DNA database retrieval method based on machine learning algorithm

Abstract

The invention relates to the technical field of database information analysis, in particular to a DNA database retrieval method based on a machine learning algorithm. Collecting map data set information, and calculating an LR parameter theta; respectively simulating individuals with a parent-child relationship, individuals with a full-sib relationship and irrelevant individuals of each known contributor; comparing each mixed DNA map with the type of each candidate individual to obtain a characteristic value; obtaining a training data set and a test data set of the features and the labels; training the regression model by using the training data set, and performing hyper-parameter optimization and feature selection in the training process; taking the test data set as input of all the trained models, and obtaining predicted values of all the models; obtaining a mixed DNA map from a site, sequentially traversing each candidate individual in the DNA database to calculate a characteristic value corresponding to each candidate individual, and performing LR prediction to obtain a target individual; in this way, the requirement for efficient and accurate screening is met.

Description

Technical Field

[0001] The present invention relates to the technical field of database information analysis, and in particular to a DNA database retrieval method based on a machine learning algorithm. Background Art

[0002] At present, my country's DNA database mainly uses STR (short tandem repeats) as genetic markers, which is an important tool for modern criminal investigation science. It aims to compare the DNA profile of biological samples obtained at the crime scene with the STR typing of the profile in the database, find the target individual, and provide clues for solving the case. At present, my country's forensic DNA database has a three-level network architecture of ministry-province-city, which is large in scale and provides strong data support for criminal investigation. With the continuous advancement of DNA testing technology and information technology, the amount of data in the DNA database continues to grow, which poses a challenge to retrieval efficiency. Efficient and accurate retrieval schemes have become a key issue to be urgently solved in the current application and development of DNA databases. The DNA profile at the crime scene may be a complex mixed profile, containing the STR typing of multiple contributors. Its retrieval strategy is to regard each individual in the DNA database as a candidate for potential contributor, and set a set of mutually exclusive hypotheses. By calculating the probability ratio under the two hypotheses, the likelihood ratio (LR), the association between each candidate individual and the mixed profile is evaluated. Among them, the plaintiff assumes that the candidate individual is a contributor to the mixed graph, and the defendant assumes that the candidate individual is not a contributor to the mixed graph. When LR is greater than 1, the plaintiff's hypothesis is supported, indicating that the candidate individual may be the target individual.

[0003] In view of the scientificity and practicality of this strategy, in the existing search, all candidate individuals in the database can be traversed, and the LR value of each individual in the mixed map can be calculated to exclude irrelevant individuals and narrow the target screening range, thereby improving the detection efficiency. To achieve this goal, a variety of optimization methods can be used to increase the search speed, such as streamlining the number of parameters required for LR calculation, limiting the possible combination range of genotypes of contributors to the mixed map, and using a hierarchical screening mechanism to gradually eliminate unlikely candidates, thereby accelerating the search process. Although the above method has shown significant advantages in improving detection efficiency and accuracy, when faced with a huge database of millions or even tens of millions of records, its computing speed and the accuracy of avoiding the erroneous inclusion of irrelevant individuals in the target range are still facing challenges, and it is difficult to meet the urgent needs of front-line detection work for efficient and accurate screening.

[0004] In summary, it is necessary to propose a DNA database retrieval method that meets the needs of efficient and accurate screening. Summary of the invention

[0005] The purpose of the present invention is to provide a DNA database retrieval method based on a machine learning algorithm, aiming to solve the technical problem that when faced with a huge database of millions or even tens of millions of records, the computing speed and the accuracy of avoiding erroneous individuals from being included in the target range are still challenged, and it is difficult to meet the urgent needs of front-line investigation work for efficient and accurate screening.

