determining the target site region, extract the sequence feature values ​​of the sequences around the target site region to form sample features; wherein the sample features include:

[0008] Feature 1) The length of homology between the consensus sequence and the target site sequence;

[0009] Feature 2) the number of degenerate bases within a preset range upstream of the start position and downstream of the end position of the target site region in the consensus sequence;

[0010] Feature 3) the difference and ratio of the number of degenerate bases in the same preset range between the downstream of the termination position and the upstream of the start position in the target site region in the consensus sequence;

[0011] Feature 4) the consistency of the secondary peak sequence and the primary peak sequence within the same preset range upstream of the start position and downstream of the end position in the target site region;

[0012] Step 2: construct a mapping from sample features to sample labels, and standardize the sample features based on the supervised classification algorithm in machine learning to establish a data set for training. Then, train the model and evaluate the accuracy of the trained model. Repeat the training until a model that meets the preset accuracy requirements is obtained.

[0013] Step 3, extracting the sequence feature value of the sequencing result of the sample to be predicted, and inputting it into the model obtained in step 3, so as to obtain a judgment result of whether a heterozygous frameshift mutation occurs in the sample to be predicted.

[0014] In the method, in step 1, the sequencing results that have been annotated with heterozygous frameshift mutation tags are obtained by the following steps:

[0015] The gene editing target site region is sequenced by the first generation, and then based on the sequencing results, it is determined whether a heterozygous frameshift mutation has occurred and the mutation is labeled.

[0016] The method described herein also includes, before performing first-generation sequencing on the gene editing target site region, the step of designing PCR primers in the region more than 200 bp upstream and downstream of the gene editing target site and performing sequence amplification.

[0017] The method described herein, for determining whether a heterozygous frameshift mutation occurs based on sequencing results, is based on the following steps:

[0018] Based on the target site sequence in the result file obtained by sequencing, the single peak on one side of the target site is found to locate the target site, and the target site is judged according to the located target site: when there is a double peak in the middle of the target site and it extends to the other side, it is judged as a heterozygous frameshift mutation, otherwise it is judged as a non-heterozygous frameshift mutation.

[0019] The method described above, when labeling, labels heterozygous frameshift mutations and non-heterozygous frameshift mutations as one of the binary classification labels.

[0020] The method, in step 1, the main peak sequence, the secondary peak sequence and the consensus sequence are extracted by the following steps:

[0021] The highest peak in the sequencing result is taken as the main peak, and then all sequences lower than 0.2 times the main peak are filtered out as background interference, and the highest one among the remaining sequences is found as the secondary peak; then the consistency sequence of the two sequences is obtained by sequence alignment of the main peak sequence and the secondary peak sequence.

[0022] In the method, in step 1, determining the target site region is to search for a region where the consensus sequence is homologous to the target site sequence, and determine the start and end positions of the target site region in the consensus sequence.

[0023] In the method, in the step 1, in the sample features, feature 2) includes:

[0024] ① The number of degenerate bases within 30-80 bp upstream of the start position of the target site region in the consensus sequence;

[0025] ②. The number of degenerate bases within 30-80bp downstream of the termination position of the target site region in the consensus sequence.

[0026] In the method, in step 1, in the sample features, feature 3) includes:

[0027] ① The difference between the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence and the number of degenerate bases within 30-80 bp upstream of the start position;

[0028] ②. The ratio of the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence to the number of degenerate bases within 30-80 bp upstream of the start position.

[0029] In the method, in step 1, in the sample features, feature 4) includes:

[0030] ①, the consistency of the sequence 30-80bp upstream of the starting position of the target site region in the secondary peak sequence and the sequence 30-80bp upstream of the starting position of the target site region in the main peak sequence;

[0031] ②. The consistency of the sequence 30-80bp downstream of the termination position of the target site region in the secondary peak sequence and the sequence 30-80bp downstream of the termination position of the target site region in the main peak sequence.

[0032] In the method, in step 2, data standardization is performed by normalization by mean variance normalization, and the formula is:

[0033]

[0034] where x scale is the standardized eigenvalue, x is the eigenvalue, x mean is the feature mean, and s is the feature variance.

