WO2020001663A2 - Gene sequencing result type detection method and apparatus, device, and storage medium - Google Patents

Abstract

Disclosed are a gene sequencing result type detection method and apparatus, a device, and a storage medium. The method comprises: obtaining a peak map of a gene sequencing result to be detected; according to the peak map, extracting characteristic data corresponding to the gene sequencing result to be detected; inputting the characteristic data into a type detection model, and obtaining a type matched with the gene sequencing result to be detected.

Description

Detection method, device, equipment and storage medium of gene sequencing result type

This application claims priority from a Chinese patent application filed with the Chinese Patent Office on June 27, 2018 with application number 201810675765.3, the entire contents of which are incorporated herein by reference.

Technical field

The present application relates to the technical field of gene detection, for example, to a method, a device, a device, and a storage medium for detecting a type of a result of gene sequencing.

Background technique

As an important subject to promote human development, gene testing has been paid more and more attention by many countries in the world. In particular, gene sequencing technology has gradually entered a period of rapid development.

Gene sequencing technology, that is, the technology for determining the sequence of Deoxyribonucleic Acid (DNA). In the research of molecular biology, the sequence analysis of DNA is the basis for the research and transformation of the target gene. During the DNA sequencing process, sequencing results may be unavailable due to various factors. Therefore, it is necessary to determine whether the DNA sequencing results are available.

The quality of genetic sequencing results is mainly determined by humans, and there may be millions of sequencing results every day. A large number of technicians are needed to view and analyze the peaks of sequencing results. Therefore, a lot of waste will be wasted. Human, material and financial resources, and inefficient. In addition, because the determination is manual, the sequencing results will be misjudged due to inconsistent determination standards between people, resulting in inaccurate determination results.

Summary of the invention

The embodiments of the present application provide a method, a device, a device, and a storage medium for detecting a type of a gene sequencing result, so as to realize automatic determination of the type of a gene sequencing result, save manpower, material resources, and financial resources, and improve determination efficiency and accuracy.

An embodiment of the present application provides a method for detecting a type of a gene sequencing result, including:

Obtain a peak shape diagram of the sequencing result of the gene to be tested;

Extracting characteristic data corresponding to the sequencing result of the gene to be tested according to the peak shape map;

The feature data is input into a type detection model to obtain a type that matches the sequencing result of the gene to be tested.

An embodiment of the present application further provides a detection device for a type of gene sequencing result, and the device includes:

The peak shape acquisition module is configured to obtain a peak shape map of the sequencing result of the gene to be tested;

A feature extraction module configured to extract feature data corresponding to the sequencing result of the gene to be tested according to the peak shape map;

The type detection module is configured to input the feature data into a type detection model to obtain a type that matches the sequencing result of the gene to be tested.

An embodiment of the present application further provides a computer device, and the device includes:

One or more processors;

Memory, set to store one or more programs;

When the one or more programs are executed by the one or more processors, the one or more processors enable the method for detecting a type of gene sequencing result described in any one of the embodiments of the present application.

An embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the method for detecting a type of gene sequencing result described in any of the embodiments of the present application is implemented.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic flowchart of a method for detecting a type of a gene sequencing result provided in Embodiment 1 of the present application; FIG.

2 is a schematic flowchart of a method for establishing a type detection model provided in Embodiment 2 of the present application;

3 is a schematic structural diagram of a detection device for a type of gene sequencing result provided in Embodiment 3 of the present application;

FIG. 4 is a schematic structural diagram of a computer device according to a fourth embodiment of the present application.

detailed description

The following describes the application with reference to the drawings and embodiments. The specific embodiments described herein are only used to explain the present application, rather than limiting the present application. For ease of description, only some, but not all, structures related to the present application are shown in the drawings.

Example one

FIG. 1 is a schematic flowchart of a method for detecting a type of a gene sequencing result provided in Embodiment 1 of the present application. The method can be applied to the determination of the quality of genetic sequencing results. The method can be implemented by a detection device of the type of genetic sequencing results. The device can be composed of hardware and / or software, and can be integrated in a computer and all the included data. Processing functions in the terminal. The method includes the following steps:

S110. Obtain a peak shape diagram of a sequencing result of the gene to be tested.