[0006] To achieve the above object, the present invention adopts a DNA database retrieval method based on a machine learning algorithm, comprising the following steps:

[0007] Collecting large-scale STR-DNA profile data set information; wherein the data set includes single-source profiles and mixed DNA profiles, each profile includes allele data and peak height data, and the mixed DNA profile is composed of DNA from at least two contributors;

[0008] The single-source atlas in the dataset is used to calculate the LR prior parameter θ;

[0009] For each mixed DNA profile in the data set, simulate the individuals with parent-child relationship, individuals with full sibling relationship, and unrelated individuals of each known contributor;

[0010] For each mixed DNA profile in the data set, the mixed DNA profile consists of K known contributors, traverse each candidate individual, set a mutually exclusive hypothesis proposition H p , H d ; Calculate the probability ratio LR under the two hypotheses, and convert Log 10 (LR) as the label value;

[0011] Among them, H p The mixed DNA map consists of candidate individuals and K-1 unknown unrelated individuals; H d The mixed DNA profile consists of K unknown unrelated individuals;

[0012] Compare each mixed DNA profile with the typing of each candidate individual to obtain characteristic values, and perform normalization on the characteristic values;

[0013] Obtain a dataset of features and labels, and divide the dataset into a training dataset and a test dataset;

[0014] Use the regression model, train the regression model using the training data set, and perform hyperparameter optimization and feature selection during the training process;

[0015] Use the test data set as the input of all trained models, obtain the predicted values ​​of all models, and judge the target model;

[0016] Obtain the mixed DNA map from the scene, traverse each candidate individual in the DNA database in turn, calculate the corresponding feature value of each candidate individual, and perform Log 10 (LR) prediction, get the target individual.

[0017] Among them, for each mixed DNA profile in the data set, in the step of simulating the individuals with parent-child relationship, individuals with full sibling relationship, and unrelated individuals of each known contributor, the definition is as follows:

[0018] The candidate individuals include: K known contributors, K individuals with parent-child relationships with known contributors, K individuals with full sibling relationships with known contributors, and K unrelated individuals;

[0019] The known contributor is typed as g, and the genotype at site m is g m ={a1,a2}, where a1 and a2 represent the two alleles at site m respectively.

[0020] Among them, for each mixed DNA profile in the data set, in the step of simulating the individuals with parent-child relationship, individuals with full sibling relationship, and unrelated individuals of each known contributor, the simulation process is as follows:

[0021] Randomly select an allele a1 at the site, and randomly select an allele a3 at the site from the population frequency data. Then the simulated individual with a parent-child relationship with the contributor at site m is typed as g m_PO ={a1,a3};

[0022] Get the simulated parent of this contributor m_F ={a1,a3}, randomly select an allele a4 at the site from the population frequency data, then the simulated maternal g of the contributor m_M ={a2,a4}; randomly select one allele from each of the simulated father and simulated mother typing to form the typing g of the simulated individual with a full sibling relationship with the contributor at site m m_FS ;

[0023] Randomly select two alleles at this site from the population frequency data to form the typing g of unrelated individuals at site m. m_UN ;

[0024] Traverse all sites and obtain the simulated individual typing g with parent-child relationship with the contributor PO , the simulated individual typing g with full sibling relationship FS , irrelevant to simulated individual typing g UN .

[0025] Among them, for each mixed DNA profile in the data set, the mixed DNA profile consists of K known contributors, traverse each candidate individual, set the mutually exclusive hypothesis proposition H p , H d In the steps:

[0026] Calculate the probability ratio LR under the two hypotheses and log 10 (LR) is used as the label value, and the corresponding prior parameter θ is used. The probability ratio LR formula is:

[0027]

[0028] Among them, in the step of comparing each mixed DNA profile with the typing of each candidate individual, obtaining characteristic values, and normalizing the characteristic values:

[0029] The characteristic values ​​include: total number of mixed map alleles, sum of mixed map allele peak heights, total number of matching alleles, proportion of alleles not lost, maximum matching allele peak height, minimum matching allele peak height, proportion of matching allele peak height, matching allele p-value, frequency product, peak height expectation, peak height coefficient of variation, and peak height degradation parameter.

[0030] Among them, in the step of using the test data set as the input of all trained models to obtain the predicted values ​​of all models:

[0031] Compare the predicted value with the true value, judge the fitting effect, and obtain the target model based on the fitting effect.

[0032] Among them, after obtaining the mixed DNA map from the scene, each candidate individual in the DNA database is traversed in turn to calculate the corresponding feature value of each candidate individual, and then log 10 (LR) prediction, the steps to obtain the target individual:

[0033] When predicting Log 10 If the (LR) value is > 0, the corresponding candidate individual is included in the target individual range;

[0034] When predicting Log 10 If the (LR) value is <0, the corresponding candidate individual is excluded from the target individual range.

