Genetic variation site processing method, phenotype prediction method and system
By determining the correlation between genetic variant sites and mapping them to two-dimensional space, the problem that neural network models cannot effectively establish site connections is solved, the accuracy of processing results is improved, and the smooth progress of biological genetic research is ensured.
Patent Information
- Application Number
- CN202411147570.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-08-20
AI Technical Summary
When the existing neural network model deals with genetic variant sites, due to the long distribution distance of the associated genetic variant sites, the model cannot effectively establish a connection between sites, resulting in low accuracy of processing results, affecting biological genetic research. smooth progress.
By determining the site sequence of biological samples to be processed, the association relationship between each genetic variant site is determined and mapped to two-dimensional space, the site object is generated, and the site processing model is input to the site processing model to establish the Contact to achieve accurate processing results.
The link between genetic variant sites has been effectively established, the accuracy of processing results has been improved, and the smooth progress of biological genetic research has been ensured.
Smart Images

Figure CN118675613B_ABST
Abstract
Description
Technical Field
[0001] The embodiments of this specification relate to the field of genetic biotechnology, and in particular, to a method for processing genetic variation sites; one or more embodiments of this specification also relate to a phenotype prediction method, another phenotype prediction method, a phenotype prediction system, a computing device, a computer-readable storage medium, and a computer program product. Background Art
[0002] In the field of biological genetics, analyzing and processing genetic variation sites in gene sequences is an important task in biological genetics research. With the continuous development of artificial intelligence technology, neural network models have also been introduced in biological genetics research to assist biological genetics research.
[0003] At present, in the process of using neural network models to analyze and process genetic variation site sequences, since the associated genetic variation sites may be distributed far apart in the genetic variation site sequence, the neural network model cannot effectively establish the connection between the genetic variation sites and perform analysis and processing, resulting in the problem of low accuracy of the processing results output by the neural network model. Therefore, the smooth progress of biological genetic research is seriously affected. Summary of the invention
[0004] In view of this, the embodiments of this specification provide a method for processing genetic variation sites. One or more embodiments of this specification also relate to a phenotype prediction method, another phenotype prediction method, a phenotype prediction system, a phenotype prediction device, another phenotype prediction device, a computing device, a computer-readable storage medium, and a computer program product to solve the technical defects existing in the prior art.
[0005] According to a first aspect of an embodiment of this specification, a method for processing a genetic variation site is provided, comprising:
[0006] Determining a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites;
[0007] Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0008] The site object is input into a site processing model to obtain processing results corresponding to the multiple genetic variation sites.
[0009] According to a second aspect of the embodiments of this specification, a device for processing a genetic variation site is provided, comprising:
[0010] A data determination module is configured to determine a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites;
[0011] an object acquisition module, configured to determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0012] The object processing module is configured to input the site object into a site processing model to obtain processing results corresponding to the multiple genetic variation sites.
[0013] According to a third aspect of the embodiments of this specification, a phenotype prediction method is provided, comprising:
[0014] Determining a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites;
[0015] Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0016] The site object is input into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
[0017] According to a fourth aspect of the embodiments of this specification, a phenotype prediction device is provided, comprising:
[0018] A data determination module is configured to determine a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites;
[0019] an object acquisition module, configured to determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0020] The object processing module is configured to input the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
[0021] According to a fifth aspect of an embodiment of this specification, a phenotype prediction method is provided, which is applied to a cloud-side device, including:
[0022] A site sequence of a biological sample to be processed sent by a receiving device, wherein the site sequence includes a plurality of genetic variation sites;
[0023] Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0024] Inputting the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites;
[0025] The phenotype prediction result is sent to the end-side device.
[0026] According to a sixth aspect of an embodiment of this specification, a phenotype prediction device is provided, which is applied to a cloud-side device, including:
[0027] A data receiving module, configured to receive a site sequence of a biological sample to be processed sent by a terminal device, wherein the site sequence includes a plurality of genetic variation sites;
[0028] an object acquisition module, configured to determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0029] An object processing module is configured to input the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites;
[0030] The result sending module is configured to send the phenotype prediction result to the terminal device.
[0031] According to a seventh aspect of the embodiments of this specification, a phenotype prediction system is provided, including a client and a server, wherein:
[0032] The client is configured to send a site sequence of a biological sample to be processed to the server, wherein the site sequence includes a plurality of genetic variation sites;
[0033] The server is configured to receive the site sequence of the biological sample to be processed sent by the client,
[0034] Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites,
[0035] The site object is input into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
[0036] According to an eighth aspect of the embodiments of this specification, a computing device is provided, including:
[0037] Memory and processor;
[0038] The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, the steps of any one of the above methods are implemented.
[0039] According to a ninth aspect of the embodiments of this specification, a computer-readable storage medium is provided, which stores a computer program / instruction, and the computer program / instruction implements the steps of any one of the above methods when executed by a processor.
[0040] According to the tenth aspect of the embodiments of this specification, a computer program product is provided, including a computer program / instruction, which implements the steps of any one of the above methods when executed by a processor.
[0041] One or more embodiments of the present specification provide a method for processing genetic variation sites. In the process of processing the genetic variation sites, the method can determine the correlation relationship between the genetic variation sites, and map the site sequence to a two-dimensional space based on the correlation relationship to obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the correlation relationship between the genetic variation sites, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the correlation relationship into a site processing model for processing, the site processing model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate processing results, avoiding the problem of low accuracy of the processing results output by the neural network model, and ensuring the smooth progress of biological genetic research. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 This is a schematic diagram of an application of a method for processing a genetic variation site provided by an embodiment of this specification;
[0043] Figure 2It is a flow chart of a method for processing a genetic variation site provided by one embodiment of this specification;
[0044] Figure 3 is a processing flow chart of a method for processing a genetic variation site provided by an embodiment of this specification;
[0045] Figure 4 is a flow chart of a phenotype prediction method provided by one embodiment of this specification;
[0046] Figure 5 is a flow chart of another phenotype prediction method provided by one embodiment of this specification;
[0047] Figure 6 is a schematic diagram of the structure of a phenotype prediction system provided by an embodiment of this specification;
[0048] Figure 7 It is a structural block diagram of a computing device provided by an embodiment of this specification. DETAILED DESCRIPTION
[0049] Many specific details are described in the following description to facilitate a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of this specification, so this specification is not limited to the specific implementation disclosed below.
[0050] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "said" and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.
[0051] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of this specification, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".
[0052] In addition, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.
[0053] In one or more embodiments of this specification, a large model refers to a deep learning model with large-scale model parameters, which usually contains hundreds of millions, tens of billions, hundreds of billions, trillions, or even more than 10 trillion model parameters. A large model can also be called a foundation model / foundation model. The large model is pre-trained with large-scale unlabeled corpus to produce a pre-trained model with more than 100 million parameters. This model can adapt to a wide range of downstream tasks, and the model has good generalization ability, such as a large-scale language model (LLM), a multi-modal pre-training model, etc.
[0054] When the big model is used in practice, only a small number of samples are needed to fine-tune the pre-trained model and it can be applied to different tasks. The big model can be widely used in natural language processing (NLP), computer vision and other fields. Specifically, it can be applied to computer vision tasks such as visual question answering (VQA), image caption (IC), image generation, as well as natural language processing tasks such as text-based sentiment classification, text summary generation, and machine translation. The main application scenarios of the big model include digital assistants, intelligent robots, search, online education, office software, e-commerce, intelligent design, etc.
[0055] First, the terms involved in one or more embodiments of this specification are explained.