The peak shape graph obtained in this embodiment may be a peak shape graph obtained by a one-generation sequencing result of a gene to be tested. Exemplarily, a file of the peak shape diagram corresponding to the sequencing result can be obtained by performing one generation sequencing of the gene to be tested, and then the peak shape diagram can be obtained from the file of the peak shape diagram. In one embodiment, the file of the peak graph is an abl format file.

In an embodiment, since different gene sequences may correspond to different peak shape diagrams, the quality of the peak shape diagrams directly affects the use of gene sequencing results. For example, under normal circumstances, if there is no abnormal peak shape in the peak shape graph, it means that the gene sequencing result corresponding to the peak shape graph is available; if there is an abnormal peak shape in the peak shape graph, it means that the gene sequencing result corresponding to the peak shape graph If abnormal, you need to analyze the sequencing results or directly re-sequence.

S120. Extract feature data corresponding to the sequencing result of the gene to be tested according to the peak shape map.

In this embodiment, the characteristic data is data that can reflect the characteristics of the peaks in the peak shape graph and / or sequence information data corresponding to the sequencing result of the gene to be tested. For example, the characteristic data includes, but is not limited to, a base sequence, a base quality value, and a peak. Wide values, lengths of base fragments with a length of more than 20 consecutive longest bases in the sequence, signal strengths of multiple bases in the sequence, and average signal strength.

For example, a specific script file is run to extract the characteristic data in the peak shape graph. In one embodiment, the script file includes different feature extraction algorithms corresponding to different feature data. In the process of running the specific script file, the feature extraction algorithm contained in the script can be used to extract corresponding feature data.

The purpose of extracting the feature data is to provide a basis for determining the type of the gene sequencing result to be tested by using a type detection model in the subsequent steps, and to automatically determine the type of the gene sequencing result. In addition, the feature data is used as the type of gene sequencing result. Compared with the manual observation of the peak shape for determination, the accuracy of the determination is also improved.

In one embodiment, extracting feature data corresponding to the sequencing result of the gene to be tested according to the peak shape map includes: extracting feature data from a file of the peak shape map; and / or, processing the peak shape map according to a preset feature extraction algorithm, Get the corresponding feature data.

Exemplarily, feature data can be directly extracted from the peak shape file obtained by one-generation sequencing, such as directly extracting sequence information, base quality values, and peak width values; or obtained through one or more preset feature extraction algorithms The one or more characteristic data corresponding to the peak shape graph, such as a length value of a base fragment with a continuous longest base mass greater than 20 in the sequence, a signal intensity of multiple bases in the sequence, and an average signal intensity. For a practical example, from the abl format file, directly extract the sequence information, base quality values, and peak width values contained in the parameter data in the file; for another example, for the peak shape diagram of the sequencing result of the gene to be tested, run the include There is a script file of the base signal intensity extraction algorithm to obtain the signal intensity of each base in the peak shape graph.

S130. The feature data is input into a type detection model, and a type matching the sequencing result of the gene to be tested is obtained.

In this embodiment, the type detection model is used to identify the input feature data to identify a type that matches the sequencing result of the gene to be tested. In an embodiment, the types that match the sequencing result of the gene to be tested may include: normal types and abnormal types. The normal type may be a type in which there is no abnormal peak shape in the peak shape diagram of the sequencing result of the gene to be tested, and the abnormal type may be a type in which there is an abnormal peak shape in the peak shape diagram of the gene sequencing result. In an embodiment, the type of abnormality may include: a poly structure or a repeating sequence type, a spectral pull-up type, a diffusion type, a bubble type, an attenuation type, a peak set type, a no signal type, a primer problem type, and an interruption type. One or more types. For example, after the type detection model recognizes the input feature data, it outputs either normal type or abnormal type. For example, after the type detection model recognizes the input feature data, it outputs normal type, poly structure, or repeated sequence. Any of the following types: type, spectral elevation type, dispersion type, bubble type, attenuation type, peak set type, no signal type, primer problem type, or interruption type. The output type of the type detection model is identified based on the characteristic data.