[0035] In the method, in step 2, training the model and evaluating the accuracy of the trained model includes:

[0036] The obtained data set is used for model training, where 80% of the data in the data set is used as a training set to train the model, and the remaining 20% ​​of the data is used as a validation set to obtain evaluation results.

[0037] A device for judging heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning, comprising the following modules:

[0038] The sample feature extraction module is used to extract the main peak sequence, secondary peak sequence and consensus sequence from the sequencing results that have been annotated with heterozygous frameshift mutation tags, and extract the sequence feature values ​​of the sequences around the target site area after determining the target site area to form sample features; the sample features include:

[0039] Feature 1) The length of homology between the consensus sequence and the target site sequence;

[0040] Feature 2) the number of degenerate bases within a preset range upstream of the start position and downstream of the end position of the target site region in the consensus sequence;

[0041] Feature 3) the difference and ratio of the number of degenerate bases in the same preset range between the downstream of the termination position and the upstream of the start position in the target site region in the consensus sequence;

[0042] Feature 4) the consistency of the secondary peak sequence and the primary peak sequence within the same preset range upstream of the start position and downstream of the end position in the target site region;

[0043] The model building module is used to construct a mapping from sample features to sample labels, and standardize the sample features based on the supervised classification algorithm in machine learning, so as to establish a data set for training, and then train the model and evaluate the accuracy of the trained model, and repeat the training until a model that meets the preset accuracy requirements is obtained;

[0044] The heterozygous frameshift mutation judgment module is used to extract the sequence feature value of the sequencing result of the sample to be predicted, and input it into the model obtained by the model construction module, so as to obtain the judgment result of whether the sample to be predicted has a frameshift mutation.

[0045] In the device and the sample feature extraction module, the sequencing result with heterozygous frameshift mutation label annotated is to perform first-generation sequencing on the gene editing target site region, and then determine whether a heterozygous frameshift mutation occurs and perform label annotating based on the sequencing result; before performing first-generation sequencing on the gene editing target site region, it also includes designing PCR primers in the region more than 200bp upstream and downstream of the gene editing target site, and performing sequence amplification; determining whether a heterozygous frameshift mutation occurs based on the sequencing result is based on the target site sequence in the result file obtained by sequencing, finding a single peak on one side of the target site to locate the target site, and determining based on the located target site: when a double peak appears in the middle of the target site and continues to the other side, it is determined to be a heterozygous frameshift mutation, otherwise it is determined to be a non-heterozygous frameshift mutation; when performing label annotating, the heterozygous frameshift mutation and the non-heterozygous frameshift mutation are respectively marked as one of the binary classification labels.

[0046] In the device and the sample feature extraction module, the main peak sequence, the secondary peak sequence and the consistency sequence are extracted by the following steps:

[0047] The highest peak in the sequencing result is taken as the main peak, and then all sequences lower than 0.2 times the main peak are filtered out as background interference, and the highest one among the remaining sequences is found as the secondary peak; then the consistency sequence of the two sequences is obtained by sequence alignment of the main peak sequence and the secondary peak sequence.

[0048] In the device and the sample feature extraction module, determining the target site region is to search for a region where the consensus sequence is homologous to the target site sequence, and determining the start and end positions of the target site region in the consensus sequence.

[0049] In the device, in the sample feature extraction module, in the sample features, feature 2) includes:

[0050] ① The number of degenerate bases within 30-80 bp upstream of the start position of the target site region in the consensus sequence;

[0051] ② The number of degenerate bases within 30-80 bp downstream of the termination position of the target site region in the consensus sequence;

[0052] Feature 3) includes:

[0053] ① The difference between the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence and the number of degenerate bases within 30-80 bp upstream of the start position;

[0054] ②. The ratio of the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence to the number of degenerate bases within 30-80 bp upstream of the start position.

[0055] Feature 4) includes:

[0056] ①, the consistency of the sequence 30-80bp upstream of the starting position of the target site region in the secondary peak sequence and the sequence 30-80bp upstream of the starting position of the target site region in the main peak sequence;

[0057] ②. The consistency of the sequence 30-80bp downstream of the termination position of the target site region in the secondary peak sequence and the sequence 30-80bp downstream of the termination position of the target site region in the main peak sequence.

[0058] In the device and the model building module, data standardization is performed by normalization of mean variance, and the formula is:

[0059]

[0060] where x scale is the standardized eigenvalue, x is the eigenvalue, x mean is the feature mean, and s is the feature variance.