[0056] Genotype: refers to the general term for all gene combinations of an individual organism; it reflects the genetic composition of the organism, that is, the sum of all genes obtained from its parents; the genotype used in genetics often refers to the genotype of a certain trait. If two organisms have a different genotype at one point, their genotypes are different, so genotype refers to all combinations of all alleles at all loci of an individual.
[0057] Phenotype: refers to the phenotype, or phenotype for short; phenotype refers to the sum of traits exhibited by an individual with a specific genotype under certain environmental conditions. The so-called traits refer to the morphological, structural, physiological, and biochemical characteristics of an organism.
[0058] Deep learning: It is the study of the inherent laws and representation levels of sample data. The information obtained in the learning process is of great help in the interpretation of data such as text, images and sounds. Its ultimate goal is to enable machines to have analytical learning capabilities like humans and to recognize data such as text, images and sounds. Deep learning is a complex machine learning algorithm that has achieved results in speech and image recognition that far exceed previous related technologies.
[0059] Convolutional neural network: It is a type of feedforward neural network with deep structure and convolutional calculation. It is one of the representative algorithms of deep learning. Convolutional neural network has the ability of representation learning and can classify input information in a translation-invariant manner according to its hierarchical structure. Therefore, it is also called "translation-invariant artificial neural network".
[0060] Transformer model: It is a deep learning model based entirely on the self-attention mechanism. It is suitable for parallel computing, and due to the complexity of the model itself, it is more accurate and performs better than the previously popular RNN recurrent neural network. The Transformer model is essentially an Encoder-Decoder architecture, consisting of an encoding component and a decoding component; the encoding component consists of multiple layers of encoders (six layers of encoders are used), and the decoding component consists of decoders with the same number of layers. Each encoder contains two sublayers, namely the Self-Attention layer and the Feed Forward Network (FFN); the Transformer model is a Self-Attention structure, in which the Q (query), K (key value), and V (value) matrices used are obtained by linear transformation of the output; there are multiple Self-Attention structures in the model, which can capture the correlation coefficient attention score on multiple dimensions.
[0061] SNP: refers to single nucleotide polymorphism, which is a DNA sequence polymorphism caused by the variation of a single nucleotide at the genome level.
[0062] AI (Artificial Intelligence): refers to artificial intelligence.
[0063] GBLUP: A model for estimating breeding values (BLUP) of individuals.
[0064] SVM (Support Vector Machine): refers to support vector machine, which is a generalized linear classifier that performs binary classification of data using supervised learning.
[0065] SVs (Structure Variantions, abbreviated as SVs): refers to structural variations. SVs refer to long-length sequence changes and positional relationship changes on the genome.
[0066] In the field of biological genetics, analyzing and processing genetic variation sites in gene sequences is an important task in biological genetics research. With the continuous development of artificial intelligence technology, neural network models have also been introduced in biological genetics research to assist biological genetics research. For example, AI-based smart breeding has become a current research hotspot, and predicting phenotypes (traits) through genetic variation SNPs is an important part of AI breeding. However, the original total amount of genetic variation sites is particularly large, usually in the millions to tens of millions, making it impossible for neural networks to perform effective analysis.
[0067] To address this problem, this specification provides a solution that uses bioinformatics knowledge or feature engineering to reduce the dimensionality of the original full genetic variation data, and then builds a prediction model on the reduced dimensionality data. The prediction models are mainly divided into two categories:
[0068] The first type is: machine learning methods.
[0069] Among them, the representative of this method is the GBLUP model, and many variant solutions have evolved with the GBLUP model as the core. Machine learning methods such as random forests and SVM can handle scenarios with higher dimensions and small sample sizes. However, since the models are often based on artificially designed features and have limited model representation capabilities, they do not perform well for complex tasks.
[0070] The second type is: deep learning methods.
[0071] This method automatically extracts useful features from the input data through a deep model. The most commonly used one is the convolutional neural network, which can be used to extract local features. However, the model in this method has major defects: first, due to the large number of learning parameters, the model is often unable to handle particularly long site inputs; second, the model cannot aggregate the associations between distant sites, which limits the further development of this type of method.
[0072] Based on this, it can be seen that this solution has major flaws. First, in the process of screening the original sites, the dimension is reduced to tens of thousands to hundreds of thousands, and this process will lose genetic information; second, the current deep learning methods mostly use convolutional neural networks. The advantage of this type of model is the extraction of local features, and the distance between the associated sites in this task may be very far, so how to more effectively extract long-distance sites is also very critical; moreover, useful genetic sites are often relatively small, and a large number of SNPs are invalid, so how to avoid the influence of massive interference factors and focus on key sites is also a problem that needs to be solved.
[0073] Based on this, in this specification, a method for processing a genetic variation site is provided. One or more embodiments of this specification also involve another method for processing a genetic variation site, a phenotype prediction method, a system for processing a genetic variation site, a processing device for a genetic variation site, another processing device for a genetic variation site, a computing device, a computer-readable storage medium, and a computer program product, which are described in detail one by one in the following embodiments.
[0074] See also Figure 1 , Figure 1 A schematic diagram of an application of a method for processing a genetic variation site according to an embodiment of the present specification is shown. Figure 1 It can be seen that the user can send the SNPs site sequence to the server 104 through the terminal 102 for phenotype prediction. After receiving the SNPs site sequence (i.e., input data) for phenotype prediction, the server 104 multiplies the site matrix M of all samples and the matrix transpose to determine the similarity of each site with all other sites, and maps all sites to a two-dimensional space based on the similarity, performs image conversion on the sites in the two-dimensional space, and obtains a site image; then the site image is input into the phenotype prediction model for processing to obtain the predicted value of the phenotype. The server 104 sends the predicted value of the phenotype output by the representation prediction framework to the terminal 102, thereby meeting the user's phenotype prediction needs.
[0075] See also Figure 2 , Figure 2 A flow chart of a method for processing a genetic variation site provided according to an embodiment of the present specification is shown, which specifically includes the following steps.
[0076] Step 202: Determine a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites.
[0077] Among them, the biological sample to be processed can be understood as a biological sample whose genetic variation sites need to be processed. When the method for processing genetic variation sites provided in this specification is applied to different scenarios, the biological sample to be processed is also different; when the method for processing genetic variation sites is applied to agricultural breeding scenarios, the biological sample to be processed can be a crop sample to be processed; when the method for processing genetic variation sites is applied to animal husbandry scenarios, the biological sample to be processed can be an animal sample to be processed; when the method for processing genetic variation sites is applied to microbial scenarios, the biological sample to be processed can be a microbial sample to be processed; that is, the biological sample to be processed can be an economic crop sample to be processed, an animal sample to be processed, a plant sample to be processed and / or a microbial sample to be processed, etc.
[0078] A site sequence can be understood as a sequence composed of multiple genetic variation sites arranged in order; the arrangement in order can be performed in a random arrangement manner, by sorting according to the acquisition time of the genetic variation sites, or by sorting according to the site position of the genetic variation sites in the gene sequence.
[0079] In one or more embodiments provided in this specification, the site sequence is one-dimensional data, for example, the site sequence is a long sequence, or a full site sequence; the long sequence or the full site sequence contains a full range of genotype genetic variation sites.
[0080] The genetic variation site can be understood as a difference site occurring on a DNA sequence with a fixed position in the genome, and the genetic variation site includes but is not limited to SNPs and SVs.
[0081] In one or more embodiments provided in this specification, in order to meet the user's processing operations on genetic variation sites, the method provided in this specification can receive a site sequence sent by the user and process the site sequence. The specific implementation method is as follows.