In an embodiment, the working principle of the type detection model may be that when the feature data is input, the type detection model recognizes the input feature data, determines which type of the input feature data is, and then outputs the type. For example, the average base quality extracted from the peak shape diagram of the first-generation sequencing results (take the peak shape diagram of the diffuse type as an example), the length of the base fragment with a continuous longest base quality greater than 20, and the average signal intensity are input to the type In the detection model, the type detection model identified and analyzed these three characteristic data, and obtained that the type of the gene sequencing result was a diffuse type, so that it can be seen that the gene sequencing result was abnormal, the sequencing failed, and the peak shape chart showed a peak shape. Wide and unrecognizable. The possible reason is that the DNA concentration is too high, which in turn provides a basis for adjusting the sequencing scheme when re-sequencing the gene.

In the technical solution of this embodiment, the characteristic data corresponding to the sequencing result of the gene to be tested is extracted based on the obtained peak shape diagram of the sequencing result of the gene to be tested, and the characteristic data is input into the type detection model, and finally the sequencing with the gene to be tested is obtained. The types that match the results, using the feature data extracted from the peak shape graph and the type detection model, solve the waste of human, material and financial resources caused by the manual determination method in the related technology, and the efficiency is low, and the accuracy of the determination is Low problem, realize the automatic determination of the type of gene sequencing results, save manpower, material and financial resources, and improve the efficiency and accuracy of determination.

Example two

FIG. 2 is a schematic flowchart of a method for establishing a type detection model applicable to Embodiment 2 of the present application. This embodiment is implemented based on the foregoing embodiment. Before the feature data is input into the type detection model to obtain a type that matches the sequencing result of the gene to be tested, the method further includes the following steps:

S210. Obtain peak shape map samples corresponding to multiple types of standard gene sequencing results.

In an embodiment, the peak shape sample may be obtained from a result of one generation sequencing. In an embodiment, the peak shape sample may be selected from a database of historical sequencing results. Exemplarily, a plurality of different types of peak shapes are selected from the historical sequencing results of one generation sequencing, and then these peak shapes are classified and labeled with corresponding classification labels, so as to obtain a plurality of different types of standard gene sequencing results. Peak shape sample. In an embodiment, there are multiple peak shape sample samples for each type of standard gene sequencing result. In an embodiment, the method for classifying the obtained peak shape map may be a manual determination classification method. For example, if it is determined manually that there is no abnormal peak shape in the peak shape map, it is determined as a normal type, and the peak shape is determined. The normal type label is marked on the figure as the peak shape sample of the standard gene sequencing result corresponding to the normal type. The peak shape icon that contains abnormal peak shapes in other peak shapes is marked with the abnormal type label as the standard gene corresponding to the abnormal type. Peak plot sample of sequencing results.

S220. Extract feature data corresponding to the sequencing result of each standard gene from the peak shape sample of each standard gene sequencing result as a feature data sample.

In an embodiment, the characteristic data sample is data capable of reflecting the peak characteristics in the peak shape sample and / or sequence information data corresponding to a standard gene sequencing result. For example, the characteristic data sample includes, but is not limited to, a base sequence and a base quality value. , The peak width value, the length of a base fragment with a length of more than 20 consecutive longest bases in the sequence, the signal intensity of multiple bases in the sequence, and the average signal intensity.

Exemplarily, feature data may be directly extracted from the obtained peak shape sample file, such as directly extracting sequence information, base quality values, peak width values, etc .; or, the feature data may be obtained through one or more preset feature extraction algorithms. One or more characteristic data corresponding to the peak shape sample, such as a length value of a base fragment with a continuous longest base mass greater than 20 in the sequence, a signal intensity of multiple bases in the sequence, and an average signal intensity. For a practical example, from the abl format file of the peak shape sample, directly extract the sequence information, base quality value and peak width value contained in the parameter data in the file; for another example, for the peak shape sample, run the include There is a script file of the base signal intensity extraction algorithm to obtain the signal intensity of each base in the peak shape graph.