[0061] In the device and the model building module, training the model and evaluating the accuracy of the trained model includes:

[0062] The obtained data set is used for model training, where 80% of the data in the data set is used as a training set to train the model, and the remaining 20% ​​of the data is used as a validation set to obtain evaluation results.

[0063] An electronic device, comprising:

[0064] one or more processors;

[0065] a storage device for storing one or more programs,

[0066] When the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method.

[0067] A computer readable medium stores a computer program, which implements the above method when executed by a processor.

[0068] The technical effect of the present invention is that the present invention trains the classification model through sequence feature values. The present method makes full use of the information contained in the first-generation sequencing results in the selection of sequence features, and can ensure that the classification results have high accuracy and stability. The present invention has high accuracy in the classification judgment of samples. Compared with the traditional manual recognition and judgment method, the entire classification process does not need to rely on the network and manual operation, that is, it can realize fast and accurate judgment of heterozygous frameshift mutation events, which can significantly reduce the labor time cost, improve recognition efficiency, and reduce data analysis costs.

[0069] The present invention is described in detail below by combining the implementation modes shown in the accompanying drawings, wherein the following drawings are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 Shown is a flow chart of a method for determining heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning of the present invention;

[0071] Figure 2 Shown are the sequencing results and feature value extraction regions of non-heterozygous frameshift mutation samples of the method for judging heterozygous frameshift mutations in the sequencing results of gene editing target sites based on machine learning of the present invention;

[0072] Figure 3 Shown are the sequencing results and feature value extraction regions of heterozygous frameshift mutation samples in the method for judging heterozygous frameshift mutations in the sequencing results of gene editing target sites based on machine learning of the present invention;

[0073] Figure 4 Shown is a flow chart of the classification model construction of the method for judging heterozygous frameshift mutations in the sequencing results of gene editing target sites based on machine learning of the present invention;

[0074] Figure 5 Shown is a flow chart for classifying samples to be classified in the method for determining heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning of the present invention. DETAILED DESCRIPTION

[0075] The following will provide a clear and complete description of the concept, specific operation and technical effects of the present invention in conjunction with the embodiments and drawings to fully understand the purpose, scheme and effect of the present invention.

[0076] Example 1

[0077] See also Figure 1-5 This embodiment proposes a method for determining heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning, comprising the following steps:

[0078] PCR primers were designed in the regions more than 200 bp upstream and downstream of the gene editing target site, and sequence amplification and first-generation sequencing were performed.

[0079] Use ab1 format file reading software (such as SnapGene) to open the sequencing result file, find the target site sequence, locate the target site by manually identifying the single peak on one side of the target site, and judge the located target site: when a double peak appears in the middle of the target site and continues to the other side, it is judged as a heterozygous frameshift mutation event, and if this feature does not appear, it is judged as a non-heterozygous frameshift mutation event. Then label the sequencing results with binary classification, such as labeling heterozygous frameshift mutations as 1 and non-heterozygous frameshift mutations as 0.

[0080] The sangerseqR software package of R language was used to read the ab1 file of the first-generation sequencing, and the makeBaseCalls method was used to filter the background interference (ratio=0.20) lower than 0.2 times the main peak, and the main peak sequence and the secondary peak sequence in the sequencing result were extracted. The highest peak of the sequence appearing in the sequencing result was taken as the main peak, and then all sequences lower than 0.2 times the main peak were filtered out as background interference, and then the highest of the remaining sequences was found as the secondary peak. Then, by comparing the sequences of the main peak sequence and the secondary peak sequence, the consistency sequence of the two sequences was obtained, that is, the sequence composed of the majority of nucleotides or amino acids appearing at each corresponding position of the two sequences was taken as the consistency sequence.

[0081] Then, the sequence alignment tool blast+ software was used to find the target site region where the consensus sequence was homologous to the target site sequence, and the start and end positions of the target site region in the consensus sequence were extracted.