[0082] Determine the site sequence of the biological sample to be processed, including:
[0083] The site sequence of the biological sample to be processed sent by a client is received, wherein the site sequence is sent by the client when a user performs a site sequence upload operation.
[0084] Specifically, the method for processing genetic variation sites provided in this specification can be applied to a server, which can receive the site sequence of the biological sample to be processed sent by a client; the specific method is:
[0085] The user performs a site sequence upload operation on the site processing page of the client, thereby providing the site sequence of the biological sample to be processed to the client, and the client uploads the site sequence of the biological sample to be processed to the server; wherein, the site processing page can be an application page, a web page, a human-computer interaction page, etc., and the user can perform a site sequence upload operation through the site processing control corresponding to the site processing page, thereby uploading the site sequence to the client; the site processing control can be a page control such as a page button, a text input box, or the site processing control can be a hardware device such as a button device, a touch screen, a mouse, a keyboard, etc.
[0086] The server can receive the site sequence of the biological sample to be processed sent by the client, and subsequently determine the processing results corresponding to multiple genetic variation sites based on the site sequence; thereby, through the interaction between the client and the server, the user's processing needs for genetic variation sites are met.
[0087] Step 204: Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object contains multiple genetic variation sites.
[0088] Among them, the association relationship can be understood as information used to characterize the association between each genetic variation site, for example, the association relationship can be understood as a relationship vector, a relationship matrix, etc.; the relationship vector or relationship matrix can characterize the relationship between one genetic variation site among multiple genetic variation sites and other genetic variation sites; in one or more embodiments provided in this specification, the association relationship is similarity.
[0089] The two-dimensional space can be understood as a two-dimensional space coordinate system composed of abscissas and ordinates, or the two-dimensional space can be a two-dimensional visualization interface for genetic variation sites.
[0090] A site object can be understood as two-dimensional data that can represent multiple genetic variation sites. For example, the site object can be two-dimensional data such as a site image, grid data, or scatter plot data. In one or more embodiments provided in this specification, the site object can be a site image. The site object can be understood as an object obtained after performing data dimensionality upgrading processing on each genetic variation site.
[0091] In one or more embodiments provided in the present specification, after determining the association relationship between each genetic variation site, it also includes: mapping the site sequence to a three-dimensional space based on the association relationship to obtain a site object corresponding to the three-dimensional space; wherein the site object can be understood as three-dimensional data that can represent multiple genetic variation sites, for example, the site object can be three-dimensional data such as a 3D model, 3D point cloud data, etc.; in one or more embodiments provided in the present specification, the site object can be site 3D point cloud data.
[0092] Taking the application of the processing method of genetic variation sites provided in this specification in the characterization prediction scenario as an example, the processing method of the genetic variation sites is explained, wherein the biological sample to be processed can be a crop sample to be processed, multiple genetic variation sites are SNPs, the site sequence can be a SNP sequence, the association relationship can be a relationship vector, the site object is a site image, and the association relationship is similarity.
[0093] Based on this, after determining the SNP sequence including SNPs, the similarity between each SNP in the SNP sequence and other SNPs is calculated to determine the correlation between the SNPs; based on the similarity, the SNPs are mapped into a two-dimensional space so that the distribution distance between the associated SNPs is closer, and then the SNPs in the two-dimensional space are converted to obtain the SNPs site image corresponding to the SNPs, so as to facilitate the subsequent phenotypic prediction through the SNPs site image.
[0094] In one or more embodiments provided in this specification, in the process of determining the association relationship between each genetic variation site, similarity can be used as the association relationship between each genetic variation site; that is, the similarity between each genetic variation site is used as the association relationship between each genetic variation site; the specific implementation method is as follows.
[0095] Determine the association between various genetic variation sites, including:
[0096] Determine the similarity between a target genetic variation site and other genetic variation sites in a site sequence, and determine the similarity as the association relationship between the genetic variation sites, wherein the target genetic variation site is any one of the multiple genetic variation sites, and the other genetic variation sites are genetic variation sites among the multiple genetic variation sites except the target genetic variation site.
[0097] Among them, the similarity can be understood as information used to characterize the degree of association between each genetic variation site and other genetic variation sites. The higher the similarity, the higher the degree of association between the two genetic variation sites. The similarity can be cosine similarity; or the similarity can be a similarity that uses distance to characterize the degree of similarity, for example, Euclidean distance.
[0098] Continuing with the above example, after determining the SNP sequence including SNPs, the similarity between each SNP in the SNP sequence (i.e., the target genetic variation site) and other SNPs in the SNP sequence (i.e., other genetic variation sites) is calculated to determine the similarity between a certain SNP and other SNPs; this similarity is then used as an association relationship that characterizes the degree of association between a certain SNP and other SNPs; thereby, in the process of processing the genetic variation sites, the association relationship between the genetic variation sites is taken into consideration, which facilitates the subsequent determination of accurate processing results.
[0099] In one or more embodiments provided in this specification, the site sequence may be multiple, and each site sequence in the multiple site sequences includes multiple corresponding genetic variation sites. For example, the multiple site sequences are M, and each site sequence includes N genetic variation sites, wherein M and N can be set according to the application scenario; in this case, in the process of calculating the similarity between each genetic variation site, the similarity between any one genetic variation site in the multiple site sequences and other genetic variation sites can be calculated, and the specific implementation method is as follows:
[0100] There are at least two site sequences, and each site sequence includes a plurality of genetic variation sites;
[0101] Determine the similarity between the target genetic variation site and other genetic variation sites in the site sequence, including:
[0102] Determine a target genetic variation site and other genetic variation sites in at least two site sequences, wherein the target genetic variation site is any one genetic variation site in the at least two site sequences, and the other genetic variation sites are other genetic variation sites in the at least two site sequences except the target genetic variation site;
[0103] The similarity between the target genetic variation site and the other genetic variation sites is determined.
[0104] Using the above example, for the problem of how to effectively convert the SNP sequence into an image, this method needs to determine each genetic variation site and other genetic variation sites, and calculate the similarity between the two, so as to obtain the similarity between each genetic variation site and other genetic variation sites. The specific method is:
[0105] First, after determining multiple SNP sequences including SNPs, determine the target SNP (i.e., target genetic variation site) in the multiple SNP sequences, where the target SNP is any one of the SNPs included in each SNP sequence; and, the SNPs other than the target SNP in the SNPs included in each SNP sequence are regarded as other SNPs.
[0106] Secondly, the similarity between the target SNP (i.e., the target genetic variation site) and other SNPs (i.e., other genetic variation sites) is calculated to determine the similarity between a certain SNP and other SNPs; the similarity can be represented by a vector; and the similarity is used as an association relationship to characterize the degree of association between a certain SNP and other SNPs; thereby, in the process of processing genetic variation sites, the association relationship between each genetic variation site is taken into consideration, which facilitates the subsequent determination of accurate processing results.
[0107] In one or more embodiments provided in this specification, in the process of determining the association relationship between each genetic variation site, multiple site sequences can be used as a site matrix, similarity calculation is performed based on the site matrix, and the similarity is used as the association relationship between each genetic variation site. That is to say, in the process of determining the association relationship between each genetic variation site, the specific way of calculating the association relationship can be: using multiple site sequences as a site matrix, wherein each site sequence contains multiple genetic variation sites; determining a transposed site matrix of the site matrix, calculating the similarity between each genetic variation site by using the site matrix and the transposed site matrix (for example, by multiplying the two matrices to obtain the similarity between each genetic variation site), and using the similarity as the association relationship between each genetic variation site; thereby facilitating the subsequent determination of accurate processing results based on the association relationship; the specific implementation method is as follows.