S230: Use the feature data samples to train the set classification algorithm model to obtain a type detection model.

The classification algorithm model in this embodiment may be a training model based on a multi-classification algorithm. For example, the multi-classification algorithm includes, but is not limited to, a gradient boosting tree classification (Gradient Boosting Classifier) algorithm, a decision tree classification (Decision tree classifier) algorithm, and an extreme random tree. Classification (Extra Tree Classifier) algorithm, Random Forest Classification (Random Forest Classifier) algorithm, etc. In an embodiment, the process of training the classification algorithm model may be a process of adjusting multiple model parameters. After continuous training, the optimal model parameters are obtained. The classification algorithm model with the optimal model parameters is the final result to be obtained. model. Exemplarily, after obtaining multiple types of multiple characteristic data samples, the classification algorithm model is trained using the multiple characteristic data samples, and the model parameters in the classification algorithm model are continuously adjusted so that the classification algorithm model has input characteristic data. The ability to make type determinations to obtain a type detection model.

In one embodiment, the classification algorithm model can be determined according to the type of data used in the feature data sample. For example, for a base fragment length and average signal intensity that includes an average base quality and a continuous longest base quality greater than 20 The characteristic data samples of the three types of data can be selected based on any of Gradient, Boosting, Classifier, Decision, TreeClassifier, Extra Tree, Classifier, and Random Forest Classifier algorithms as the set classification algorithm model.

In one embodiment, using the feature data samples to train the set classification algorithm model to obtain a type detection model includes: using the feature data samples to train a plurality of different classification algorithm models; obtaining a plurality of different classification algorithms after a set number of trainings The recognition accuracy corresponding to the models respectively; the classification algorithm model with the highest recognition accuracy is determined as the type detection model.

In one embodiment, in order to improve the recognition accuracy of the type detection model, multiple different classification algorithm models may be trained at the same time, and the model with the highest recognition accuracy rate among the multiple different classification algorithm models is selected as the type detection model. Exemplarily, a feature data sample including three types of data including an average base quality, a base fragment length of more than 20 consecutive base lengths, and an average signal intensity is used. The classification algorithm model is trained, and after a set number of trainings, the recognition accuracy corresponding to multiple classification algorithm models can be obtained.

Table 1. The algorithm used under specific feature data and the classification algorithm model based on the algorithm after training

Recognition accuracy

algorithm	Recognition accuracy
Gradient Boosting Classifier	86.3%
Bernoulli NB	49.7%
Decision TreeClassifier	80.5%
Extra TreeClassifier	79.4%
Gaussian NB	72.7%

K Neighbors Classifier	74.1%
Label Propagation	17.2%
Label Spreading	17.2%
Linear Discriminant Analysis	71.7%
Linear SVC	71.7%
MLPC lassifier	61.0%
Nearest Centroid	34.2%
Quadratic Discriminant Analysis	73.5%
Random Forest Classifier	85.1%
SVC	52.5%
AdaBoost Classifier	63.5%
Gaussian ProcessClassifier	41.6%

In Table 1 above, the classification algorithm models with recognition accuracy rates above 60% are all optional classification algorithm models. In an embodiment, the classification algorithm model with the highest recognition accuracy in Table 1 may also be selected, for example, a training model based on the Gradient Boosting Classifier algorithm, as the input feature data is the average base quality and the continuous longest base quality is greater than 20 The base fragment length and average signal intensity are the three data when the training model is selected.

In the technical solution of this embodiment, before using a classification detection model to determine the type of the sequencing result of the gene to be tested, the peak shape map samples of the standard gene sequencing results corresponding to multiple types are obtained, and the corresponding ones are extracted from the peak shape map samples. The feature data is used as the feature data sample, and then the set classification algorithm model is trained using the feature data sample to obtain the type detection model, and the establishment of the type detection model is realized, so as to provide a model for automatic determination of the type of gene sequencing results. Basically, it improves the accuracy of judgment.