[0082] Next, the sequence feature values ​​of the sequence around the target site region are extracted to form sample features. The sequence feature values ​​include the following: (1) The length of homology between the consensus sequence and the target site sequence. (2) The number of degenerate bases within a preset range upstream of the starting position of the target site region in the consensus sequence. (3) The number of degenerate bases within a preset range downstream of the ending position of the target site region in the consensus sequence. (4) The difference between the number of degenerate bases within a preset range downstream of the ending position of the target site region in the consensus sequence and the number of degenerate bases within the same preset range upstream of the starting position. (5) The ratio of the number of degenerate bases within a preset range downstream of the ending position of the target site region in the consensus sequence to the number of degenerate bases within the same preset range upstream of the starting position. (6) The consistency of the sequence within the preset range upstream of the starting position of the target site region in the secondary peak sequence and the same preset range upstream of the starting position of the target site region in the main peak sequence. (7) The consistency of the sequence within the preset range downstream of the ending position of the target site region in the secondary peak sequence and the same preset range downstream of the ending position of the target site region in the main peak sequence. The preset range can be set to 30-80bp, and can be set to 50bp during specific operation.

[0083] After the extraction is completed, the sample features are standardized using the mean variance normalization method to obtain a data set for model training. Data standardization is performed by mean variance normalization, and the formula is:

[0084]

[0085] where x scale is the standardized eigenvalue, x is the eigenvalue, x mean is the feature mean, and s is the feature variance.

[0086] The model is constructed based on the supervised classification algorithm in machine learning, and the obtained data set is used for model training, where 80% of the data in the data set is used to train the model, and the other 20% of the data is used for verification and evaluation results. If the evaluation result does not meet the preset requirements, the training is repeated until the requirements are met, so as to obtain a classification model that can be used for judgment.

[0087] Then, the sequence feature value of the sequencing result of the sample to be predicted is extracted and input into the constructed classification model to determine whether the sample to be predicted has a frameshift mutation and output the judgment result.

[0088] When judging heterozygous frameshift mutations in the sequencing results of gene editing target sites of samples, this method obtains the data set used for model training through manual identification of a small number of samples, and then uses it to judge heterozygous frameshift mutations in the sequencing results of gene editing target sites of large quantities of samples.

[0089] Example 2

[0090] This embodiment proposes a device for judging heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning, including:

[0091] The sample feature extraction module is used to extract the main peak sequence, secondary peak sequence and consensus sequence from the sequencing results that have been annotated with heterozygous frameshift mutation tags, and extract the sequence feature values ​​of the sequences around the target site area after determining the target site area to form sample features; the sample features include:

[0092] (1) The length of homology between the consensus sequence and the target site sequence. (2) The number of degenerate bases within a preset range upstream of the starting position of the target site region in the consensus sequence. (3) The number of degenerate bases within a preset range downstream of the ending position of the target site region in the consensus sequence. (4) The difference between the number of degenerate bases within a preset range downstream of the ending position of the target site region in the consensus sequence and the number of degenerate bases within the same preset range upstream of the starting position. (5) The ratio of the number of degenerate bases within a preset range downstream of the ending position of the target site region in the consensus sequence to the number of degenerate bases within the same preset range upstream of the starting position. (6) The consistency of the sequence within the preset range upstream of the starting position of the target site region in the secondary peak sequence and the same preset range upstream of the starting position of the target site region in the main peak sequence. (7) The consistency of the sequence within the preset range downstream of the ending position of the target site region in the secondary peak sequence and the same preset range downstream of the ending position of the target site region in the main peak sequence. The preset range can be set to 30-80bp, and can be set to 50bp during specific operations.

[0093] The sequencing results in the sample feature extraction module that have been annotated with heterozygous frameshift mutation labels are first-generation sequencing of the gene editing target site region, and then based on the sequencing results, it is judged whether a heterozygous frameshift mutation occurs and the label is annotated; before the first-generation sequencing of the gene editing target site region, PCR primers are designed in the region more than 200bp upstream and downstream of the gene editing target site, and sequence amplification is performed; based on the sequencing results, it is judged whether a heterozygous frameshift mutation occurs, based on the target site sequence in the result file obtained by sequencing, a single peak on one side of the target site is found to locate the target site, and a judgment is made based on the located target site: when there is a double peak in the middle of the target site and it extends to the other side, it is judged as a heterozygous frameshift mutation, otherwise it is judged as a non-heterozygous frameshift mutation; when labeling, the heterozygous frameshift mutation and the non-heterozygous frameshift mutation are respectively annotated as one of the binary classification labels.