[0108] Determining the similarity between the target genetic variation site and the other genetic variation sites comprises:
[0109] Determine the at least two site sequences as a site matrix, and determine a transposed site matrix corresponding to the site matrix;
[0110] The site matrix and the transposed site matrix are multiplied to obtain a product result, and the similarity between the target genetic variation site and the other genetic variation sites is determined based on the product result.
[0111] The transposed site matrix may be understood as a matrix obtained by transposing the site matrix.
[0112] Wherein, determining the similarity between the target genetic variation site and the other genetic variation sites based on the multiplication result can be understood as taking the multiplication result as the similarity between the target genetic variation site and the other genetic variation sites.
[0113] Wherein, determining the similarity between the target genetic variation site and the other genetic variation sites based on the product result can also be understood as calculating the similarity between the target genetic variation site and the other genetic variation sites based on the product result; for example, a site matrix A, whose size is , then the transposed site matrix of the site matrix A is It will be a A and Multiplying them, we get The symmetric matrix (i.e., the product result) is usually used to calculate the similarity between row vectors; Multiplying by A, we get A symmetric matrix (i.e., the product result), which is usually used to calculate the similarity between column vectors.
[0114] Continuing with the above example, for the problem of how to effectively convert SNP sequences into images, this method needs to determine the relationship vectors between each genetic variation site. The specific method is: first, the site matrix M composed of multiple SNP sequences is transposed to obtain a site transposed matrix; second, the site matrix M of all samples and the site matrix transpose are multiplied to calculate the similarity between each genetic variation site and other genetic variation sites; finally, the similarity is used as the relationship vector between each genetic variation site and all other genetic variation sites.
[0115] In one or more embodiments provided in this specification, mapping the site sequence to a two-dimensional space based on the association relationship, and obtaining a site object corresponding to the two-dimensional space includes:
[0116] Determining distance parameters between the genetic variation sites based on the association relationship between the genetic variation sites, and mapping the multiple genetic variation sites included in the site sequence to a two-dimensional space based on the distance parameters;
[0117] Clustering the multiple genetic variation sites in the two-dimensional space to obtain at least one site set, wherein each site set contains associated genetic variation sites;
[0118] The genetic variation sites included in each site set are subjected to data conversion to obtain at least one site object.
[0119] Among them, the site set can be understood as a set of at least two associated genetic variation sites, and the genetic variation sites contained in the site set are genetic variation sites of the same type; based on this, the at least one site set can be at least one type of site set.
[0120] Following the above example, after obtaining the relationship vector R between each site (i.e., SNP) and all other sites, the site distance is determined according to the similarity between each genetic variation site, and these genetic variation sites are mapped into a two-dimensional space based on the site distance; then, a clustering algorithm is used to cluster the genetic variation sites in the two-dimensional space, so that the genetic variation sites are clustered into k categories, among which genetic variation sites with higher correlation are included in the same category;
[0121] Finally, the sites in each cluster category (i.e., site set) are rearranged into images. All cluster categories will form k images, forming an image sequence. It should be noted that in the process of converting to images, padding is used to ensure the consistency of image size between each category.
[0122] In the above embodiment, in the process of mapping the site sequence to a two-dimensional space based on the association relationship and obtaining the site object corresponding to the two-dimensional space, multiple genetic variation sites in the site sequence can be mapped to a two-dimensional space based on the association relationship, and the genetic variation sites in the two-dimensional space can be clustered to obtain at least one site set; then, the genetic variation sites in each site set are converted into a site object, thereby adjusting the site sequence based on the association relationship between the genetic variation sites, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the association relationship into the site processing model for processing, the site processing model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate processing results.
[0123] In one or more embodiments provided in this specification, the site set is one;
[0124] Performing data conversion on the genetic variation sites included in each site set to obtain at least one site object includes:
[0125] Determining the genetic variation sites included in the site set, and performing image data conversion on the genetic variation sites to obtain site image data;
[0126] Continuing with the above example, after obtaining the relationship vector R between each site and all other sites, the site distance is determined according to the similarity between the sites, and based on the site distance, these sites are mapped into a two-dimensional space, and clustered using a clustering algorithm to cluster the sites into one category, and the sites in the cluster category are converted into images to obtain site images.
[0127] The set of sites is at least two;
[0128] Performing data conversion on the genetic variation sites included in each site set to obtain at least one site object includes:
[0129] The genetic variation sites included in each of the at least two site sets are converted into image data to obtain at least two site image data, wherein the at least two site image data correspond to the at least two site sets.
[0130] Continuing with the above example, clustering is performed using a clustering algorithm to cluster the sites into k categories. Image conversion is then performed on the sites in each clustering category to obtain site images. Finally, all clustering categories will form k images, forming an image sequence.
[0131] It should be noted that in the process of converting the genetic variation site into image data, it is necessary to convert the genetic variation site into image data according to the attribute information of the image data (such as the number of sites, the distribution range of sites, the coordinates of sites, etc.), so as to obtain multiple site image data, and the specific methods include:
[0132] Determine the number of genetic variation sites included in each site set, and determine a maximum site set from at least two site sets based on the number of sites;
[0133] Determine the target site image size according to the site distribution range of the genetic variation sites included in the maximum site set in the two-dimensional space;
[0134] The site distribution range may be an irregular range, the site image size may be the size of the site image data, and the site image size may be represented by a rectangle or circle of a specific size.
[0135] Determine the size of other site images according to the site distribution range of the genetic variation sites included in other site sets in two-dimensional space, wherein the other site sets are other site sets in at least two site sets except the largest site set;
[0136] When the image size of other sites is smaller than the image size of the target site, the site expansion is performed on the other site sets to obtain an expanded genetic variation site set, wherein the expanded genetic variation site set includes genetic variation nodes and virtual expanded genetic variation sites, and the site distribution range of the genetic variation nodes and the virtual expanded genetic variation sites included in the expanded genetic variation site set in the two-dimensional space is the same as the site distribution range of the genetic variation sites included in the maximum site set in the two-dimensional space.
[0137] Image conversion is performed on the site distribution range in two-dimensional space according to the genetic variation sites included in the maximum site set to obtain a target site image, and image conversion is performed on the site distribution range in two-dimensional space according to the genetic variation sites included in the expanded genetic variation site set and the expanded genetic variation sites to obtain a target site image; wherein the image size of the target site image is consistent with that of the other site images.
[0138] Continuing with the above example, since the number of sites in each set is different, the image size (rectangular box) corresponding to the site image is different, which in turn leads to different sizes of the site image. To avoid this problem, some virtual sites can be filled in for site sets with fewer sites and more concentrated distribution ranges (i.e., expanding genetic variation sites) to ensure consistency in image size.
[0139] Step 206: Input the site object into a site processing model to obtain processing results corresponding to the plurality of genetic variation sites.
[0140] The site processing model can be understood as a model that can process genetic variation sites; the processing result can be understood as the processing result of the site processing model for the genetic variation site; when the genetic variation site processing method provided in this specification is applied to different scenarios, the site processing model is different, and the processing result is also different;
[0141] In the case where the processing method of genetic variation sites is applied to a phenotype prediction scenario, the site processing model may be a phenotype prediction model, which is used to predict phenotypes through genetic variation SNPs, and correspondingly, the processing result may be a phenotype prediction result; in the case where the processing method of genetic variation sites is applied to a genetic disease analysis scenario, the site processing model may be a genetic disease prediction model, which is used to predict genetic diseases through genetic variation SNPs, and correspondingly, the processing result may be a genetic disease prediction result; in the case where the processing method of genetic variation sites is applied to an identity recognition scenario, the site processing model may be an identity recognition model, which is used to perform identity recognition through genetic variation SNPs, and correspondingly, the processing result may be an identity recognition result; it should be noted that the site processing model may be a large model, and the site processing model (e.g., a phenotype prediction model) may include a feature extraction unit and a feature processing unit; the processing result may be a phenotype prediction result, and the phenotype prediction result may be understood as a result that characterizes the phenotype corresponding to the genetic variation site; the phenotype prediction result may be a phenotype, a predicted value of a phenotype, etc., and the predicted value of the phenotype may be represented by a preset numerical value, for example, the predicted value may be represented by a floating-point numerical value.