Example three

FIG. 3 is a schematic structural diagram of a detection device for a type of gene sequencing result provided in Embodiment 3 of the present application. Referring to FIG. 3, the type detection device for the type of genetic sequencing results includes a peak shape acquisition module 310, a feature extraction module 320, and a type detection module 330. Each module is described below.

A peak shape acquisition module 310 is configured to obtain a peak shape map of a sequencing result of a gene to be tested; a feature extraction module 320 is configured to extract feature data corresponding to the sequencing result of the gene to be tested according to the peak shape map; a type detection module 330, It is configured to input the feature data into a type detection model to obtain a type matching the sequencing result of the gene to be tested.

In an embodiment, the device for detecting the type of the genetic sequencing result may further include a sample acquisition module configured to input the feature data into a type detection model and obtain a type matching the sequencing result of the gene to be tested. Acquire peak shape map samples of standard gene sequencing results corresponding to multiple types; the data extraction module is configured to extract feature data corresponding to each standard gene sequencing result from the peak shape map samples of each standard gene sequencing result, as Feature data samples; a model training module configured to use the feature data samples to train a set classification algorithm model to obtain the type detection model.

In an embodiment, the model training module is configured to: use the feature data samples to train a plurality of different classification algorithm models; and obtain a recognition accuracy rate corresponding to each of the plurality of different classification algorithm models after a set number of trainings. Determining the classification algorithm model with the highest recognition accuracy as the type detection model.

In an embodiment, the feature extraction module 320 is configured to: extract feature data from a file of the peak shape map; and / or process the peak shape map according to a preset feature extraction algorithm to obtain corresponding feature data .

In an embodiment, the types include a normal type and an abnormal type.

In an embodiment, the type of abnormality includes one of a poly structure or a repeating sequence type, a spectral pull-up type, a diffusion type, a bubble type, an attenuation type, a peak set type, a no signal type, a primer problem type, and an interruption type. Or multiple types.

The above product can execute the method provided by any embodiment of the present application, and has corresponding functional modules and beneficial effects for executing the method.

Example 4

FIG. 4 is a schematic structural diagram of a computer device provided in Embodiment 4 of the present application. As shown in FIG. 4, a computer device provided in this embodiment includes a processor 41 and a memory 42. The processor in the computer device may be one or more. In FIG. 4, a processor 41 is used as an example. The processor 41 and the memory 42 in the computer device may be connected through a bus or other methods. Take bus connection as an example.

The processor 41 of the computer equipment in this embodiment integrates the detection device of the type of the result of gene sequencing provided in the above embodiment. In addition, the memory 42 in the computer device serves as a computer-readable storage medium, and may be configured to store one or more programs. The programs may be software programs, computer-executable programs, and modules, such as genes in the embodiments of the present application. Program instructions / modules corresponding to the detection method of the sequencing result type (for example, the modules in the genetic sequencing result type detection device shown in FIG. 3 include a peak shape acquisition module 310, a feature extraction module 320, and a type detection module 330) . The processor 41 executes various functional applications and data processing of the device by running software programs, instructions, and modules stored in the memory 42, that is, a method for detecting a type of genetic sequencing result in the foregoing method embodiment.

The memory 42 may include a program storage area and a data storage area. In one embodiment, the storage program area may store an operating system and application programs required for at least one function; the storage data area may store data created according to the use of the device, and the like. In addition, the memory 42 may include a high-speed random access memory, and may further include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some examples, the memory 42 may include memory remotely set relative to the processor 41, and these remote memories may be connected to the device through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

In addition, when the one or more programs included in the computer device are executed by the one or more processors 41, the program performs the following operations: obtaining a peak shape diagram of the sequencing result of the gene to be tested; and extracting the gene to be tested according to the peak shape diagram. The characteristic data corresponding to the sequencing result; the characteristic data is input into the type detection model to obtain a type that matches the sequencing result of the gene to be tested.