[0094] The main peak sequence, secondary peak sequence and consensus sequence in the sample feature extraction module are to take the highest peak of the sequence appearing in the sequencing result as the main peak, then filter out all sequences lower than 0.2 times the main peak as background interference, and then find the highest of the remaining sequences as the secondary peak; then, by comparing the main peak sequence and the secondary peak sequence, the consensus sequence of the two sequences is obtained. Determining the target site region is to find the region where the consensus sequence is homologous to the target site sequence, and determine the starting and ending positions of the target site region in the consensus sequence.

[0095] The model building module is used to construct a mapping from sample features to sample labels, and standardize the sample features based on the supervised classification algorithm in machine learning to establish a data set for training. The model is then trained and the accuracy of the trained model is evaluated. The training is repeated until a model that meets the preset accuracy requirements is obtained.

[0096] The data standardization is carried out by normalizing the mean variance, and the formula is:

[0097]

[0098] where x scale is the standardized eigenvalue, x is the eigenvalue, x mean is the feature mean, and s is the feature variance.

[0099] Training the model and evaluating the accuracy of the trained model includes: using the obtained data set for model training, wherein 80% of the data in the data set is used as a training set to train the model, and the remaining 20% ​​of the data is used as a validation set to obtain an evaluation result.

[0100] The heterozygous frameshift mutation judgment module is used to extract the sequence feature value of the sequencing result of the sample to be predicted, and input it into the model obtained by the model construction module, so as to obtain the judgment result of whether the sample to be predicted has a frameshift mutation.

[0101] According to an embodiment of the present invention, the present invention also provides an electronic device and a computer-readable medium.

[0102] Electronic equipment includes:

[0103] one or more processors;

[0104] a storage device for storing one or more programs,

[0105] When the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method.

[0106] In specific use, users can interact with servers, which are also electronic devices, through electronic devices as terminal devices and based on the network to achieve functions such as receiving or sending messages. Terminal devices are generally various electronic devices equipped with display devices and used based on human-computer interfaces, including but not limited to smart phones, tablet computers, laptops and desktop computers. Various specific application software can be installed on the terminal devices as needed, including but not limited to web browser software, instant messaging software, social platform software, shopping software, etc.

[0107] The server is a network service end used to provide various services, such as a backend server that provides corresponding judgment services for the sequencing results of the samples to be predicted transmitted from the terminal device. This is to train the historical model with the received training data set, and perform heterozygous frameshift mutation judgment based on the trained model, so as to return the final judgment result to the terminal device.

[0108] The heterozygous frameshift mutation judgment method provided in this embodiment is generally executed by a server. In actual application, under the necessary conditions, the terminal device can also directly perform the heterozygous frameshift mutation judgment. Accordingly, the heterozygous frameshift mutation judgment device can be set in the server. Similarly, under the necessary conditions, the heterozygous frameshift mutation judgment device can also be set in the terminal device.

[0109] Similarly, the computer-readable medium of the present invention stores a computer program thereon, and when the program is executed by a processor, the method for determining a heterozygous frameshift mutation according to an embodiment of the present invention is implemented. Although the description of the present invention has been quite detailed, it is not intended to be limited to any of these details or embodiments, but should be regarded as providing a broad possible interpretation for these claims by reference to the attached claims in view of the prior art, thereby effectively covering the intended scope of the present invention.

[0110] The above is only a preferred embodiment of the present invention. The present invention is not limited to the above implementation mode. As long as the technical effects of the present invention are achieved by the same means, they should fall within the protection scope of the present invention.

Claims

1. A method for determining heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning, characterized in that: The following steps are involved: Step 1: extract the main peak sequence, the secondary peak sequence and the consensus sequence from the sequencing results that have been annotated with heterozygous frameshift mutation tags, and after determining the target site region, extract the sequence feature values ​​of the sequences around the target site region to form sample features; in Sample characteristics include: Feature 1) The length of homology between the consensus sequence and the target site sequence; Feature 2) the number of degenerate bases within a preset range upstream of the start position and downstream of the end position of the target site region in the consensus sequence; Feature 3) the difference and ratio of the number of degenerate bases in the same preset range between the downstream of the termination position and the upstream of the start position in the target site region in the consensus sequence; Feature 4) the consistency of the secondary peak sequence and the primary peak sequence within the same preset range upstream of the start position and downstream of the end position in the target site region; Step 2: construct a mapping from sample features to sample labels, and standardize the sample features based on the supervised classification algorithm in machine learning to establish a data set for training. Then, train the model and evaluate the accuracy of the trained model. Repeat the training until a model that meets the preset accuracy requirements is obtained. Step 3, extracting the sequence feature value of the sequencing result of the sample to be predicted, and inputting it into the model obtained in step 3, so as to obtain a judgment result of whether a heterozygous frameshift mutation occurs in the sample to be predicted.