[0142] In one or more embodiments provided in this specification, inputting the site object into a site processing model to obtain processing results corresponding to the plurality of genetic variation sites includes:
[0143] Inputting the site object into the site processing model, wherein the site processing model includes a feature extraction unit and a feature processing unit;
[0144] Using the feature extraction unit to extract features from the site object to obtain site object features corresponding to the site object;
[0145] The feature processing unit is used to process the site object features and the site object to obtain processing results corresponding to the multiple genetic variation sites.
[0146] Among them, the feature extraction unit can be understood as a unit used to extract features from site objects. The feature extraction unit can be one or more network layers in the site processing model; the feature extraction unit can be a sub-model in the site processing model; or, the feature extraction unit can be one or more convolutional neural networks in the site processing model, for example, the feature processing unit can be a CNN.
[0147] The feature processing unit can be understood as a unit used to predict processing results based on site object features and site objects. For example, the feature processing unit can be a unit that performs phenotype prediction; the feature processing unit can be one or more network layers in a site processing model; the feature processing unit can be a sub-model in a site processing model; or, the feature processing unit can be a transformer.
[0148] In one or more embodiments provided in this specification, the site processing model can be understood as a deep learning model including a convolutional neural network and a transformer, which is used for phenotype prediction.
[0149] Continuing with the above example, the deep learning model (i.e., site processing model) is composed of network modules of CNNs (i.e., multiple convolutional neural networks) and transformers, where CNNs process the generated site images to produce corresponding image features, and transformers extract features from the input image features and site images to process the relationship between images. The entire deep learning model extracts features from two levels and finally predicts the final phenotypic value (i.e., the processing result) through a fully connected output layer.
[0150] In the above embodiment, the site processing model can use CNN and transformer to process the site object to obtain the processing results corresponding to the multiple genetic variation sites. By using the CNN combined with the transformer model structure as the backbone network to extract features, the CNN module can extract local features, and the transformer can associate global features, and the built-in self-attention machine can make the model focus on key sites; therefore, not only can it process extremely long site sequences without losing site information, but it can also learn the association between distant sites.
[0151] In one or more embodiments provided in this specification, there are at least two site objects, each of which contains a portion of genetic variation sites, and the portion of genetic variation sites is a subset of the plurality of genetic variation sites;
[0152] Utilizing the feature extraction unit to extract features from the site object to obtain site object features corresponding to the site object includes:
[0153] The feature extraction unit is used to extract features from each of the site objects to obtain site object features corresponding to each of the site objects.
[0154] Continuing with the above example, the deep learning model (i.e., site processing model) is composed of network modules of CNNs (i.e., multiple convolutional neural networks) and network modules of transformers, where CNNs process each image in the image sequence individually, extract more accurate image features, and finally generate a feature sequence of k images, thereby improving the accuracy of subsequent processing results.
[0155] In one or more embodiments provided in this specification, the using the feature processing unit to process the site object feature and the site object to obtain processing results corresponding to the multiple genetic variation sites includes:
[0156] Sorting the site object features corresponding to each site object to determine a site object feature sequence corresponding to each site object;
[0157] Sort the site objects to obtain a site object sequence;
[0158] The site object feature sequence and the site object sequence are input into the feature processing unit for processing to obtain processing results corresponding to the multiple genetic variation sites.
[0159] Continuing with the above example, the feature sequence of k images and the image sequence composed of k images are input into transformers; transformers process the relationship between these image sequences, and the entire model extracts features from two levels, and finally predicts the final phenotypic value through a fully connected output layer.
[0160] In the above embodiment, by using the 2D CNN+Transformer model structure, long-distance site features can be extracted and key sites can be focused on, thereby obtaining accurate phenotypic prediction values.
[0161] In one or more embodiments provided in this specification, after inputting the site object into the site processing model to obtain the processing results corresponding to the plurality of genetic variation sites, the method further includes:
[0162] The processing result is sent to the client for display.
[0163] Specifically, after obtaining the processing result, the processing can be sent to the client, so that the processing result can be displayed to the user through the client, thereby meeting the user's processing needs for genetic variation sites.
[0164] One or more embodiments of the present specification provide a method for processing genetic variation sites. In the process of processing the genetic variation sites, the method can determine the correlation relationship between the genetic variation sites, and map the site sequence to a two-dimensional space based on the correlation relationship to obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the correlation relationship between the genetic variation sites, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the correlation relationship into a site processing model for processing, the site processing model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate processing results, avoiding the problem of low accuracy of the processing results output by the neural network model, and ensuring the smooth progress of biological genetic research.
[0165] The following combination Figure 3 Taking the application of the method for processing genetic variation sites provided in this specification in a phenotype prediction scenario as an example, the method for processing genetic variation sites is further described. Figure 3 A processing flow chart of a method for processing a genetic variation site provided in an embodiment of the present specification is shown, which specifically includes the following steps.
[0166] Step 302: Determine the SNP site sequence of the long sequence type.
[0167] Among them, there are multiple SNP site sequences; each SNP site sequence is one-dimensional data; and multiple SNP site sequences can constitute a site matrix M; the size of the site matrix M can be N*M, wherein N represents the number of samples contained in each SNP site sequence, and M represents the number of SNP site sequences; each SNP site in the SNP site sequence or site matrix M represents the feature / gene expression value.
[0168] It should be noted that this solution is used to predict phenotypes based on SNP site sequences, and in the process of phenotype prediction, it can solve problems such as site loss, limited input sequence length, and the inability of the model to effectively link long-distance site relationships.
[0169] In addition, this solution can be applied to an agricultural breeding platform, and the SNP site sequence can be provided by the user to the agricultural breeding platform for phenotypic prediction through human-computer interaction.
[0170] Step 304: Determine the relationship vector of the SNP site sequence.
[0171] Specifically, this scheme can convert the input site sequence into an image form according to the similarity. In the process of image conversion, the relationship vector between each genetic variation site and all other genetic variation sites must first be determined; the specific implementation steps are:
[0172] 1. Transpose the site matrix M composed of multiple SNP sequences to obtain a site transposed matrix;
[0173] 2. The site matrix M of all samples is transposed and multiplied by the site matrix to calculate the similarity between each genetic variation site and other genetic variation sites;
[0174] 3. The similarity is used as the relationship vector between each genetic variation site and all other genetic variation sites.
[0175] The specific method is as follows: the similarity between a genetic variation site and all other genetic variation sites can be constructed as a relationship vector R that represents the degree of association; and the relationship vector R between all genetic variation sites and all other genetic variation sites can be constructed as a two-dimensional relationship vector matrix.
[0176] Step 306: Map the genetic variation sites to a two-dimensional space based on the relationship vectors and perform clustering.
[0177] Specifically, in the process of image conversion of genetic variation sites based on relationship vectors, all genetic variation sites can be mapped to a two-dimensional space based on the similarity in the relationship vectors, and clustering processing can be performed to obtain sets of genetic variation sites of multiple categories, which specifically includes the following steps:
[0178] 1. Determine the site distance between each site according to the similarity between each site, and map these sites into a two-dimensional space based on the site distance;
[0179] 2. Use a clustering algorithm to cluster the genetic variation sites in the two-dimensional space, thereby clustering the sites into K categories and obtaining a set of genetic variation sites in K categories.