[0035] The invention discloses a DNA database retrieval method based on a machine learning algorithm, which collects large-scale STR-DNA profile data set information; wherein the data set includes a single-source profile and a mixed DNA profile, each profile includes allele data and peak height data, and the mixed DNA profile is composed of DNA from at least two contributors; the single-source profile in the data set is used to calculate the LR prior parameter θ; for each mixed DNA profile in the data set, the individuals with parent-child relationship, individuals with full sibling relationship, and unrelated individuals of each known contributor are simulated respectively; for each mixed DNA profile in the data set, the mixed DNA profile is composed of K known contributors, each candidate individual is traversed, and mutually exclusive hypothesis propositions H are set p , H d ; where H p The mixed DNA map consists of candidate individuals and K-1 unknown unrelated individuals; H d The mixed DNA map is composed of K unknown unrelated individuals; each mixed DNA map is compared with the typing of each candidate individual, the characteristic value is obtained, and the characteristic value is normalized; a data set of features and labels is obtained, the data set is divided into a training data set and a test data set, and the characteristics of the data set are normalized; a regression model is used to train the regression model using the training data set, and hyper-parameter optimization and feature selection are performed during the training process; the test data set is used as the input of all trained models, the predicted values ​​of all models are obtained, and the target model is judged; a mixed DNA map from the scene is obtained, each candidate individual in the DNA database is traversed in turn, the characteristic value corresponding to each candidate individual is calculated, and Log 10 (LR) prediction to obtain the target individual; to meet the needs of efficient and accurate screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 It is a flow chart of the steps of the DNA database retrieval method based on the machine learning algorithm of the present invention.

[0038] Figure 2 It is a schematic diagram of some features and label data sets of the present invention. DETAILED DESCRIPTION

[0039] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0040] See also Figure 1 and Figure 2 The present invention provides a DNA database retrieval method based on a machine learning algorithm, comprising the following steps:

[0041] Step 1: Collect large-scale STR-DNA profile data set information; wherein the data set includes single-source profiles and mixed DNA profiles, each profile includes allele data and peak height data, and the mixed DNA profile is composed of DNA from at least two contributors.

[0042] In this embodiment, large-scale STR-DNA profile data set information is collected; wherein the data set includes single-source profiles and mixed DNA profiles, each profile includes allele data and peak height data, the mixed DNA profile is composed of DNA from at least two contributors, and the experimental conditions for generating the mixed DNA profile are: using different DNA template amounts, applying different STR kits, performing different PCR reaction conditions, and adopting different electrophoresis platforms and electrophoresis injection conditions; the DNA templates include: original DNA templates without any treatment, DNA templates that have undergone degradation, and DNA templates with PCR inhibitors.

[0043] For example, the STR-DNA profile online dataset PROVEDIT is used. This dataset contains single-source profiles and mixed DNA profiles. Each profile contains allele data and peak height data. And the mixed profile contains at least two contributors. The kits for generating mixed DNA profiles include Identifiler kit, PowerPlex16 kit, Fusion 6C kit, and GlobalFiler kit, and their experimental conditions are shown in Table 1. Its DNA templates cover a variety of types.

[0044] Table 1 Prior parameter θ values ​​of single source maps under different experimental conditions

[0045]

[0046]

[0047] Step 2: Use a single source map in the dataset to calculate the LR prior parameter θ.

[0048] In this embodiment, a single source profile in a data set is used to calculate the LR prior parameter θ, wherein the single source profiles generated by different kits, different cycle parameters, different electrophoresis platforms, and different injection times have different LR prior parameters θ.

[0049] For example, the LR related prior parameters θ are calculated using the single source map in PROVEDIT.

[0050] For example, θ={AT, pC, λ}, where AT represents the analysis threshold (AT), pC represents the probability of allele insertion, and λ represents the parameter of the exponential distribution to which the inserted allele obeys.

[0051] The calculation results are shown in Table 1, and the calculation method is as follows:

[0052] 1. For the single-source atlas generated with the same kit, the same cycle parameters, the same electrophoresis platform, and the same injection time, remove the OL peak, the true allele peak, and its corresponding -1stutter peak and +1stutter peak in the single-source atlas. All the remaining peaks are noise peaks, and the 95% quantile of the peak empirical cumulative distribution is AT.

[0053] 2.

[0054] Where n represents the number of noise peaks, N represents the number of single-source maps, and L represents the number of alleles in each single-source map;

[0055]

[0056] x i represents the peak height of the i-th noise peak.