2. The method according to claim 1, characterized in that In step 1, the sequencing results annotated with heterozygous frameshift mutation tags are obtained by the following steps: The gene editing target site region is sequenced by the first generation, and then based on the sequencing results, it is determined whether a heterozygous frameshift mutation has occurred and the mutation is labeled.

3. The method according to claim 2, characterized in that Before performing first-generation sequencing on the gene editing target site region, the process also includes designing PCR primers in the region more than 200 bp upstream and downstream of the gene editing target site and performing sequence amplification steps.

4. The method according to claim 2, characterized in that: The determination of whether a heterozygous frameshift mutation occurs based on the sequencing results is based on the following steps: Based on the target site sequence in the result file obtained by sequencing, the single peak on one side of the target site is found to locate the target site, and the target site is judged according to the located target site: when there is a double peak in the middle of the target site and it extends to the other side, it is judged as a heterozygous frameshift mutation, otherwise it is judged as a non-heterozygous frameshift mutation.

5. The method according to claim 2, characterized in that: When labeling, heterozygous frameshift mutations and non-heterozygous frameshift mutations are respectively labeled as one of the binary classification labels.

6. The method according to claim 1, characterized in that In step 1, the main peak sequence, the secondary peak sequence and the consensus sequence are extracted by the following steps: The highest peak of the sequence appearing in the sequencing results is taken as the main peak, and then all sequences lower than 0.2 times the main peak are filtered out as background interference, and then the highest one among the remaining sequences is found as the secondary peak; then the sequence consistency of the two sequences is obtained by sequence alignment of the main peak sequence and the secondary peak sequence.

7. The method according to claim 1, characterized in that In the step 1, determining the target site region is to search for a region where the consensus sequence is homologous to the target site sequence, and determine the start and end positions of the target site region in the consensus sequence.

8. The method according to claim 1, characterized in that: In the step 1, among the sample features, feature 2) includes: ① The number of degenerate bases within 30-80 bp upstream of the start position of the target site region in the consensus sequence; ②. The number of degenerate bases within 30-80bp downstream of the termination position of the target site region in the consensus sequence.

9. The method according to claim 1, characterized in that: In the step 1, among the sample features, feature 3) includes: ① The difference between the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence and the number of degenerate bases within 30-80 bp upstream of the start position; ②. The ratio of the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence to the number of degenerate bases within 30-80 bp upstream of the start position.

10. The method according to claim 1, characterized in that In the step 1, among the sample features, feature 4) includes: ①, the consistency of the sequence 30-80bp upstream of the starting position of the target site region in the secondary peak sequence and the sequence 30-80bp upstream of the starting position of the target site region in the main peak sequence; ②. The consistency of the sequence 30-80bp downstream of the termination position of the target site region in the secondary peak sequence and the sequence 30-80bp downstream of the termination position of the target site region in the main peak sequence.

11. The method according to claim 1, characterized in that In step 2, data standardization is performed by normalization of mean variance, and the formula is: where x scale is the standardized eigenvalue, x is the eigenvalue, x mean is the feature mean, and s is the feature variance.

12. The method according to claim 1, characterized in that In step 2, training the model and evaluating the accuracy of the trained model include: The obtained data set is used for model training, where 80% of the data in the data set is used as a training set to train the model, and the remaining 20% ​​of the data is used as a validation set to obtain evaluation results.