[0180] Step 308: Perform image conversion on the genetic variation site to obtain a site image.
[0181] Specifically, this scheme can rearrange the sites contained in the genetic variation site set of each cluster category and convert it into an image. The specific implementation method is as follows:
[0182] 1. Determine the genetic variation sites contained in the genetic variation site sets of multiple clustering categories;
[0183] 2. Determine a rectangular frame based on the coordinate information of the genetic variation site;
[0184] 3. Visualize all genetic variation sites framed in the rectangular frame to obtain a site image I (i.e., genotype image) corresponding to each set of genetic variation sites.
[0185] It should be noted that, in the process of converting into images using the above method, padding can also be used to ensure the consistency of image size between each category.
[0186] The padding method means that since the number of sites in each set is different, the size of the rectangular box is different, and thus the size of the site image is different. To avoid this problem, some virtual sites can be filled for the set with a smaller number of sites to ensure the consistency of the image size.
[0187] Step 310: Extract features from the genotype image using a deep learning model.
[0188] The processing method of the deep learning model (i.e. the backbone network) can be found in Figure 6 , the deep learning model can be a representation prediction model, which is composed of a CNNs network module and a transformers network module; the specific implementation steps are:
[0189] 1. Input the site image sequence into the deep learning model, use CNNs to process each image individually, and finally generate an image feature sequence of k images.
[0190] 2. Input the site image sequence and image feature sequence into transformers, and use transformers to process the relationship between these image sequences. The entire model extracts features from two levels: image data and type data.
[0191] Step 312: Use the fully connected output layer to process the features output by the deep learning model to obtain the predicted value of the phenotype.
[0192] Based on the above steps, it can be seen that a method for processing genetic variation sites in this specification provides a deep learning method that supports the prediction of phenotypes from all genotype sites. This solution uses DeepInsight to convert SNP sequences into two-dimensional images, thereby converting one-dimensional problems into two dimensions, and building a 2D CNN combined with a transformer model structure. The 2D CNN+Transformer model structure is used to extract long-distance features and focus on key sites, thereby solving the problems of large input data dimensions making training difficult, loss of useful information in initial screening of input data, and the inability of existing models to effectively establish connections between long-distance sites.
[0193] In order to solve the two problems of large input data dimensions making training difficult and initial screening of input data resulting in loss of useful information, we adopted the problem conversion method, converting problems that are difficult to solve in one dimension to higher dimensions. The specific method is to use DeepInsight to convert the input sequence so that the model can not only process the original data without dimensionality reduction, but also gather similar information in closer locations.
[0194] In order to solve the problem that the model cannot effectively establish connections between distant sites, a 2D CNN combined with a Transformer solution was used. 2D CNN is used to extract local features, and Transformer is used to extract long-distance features. It focuses on key sites and automatically assigns higher weights to key information, thus solving the existing problem. In the end, this solution achieved results that were superior to traditional methods and existing deep learning models on long sequence input data.
[0195] See also Figure 4 , Figure 4 A flowchart of a phenotype prediction method provided according to an embodiment of the present specification is shown, which specifically includes the following steps.
[0196] Step 402: Determine a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites.
[0197] Step 404: Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object contains multiple genetic variation sites.
[0198] Step 406: Input the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
[0199] One or more embodiments of the present specification provide a phenotype prediction method. In the process of phenotype prediction based on genetic variation sites, the method can determine the correlation between each genetic variation site, and map the site sequence to a two-dimensional space based on the correlation to obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the correlation between each genetic variation site, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the correlation into a phenotype prediction model for processing, the phenotype prediction model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate phenotype prediction results, avoiding the problem of low accuracy of the phenotype prediction results output by the neural network model, and ensuring the smooth progress of biological genetic research.
[0200] The above is a schematic scheme of a phenotype prediction method of this embodiment. It should be noted that the technical scheme of the phenotype prediction method and the technical scheme of the above-mentioned method for processing a genetic variation site belong to the same concept, and the details not described in detail in the technical scheme of the phenotype prediction method can be referred to the description of the technical scheme of the above-mentioned method for processing a genetic variation site.
[0201] See also Figure 5 , Figure 5 A flowchart of another phenotype prediction method provided according to an embodiment of the present specification is shown. The phenotype prediction method is applied to a cloud-side device and specifically includes the following steps.
[0202] Step 502: receiving a site sequence of a biological sample to be processed sent by a terminal device, wherein the site sequence includes a plurality of genetic variation sites.
[0203] Step 504: Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object contains multiple genetic variation sites.
[0204] Step 506: Input the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
[0205] Step 508: Send the phenotype prediction result to the terminal device.
[0206] In one or more embodiments provided in this specification, the cloud-side device may be a central cloud device of a distributed architecture or an edge cloud device of a distributed architecture. The cloud-side device may be a cloud-side device with a cloud desktop system or cloud desktop software installed and deployed, such as a cloud server, cloud host, etc. The terminal-side device may be understood as any terminal that interacts with the cloud-side device for data exchange. The terminal may be a laptop, desktop computer, tablet computer, smart device, server, etc.
[0207] One or more embodiments of the present specification provide a phenotype prediction method applied to a cloud-side device. In the process of processing genetic variation sites, the method can determine the correlation between the genetic variation sites provided by the terminal-side device, and map the site sequence to a two-dimensional space based on the correlation, and obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the correlation between the genetic variation sites, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the correlation into a phenotype prediction model for processing, the phenotype prediction model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate phenotype prediction results, avoiding the problem of low accuracy of the phenotype prediction results output by the neural network model, and ensuring the smooth progress of biological genetic research.
[0208] The above is a schematic scheme of another phenotype prediction method of this embodiment. It should be noted that the technical scheme of the other phenotype prediction method and the technical scheme of the above-mentioned one phenotype prediction method belong to the same concept, and the details of the technical scheme of the other phenotype prediction method that are not described in detail can all be referred to the description of the technical scheme of the above-mentioned one phenotype prediction method.
[0209] See also Figure 6 , Figure 6 60 shows a schematic diagram of the structure of a phenotype prediction system provided by an embodiment of the present specification; the device includes a client 602 and a server 604, wherein:
[0210] The client 602 is configured to send a site sequence of a biological sample to be processed to the server 604, wherein the site sequence includes a plurality of genetic variation sites;
[0211] The server 604 is configured to receive the site sequence of the biological sample to be processed sent by the client 602,
[0212] Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites,
[0213] The site object is input into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
[0214] One or more embodiments of the present specification provide a phenotype prediction system. In the process of processing genetic variation sites, the client can send a site sequence of a biological sample to be processed to a server. The server can determine the association relationship between each genetic variation site in the site sequence, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the association relationship between each genetic variation site, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the association relationship into a phenotype prediction model for processing, the phenotype prediction model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate phenotype prediction results, avoiding the problem of low accuracy of the phenotype prediction results output by the neural network model, and ensuring the smooth progress of biological genetic research.
[0215] The above is a schematic scheme of a phenotype prediction system of this embodiment. It should be noted that the technical scheme of the phenotype prediction system and the technical scheme of the phenotype prediction method described above belong to the same concept, and the details of the technical scheme of the phenotype prediction system that are not described in detail can all be referred to the description of the technical scheme of the phenotype prediction method described above.