Example 5

Embodiment 5 of the present application further provides a computer-readable storage medium having stored thereon a computer program that, when executed by a detection device for a type of genetic sequencing result, implements detection of the type of gene sequencing result provided in Embodiment 1 of the present application. The method includes: obtaining a peak shape diagram of the sequencing result of the gene to be tested; extracting characteristic data corresponding to the sequencing result of the gene to be tested according to the peak shape diagram; and inputting the characteristic data into the type detection model to obtain a result that is comparable to the sequencing result of the gene to be tested Match type.

The computer-readable storage medium provided in the embodiments of the present application is not limited to the implementation of the method operations described above when the computer program stored thereon is executed, and can also implement the type of gene sequencing results provided by any embodiment of the present application. Relevant operations in the detection method.

Through the foregoing description of the embodiments, those skilled in the art may understand that the present application may be implemented by software and general hardware, and may also be implemented by hardware. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a computer's floppy disk, Read-Only Memory (ROM), Random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including multiple instructions to enable a computer device (can be a personal computer, a server, or a network device, etc.) to execute any of this application The method described in the examples.

In the embodiment of the above-mentioned type of genetic sequencing result detection device, the multiple units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, each The names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the present application.

Claims (10)

1. A method for detecting the type of genetic sequencing results, including:

   Obtain a peak shape diagram of the sequencing result of the gene to be tested;

   Extracting characteristic data corresponding to the sequencing result of the gene to be tested according to the peak shape map;

   The feature data is input into a type detection model to obtain a type that matches the sequencing result of the gene to be tested.
2. The method according to claim 1, before the inputting the characteristic data into a type detection model to obtain a type matching the sequencing result of the gene to be tested, further comprising:

   Obtain the peak shape samples of the standard gene sequencing results corresponding to multiple types respectively;

   Extracting characteristic data corresponding to each standard gene sequencing result from the peak shape sample corresponding to each standard gene sequencing result as a characteristic data sample;

   The set classification algorithm model is trained using the feature data samples to obtain the type detection model.
3. The method according to claim 2, wherein the training the set classification algorithm model using the feature data samples to obtain the type detection model comprises:

   Training multiple different classification algorithm models using the feature data samples;

   Obtaining the recognition accuracy rates corresponding to the multiple different classification algorithm models after a set number of trainings;

   The classification algorithm model with the highest recognition accuracy rate is determined as the type detection model.
4. The method according to any one of claims 1 to 3, wherein the extracting characteristic data corresponding to the sequencing result of the gene to be tested according to the peak shape graph comprises at least one of the following:

   Feature data is extracted from a file of the peak shape map; the peak shape map is processed according to a preset feature extraction algorithm to obtain corresponding feature data.
5. The method according to any one of claims 1-4, wherein the types include a normal type and an abnormal type.
6. The method according to claim 5, wherein the type of abnormality includes: poly-poly structure or repeating sequence type, spectral pull-up type, dispersion type, bubble type, attenuation type, set peak type, no signal type, primer problem At least one of a type and an interrupt type.
7. A detection device for a type of gene sequencing result, including:

   The peak shape acquisition module is configured to obtain a peak shape map of the sequencing result of the gene to be tested;

   A feature extraction module configured to extract feature data corresponding to the sequencing result of the gene to be tested according to the peak shape map;

   The type detection module is configured to input the feature data into a type detection model to obtain a type that matches the sequencing result of the gene to be tested.
8. The apparatus according to claim 7, further comprising:

   The sample acquisition module is configured to obtain peak shapes of standard gene sequencing results corresponding to multiple types before the characteristic data is input into a type detection model and a type matching the sequencing result of the gene to be tested is obtained. Sample map

   A data extraction module configured to extract feature data corresponding to each standard gene sequencing result from a peak shape sample of each standard gene sequencing result as a feature data sample;

   A model training module is configured to use the feature data samples to train a set classification algorithm model to obtain the type detection model.
9. A computer device including:

   At least one processor;

   A memory configured to store at least one program;

   When the at least one program is executed by the at least one processor, the at least one processor implements a method for detecting a type of a gene sequencing result according to any one of claims 1-6.
10. A computer-readable storage medium having stored thereon a computer program that, when executed by a processor, implements a method for detecting a type of genetic sequencing result according to any one of claims 1-6.

Images