13. A device for judging heterozygous frameshift mutations in gene editing target site sequencing results based on machine learning, characterized in that: Includes the following modules: The sample feature extraction module is used to extract the main peak sequence, the secondary peak sequence and the consensus sequence from the sequencing results that have been annotated with heterozygous frameshift mutation tags, and after determining the target site region, extract the sequence feature values ​​of the sequences around the target site region to form sample features; in Sample characteristics include: Feature 1) The length of homology between the consensus sequence and the target site sequence; Feature 2) the number of degenerate bases within a preset range upstream of the start position and downstream of the end position of the target site region in the consensus sequence; Feature 3) the difference and ratio of the number of degenerate bases in the same preset range between the downstream of the termination position and the upstream of the start position in the target site region in the consensus sequence; Feature 4) the consistency of the secondary peak sequence and the primary peak sequence within the same preset range upstream of the start position and downstream of the end position in the target site region; The model building module is used to construct a mapping from sample features to sample labels, and standardize the sample features based on the supervised classification algorithm in machine learning, so as to establish a data set for training, and then train the model and evaluate the accuracy of the trained model, and repeat the training until a model that meets the preset accuracy requirements is obtained; The heterozygous frameshift mutation judgment module is used to extract the sequence feature value of the sequencing result of the sample to be predicted, and input it into the model obtained by the model construction module, so as to obtain the judgment result of whether the sample to be predicted has a frameshift mutation.

14. The device according to claim 13, characterized in that In the sample feature extraction module, the sequencing results that have been annotated with heterozygous frameshift mutation labels are first-generation sequencing of the gene editing target site region, and then based on the sequencing results, it is determined whether a heterozygous frameshift mutation occurs and the label is annotated; Before performing first-generation sequencing on the gene editing target site region, PCR primers are designed in the region more than 200bp upstream and downstream of the gene editing target site, and sequence amplification is performed; judging whether a heterozygous frameshift mutation occurs based on the sequencing results is based on the target site sequence in the result file obtained by sequencing, and the target site is located by finding a single peak on one side of the target site, and judging based on the located target site: when there is a double peak in the middle of the target site and it extends to the other side, it is judged as a heterozygous frameshift mutation, otherwise it is judged as a non-heterozygous frameshift mutation; when labeling, the heterozygous frameshift mutation and non-heterozygous frameshift mutation are respectively marked as one of the binary classification labels.

15. The device according to claim 13, characterized in that In the sample feature extraction module, the main peak sequence, the secondary peak sequence and the consensus sequence are extracted by the following steps: The highest peak of the sequence appearing in the sequencing results is taken as the main peak, and then all sequences lower than 0.2 times the main peak are filtered out as background interference, and then the highest one among the remaining sequences is found as the secondary peak; then the sequence consistency of the two sequences is obtained by sequence alignment of the main peak sequence and the secondary peak sequence.

16. The device according to claim 13, characterized in that In the sample feature extraction module, determining the target site region is to search for a region where the consensus sequence is homologous to the target site sequence, and to determine the start and end positions of the target site region in the consensus sequence.

17. The device according to claim 13, characterized in that In the sample feature extraction module, among the sample features, feature 2) includes: ① The number of degenerate bases within 30-80 bp upstream of the start position of the target site region in the consensus sequence; ② The number of degenerate bases within 30-80 bp downstream of the termination position of the target site region in the consensus sequence; Feature 3) includes: ① The difference between the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence and the number of degenerate bases within 30-80 bp upstream of the start position; ②. The ratio of the number of degenerate bases within 30-80 bp downstream of the end position of the target site region in the consensus sequence to the number of degenerate bases within 30-80 bp upstream of the start position. Feature 4) includes: ①, the consistency of the sequence 30-80bp upstream of the starting position of the target site region in the secondary peak sequence and the sequence 30-80bp upstream of the starting position of the target site region in the main peak sequence; ②. The consistency of the sequence 30-80bp downstream of the termination position of the target site region in the secondary peak sequence and the sequence 30-80bp downstream of the termination position of the target site region in the main peak sequence.

18. The device according to claim 13, characterized in that In the model building module, data standardization is performed by normalizing the mean variance, and the formula is: where x scale is the standardized eigenvalue, x is the eigenvalue, x mean is the feature mean, and s is the feature variance.

19. The device according to claim 13, characterized in that In the model building module, training the model and evaluating the accuracy of the trained model include: The obtained data set is used for model training, where 80% of the data in the data set is used as a training set to train the model, and the remaining 20% ​​of the data is used as a validation set to obtain evaluation results.

20. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 12.

21. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 12 is implemented.

Images