[0216] Corresponding to the above method embodiment, this specification also provides an embodiment of a device for processing genetic variation sites, the device comprising:
[0217] A data determination module is configured to determine a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites;
[0218] an object acquisition module, configured to determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0219] The object processing module is configured to input the site object into a site processing model to obtain processing results corresponding to the multiple genetic variation sites.
[0220] Optionally, the object processing module is further configured to:
[0221] Inputting the site object into the site processing model, wherein the site processing model includes a feature extraction unit and a feature processing unit;
[0222] Using the feature extraction unit to extract features from the site object to obtain site object features corresponding to the site object;
[0223] The feature processing unit is used to process the site object features and the site object to obtain processing results corresponding to the multiple genetic variation sites.
[0224] Optionally, there are at least two site objects, each of which contains some genetic variation sites, and the some genetic variation sites are subsets of the multiple genetic variation sites;
[0225] The object processing module is further configured to:
[0226] The feature extraction unit is used to extract features from each of the site objects to obtain site object features corresponding to each of the site objects.
[0227] Optionally, the object processing module is further configured to:
[0228] Sorting the site object features corresponding to each site object to determine a site object feature sequence corresponding to each site object;
[0229] Sort the site objects to obtain a site object sequence;
[0230] The site object feature sequence and the site object sequence are input into the feature processing unit for processing to obtain processing results corresponding to the multiple genetic variation sites.
[0231] Optionally, the object acquisition module is further configured to:
[0232] Determine the similarity between a target genetic variation site and other genetic variation sites in a site sequence, and determine the similarity as the association relationship between the genetic variation sites, wherein the target genetic variation site is any one of the multiple genetic variation sites, and the other genetic variation sites are genetic variation sites among the multiple genetic variation sites except the target genetic variation site.
[0233] Optionally, there are at least two site sequences, and each site sequence includes a plurality of genetic variation sites;
[0234] The object acquisition module is further configured to:
[0235] Determine a target genetic variation site and other genetic variation sites in at least two site sequences, wherein the target genetic variation site is any one genetic variation site in the at least two site sequences, and the other genetic variation sites are other genetic variation sites in the at least two site sequences except the target genetic variation site;
[0236] The similarity between the target genetic variation site and the other genetic variation sites is determined.
[0237] Optionally, the object acquisition module is further configured to:
[0238] Determine the at least two site sequences as a site matrix, and determine a transposed site matrix corresponding to the site matrix;
[0239] The site matrix and the transposed site matrix are multiplied to obtain a product result, and the similarity between the target genetic variation site and the other genetic variation sites is determined based on the product result.
[0240] Optionally, the object acquisition module is further configured to:
[0241] Determining distance parameters between the genetic variation sites based on the association relationship between the genetic variation sites, and mapping the multiple genetic variation sites included in the site sequence to a two-dimensional space based on the distance parameters;
[0242] Clustering the multiple genetic variation sites in the two-dimensional space to obtain at least one site set, wherein each site set contains associated genetic variation sites;
[0243] The genetic variation sites included in each site set are subjected to data conversion to obtain at least one site object.
[0244] Optionally, the site set is one;
[0245] The object acquisition module is further configured to:
[0246] Determining the genetic variation sites included in the site set, and performing image data conversion on the genetic variation sites to obtain site image data;
[0247] The set of sites is at least two;
[0248] The object acquisition module is further configured to:
[0249] The genetic variation sites included in each of the at least two site sets are converted into image data to obtain at least two site image data, wherein the at least two site image data correspond to the at least two site sets.
[0250] Optionally, the data determination module is further configured to:
[0251] receiving the site sequence of the biological sample to be processed sent by a client, wherein the site sequence is sent by the client when the user performs a site sequence upload operation;
[0252] The genetic variation site processing device further includes a result sending module configured to:
[0253] The processing result is sent to the client for display.
[0254] One or more embodiments of the present specification provide a device for processing genetic variation sites. In the process of processing the genetic variation sites, the device can determine the correlation relationship between the genetic variation sites, and map the site sequence to a two-dimensional space based on the correlation relationship to obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the correlation relationship between the genetic variation sites, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the correlation relationship into a site processing model for processing, the site processing model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate processing results, avoiding the problem of low accuracy of the processing results output by the neural network model, and ensuring the smooth progress of biological genetic research.
[0255] The above is a schematic scheme of a processing device for a genetic variation site of this embodiment. It should be noted that the technical scheme of the processing device for a genetic variation site and the technical scheme of the processing method for a genetic variation site described above belong to the same concept, and the details of the technical scheme of the processing device for a genetic variation site that are not described in detail can all be referred to the description of the technical scheme of the processing method for a genetic variation site described above.
[0256] Corresponding to the above method embodiment, this specification also provides a phenotype prediction device embodiment, the device comprising:
[0257] A data determination module is configured to determine a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites;
[0258] an object acquisition module, configured to determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0259] The object processing module is configured to input the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
[0260] One or more embodiments of the present specification provide a phenotype prediction device. In the process of phenotype prediction based on genetic variation sites, the device can determine the correlation relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the correlation relationship to obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the correlation relationship between each genetic variation site, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the correlation relationship into a phenotype prediction model for processing, the phenotype prediction model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate phenotype prediction results, avoiding the problem of low accuracy of the phenotype prediction results output by the neural network model, and ensuring the smooth progress of biological genetic research.
[0261] The above is a schematic scheme of a phenotype prediction device of this embodiment. It should be noted that the technical scheme of the phenotype prediction device and the technical scheme of the above-mentioned phenotype prediction method belong to the same concept, and the details not described in detail in the technical scheme of the phenotype prediction device can be referred to the description of the technical scheme of the above-mentioned phenotype prediction method.
[0262] Corresponding to the above method embodiment, this specification also provides another phenotype prediction device embodiment, which is applied to a cloud-side device, including:
[0263] A data receiving module, configured to receive a site sequence of a biological sample to be processed sent by a terminal device, wherein the site sequence includes a plurality of genetic variation sites;
[0264] an object acquisition module, configured to determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites;
[0265] An object processing module is configured to input the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites;
[0266] The result sending module is configured to send the phenotype prediction result to the terminal device.
[0267] One or more embodiments of the present specification provide a phenotype prediction device applied to a cloud-side device. During the process of processing genetic variation sites, the device can determine the correlation between the genetic variation sites provided by the terminal-side device, and map the site sequence to a two-dimensional space based on the correlation, and obtain a site object corresponding to the two-dimensional space, thereby adjusting the site sequence based on the correlation between the genetic variation sites, thereby distributing the associated genetic variation sites in similar positions, and by inputting the site objects determined based on the correlation into a phenotype prediction model for processing, the phenotype prediction model effectively establishes the connection between the genetic variation sites and performs analysis and processing, thereby obtaining accurate phenotype prediction results, avoiding the problem of low accuracy of the phenotype prediction results output by the neural network model, and ensuring the smooth progress of biological genetic research.
[0268] The above is a schematic scheme of another phenotype prediction device of this embodiment. It should be noted that the technical scheme of the another phenotype prediction device and the technical scheme of the another phenotype prediction method mentioned above belong to the same concept, and the details of the technical scheme of the another phenotype prediction device that are not described in detail can all be referred to the description of the technical scheme of the another phenotype prediction method mentioned above.
[0269] Figure 7 The block diagram of a computing device 700 according to an embodiment of the present specification is shown. The components of the computing device 700 include but are not limited to a memory 710 and a processor 720. The processor 720 is connected to the memory 710 via a bus 730, and the database 750 is used to store data.
[0270] The computing device 700 also includes an access device 740 that enables the computing device 700 to communicate via one or more networks 760. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 740 may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world-wide interoperability for microwave access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, and a near field communication (NFC).
[0271] In one embodiment of the present specification, the above components of the computing device 700 and Figure 7 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Figure 7 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.
[0272] The computing device 700 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 700 may also be a mobile or stationary server.
[0273] The processor 720 is used to execute the following computer executable instructions, which implement the steps of any of the above methods when executed by the processor.
[0274] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computing device embodiment, since it is basically similar to any method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of any method embodiment.
[0275] An embodiment of the present specification further provides a computer-readable storage medium storing a computer program / instruction, which implements the steps of any of the above methods when executed by a processor.
[0276] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer-readable storage medium embodiment, since it is basically similar to any method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of any method embodiment.
[0277] An embodiment of the present specification also provides a computer program product, including a computer program / instruction, which implements the steps of any one of the above methods when executed by a processor.
[0278] The above is a schematic solution of a computer program product of this embodiment. It should be noted that the technical solution of the computer program product and the technical solution of any of the above methods belong to the same concept, and the details not described in detail in the technical solution of the computer program product can be referred to the description of the technical solution of any of the above methods.
[0279] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0280] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.
[0281] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of this specification are not limited by the order of the actions described, because according to the embodiments of this specification, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of this specification.
[0282] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0283] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The optional embodiments do not describe all the details in detail, nor do they limit the invention to only the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of this specification, so that technicians in the relevant technical field can well understand and use this specification. This specification is only limited by the claims and their full scope and equivalents.
Claims
1. A method for processing genetic variation sites, comprising: Determining a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites; Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites, the site object includes a site image, the site image is obtained by performing image data conversion on each genetic variation site in the two-dimensional space, and the association relationship is obtained by determining multiple site sequences into a site matrix; The site object is input into a site processing model to obtain processing results corresponding to the multiple genetic variation sites.
2. The method of claim 1, inputting the site object into a site processing model to obtain processing results corresponding to the plurality of genetic variation sites, comprising: Inputting the site object into the site processing model, wherein the site processing model includes a feature extraction unit and a feature processing unit; Using the feature extraction unit to extract features from the site object to obtain site object features corresponding to the site object; The feature processing unit is used to process the site object features and the site object to obtain processing results corresponding to the multiple genetic variation sites.
3. The method according to claim 2, wherein the number of the site objects is at least two, each of the site objects contains a portion of genetic variation sites, and the portion of genetic variation sites is a subset of the plurality of genetic variation sites; Utilizing the feature extraction unit to extract features from the site object to obtain site object features corresponding to the site object includes: The feature extraction unit is used to extract features from each of the site objects to obtain site object features corresponding to each of the site objects.
4. The method according to claim 3, wherein the step of processing the site object features and the site object using the feature processing unit to obtain processing results corresponding to the plurality of genetic variation sites comprises: Sorting the site object features corresponding to each site object to determine a site object feature sequence corresponding to each site object; Sort the site objects to obtain a site object sequence; The site object feature sequence and the site object sequence are input into the feature processing unit for processing to obtain processing results corresponding to the multiple genetic variation sites.
5. The method of claim 1, wherein determining the association between each genetic variation site comprises: Determine the similarity between a target genetic variation site and other genetic variation sites in a site sequence, and determine the similarity as the association relationship between the genetic variation sites, wherein the target genetic variation site is any one of the multiple genetic variation sites, and the other genetic variation sites are genetic variation sites among the multiple genetic variation sites except the target genetic variation site.
6. The method according to claim 5, wherein the number of the site sequences is at least two, and each site sequence comprises a plurality of genetic variation sites; Determine the similarity between the target genetic variation site and other genetic variation sites in the site sequence, including: Determine a target genetic variation site and other genetic variation sites in at least two site sequences, wherein the target genetic variation site is any one genetic variation site in the at least two site sequences, and the other genetic variation sites are other genetic variation sites in the at least two site sequences except the target genetic variation site; The similarity between the target genetic variation site and the other genetic variation sites is determined.
7. The method of claim 6, wherein determining the similarity between the target genetic variation site and the other genetic variation sites comprises: Determine the at least two site sequences as a site matrix, and determine a transposed site matrix corresponding to the site matrix; The site matrix and the transposed site matrix are multiplied to obtain a product result, and the similarity between the target genetic variation site and the other genetic variation sites is determined based on the product result.
8. The method according to claim 1, mapping the site sequence to a two-dimensional space based on the association relationship, and obtaining the site object corresponding to the two-dimensional space, comprising: Determining distance parameters between the genetic variation sites based on the association relationship between the genetic variation sites, and mapping the multiple genetic variation sites included in the site sequence to a two-dimensional space based on the distance parameters; Clustering the multiple genetic variation sites in the two-dimensional space to obtain at least one site set, wherein each site set contains associated genetic variation sites; The genetic variation sites included in each site set are subjected to data conversion to obtain at least one site object.
9. The method of claim 8, wherein the set of sites is one; Performing data conversion on the genetic variation sites included in each site set to obtain at least one site object includes: Determining the genetic variation sites included in the site set, and performing image data conversion on the genetic variation sites to obtain site image data; The set of sites is at least two; Performing data conversion on the genetic variation sites included in each site set to obtain at least one site object includes: The genetic variation sites included in each of the at least two site sets are converted into image data to obtain at least two site image data, wherein the at least two site image data correspond to the at least two site sets.
10. The method of claim 1, wherein determining the site sequence of the biological sample to be processed comprises: receiving the site sequence of the biological sample to be processed sent by a client, wherein the site sequence is sent by the client when a user performs a site sequence upload operation; After inputting the site object into the site processing model to obtain the processing results corresponding to the plurality of genetic variation sites, the method further includes: The processing result is sent to the client for display.
11. A phenotype prediction method, comprising: Determining a site sequence of a biological sample to be processed, wherein the site sequence includes a plurality of genetic variation sites; Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites, the site object includes a site image, the site image is obtained by performing image data conversion on each genetic variation site in the two-dimensional space, and the association relationship is obtained by determining multiple site sequences into a site matrix; The site object is input into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
12. A phenotype prediction method, applied to a cloud-side device, comprising: A site sequence of a biological sample to be processed sent by a receiving device, wherein the site sequence includes a plurality of genetic variation sites; Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites, the site object includes a site image, the site image is obtained by performing image data conversion on each genetic variation site in the two-dimensional space, and the association relationship is obtained by determining multiple site sequences into a site matrix; Inputting the site object into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites; The phenotype prediction result is sent to the end-side device.
13. A phenotype prediction system, comprising a client and a server, wherein: The client is configured to send a site sequence of a biological sample to be processed to the server, wherein the site sequence includes a plurality of genetic variation sites; The server is configured to receive the site sequence of the biological sample to be processed sent by the client, Determine the association relationship between each genetic variation site, and map the site sequence to a two-dimensional space based on the association relationship to obtain a site object corresponding to the two-dimensional space, wherein the dimension of the site object is greater than the dimension of the site sequence, and the site object includes multiple genetic variation sites, the site object includes a site image, and the site image is obtained by performing image data conversion on each genetic variation site in the two-dimensional space, and the association relationship is obtained by determining multiple site sequences into a site matrix, The site object is input into a phenotype prediction model to obtain phenotype prediction results corresponding to the multiple genetic variation sites.
14. A computing device comprising: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, the steps of the method described in any one of claims 1 to 12 are implemented.
15. A computer-readable storage medium storing a computer program / instruction, wherein the computer program / instruction, when executed by a processor, implements the steps of the method according to any one of claims 1 to 12.
16. A computer program product, comprising a computer program / instruction, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 12.