[0057] Step 3: For each mixed DNA profile in the data set, simulate the individuals with parent-child relationships, individuals with full sibling relationships, and unrelated individuals of each known contributor.

[0058] In this embodiment, the definitions are as follows:

[0059] The candidate individuals include: K known contributors, K individuals with parent-child relationships with known contributors, K individuals with full sibling relationships with known contributors, and K unrelated individuals; for each mixed graph E in the data set, the number of contributors participating in the mixed graph is K. When K=2, simulate one individual with parent-child relationship, one individual with full sibling relationship, and one unrelated individual for each known contributor. Then the candidate individuals include: 2 known contributors, 2 individuals with parent-child relationship with known contributors, 2 individuals with full sibling relationship with known contributors, and 2 unrelated individuals, a total of 8 individuals. When K=3, there are 12 candidate individuals in total. And so on.

[0060] The known contributor is typed as g, and the genotype at site m is g m ={a1,a2}, a1 and a2 represent the two alleles at site m respectively; taking site D3S1358 as an example, the genotype of this site is g m ={13,15}

[0061] The simulation process is:

[0062] Randomly select an allele 13 at the site, and randomly select an allele 16 at the site from the population frequency data. Then the simulated individual with a parent-child relationship with the contributor at the site D3S1358 is typed as g m_PO ={13,16};

[0063] Follow the above steps to get the simulated parent of the contributor m_F ={13,16}, randomly select an allele 18 at this site from the population frequency data, then the simulated maternal g of this contributor m_M ={15,18}; randomly select one allele from each of the simulated father and simulated mother typing to form the typing g of the simulated individual with a full sibling relationship with the contributor at site m m_FS ={15,16};

[0064] Randomly select two alleles at this site from the population frequency data to form the typing g of unrelated individuals at site m. m_UN ={19, 20};

[0065] Traverse all sites and obtain the simulated individual typing g with parent-child relationship with the contributor PO , the simulated individual typing g with full sibling relationship FS , irrelevant to simulated individual typing g UN .

[0066] Step 4: For each mixed DNA profile in the data set, the mixed DNA profile consists of K known contributors, traverse each candidate individual, and set a mutually exclusive hypothesis proposition H p , H d ; where H p The mixed DNA map consists of candidate individuals and K-1 unknown unrelated individuals; H d The mixed DNA profile consists of K unknown unrelated individuals.

[0067] In this embodiment, for each mixed graph E, the mixed graph E consists of K known contributors, traverse each candidate individual, set a mutually exclusive hypothesis proposition H p , H d ; When the number of contributors participating in the mixed graph is K = 2, H p The mixed graph consists of candidate individuals and one unknown unrelated individual; H d The mixed graph consists of two unknown unrelated individuals; when K = 3, H p The mixed graph consists of a candidate individual and two unknown unrelated individuals; H d The mixed graph consists of three unknown unrelated individuals; Similarly, the probability ratio LR under the two hypotheses is calculated, and Log 10 (LR) is used as the label value, and the probability ratio LR formula is:

[0068]

[0069] Step 5: Compare each mixed DNA profile with the typing of each candidate individual, obtain the characteristic values, and normalize the characteristic values.

[0070] In this embodiment, for each mixed pattern E in the data set, the following characteristic values ​​are obtained by comparing with the typing of each candidate individual: the total number of mixed pattern alleles, the sum of mixed pattern allele peak heights, the total number of matching alleles, the proportion of alleles not lost, the maximum peak height of matching alleles, the minimum peak height of matching alleles, the peak height proportion of matching alleles, the p-value of matching alleles, the frequency product, the expected peak height, the coefficient of variation of the peak height, and the peak height degradation parameter; and the characteristic values ​​are normalized;

[0071] The characteristic data definitions are as follows:

[0072] Total number of alleles in a hybrid map: The sum of the numbers of alleles at all sites in a hybrid map.

[0073] Sum of allele peak heights in a mixed map: The sum of the peak heights of alleles at all sites in a mixed map.

[0074] Matching alleles: Alleles that are consistent at site m in the mixed map and the candidate individual typing.

[0075] Total number of matching alleles: the sum of the number of matching alleles at all sites.

[0076] Proportion of alleles not missing: the total number of matching alleles divided by the total number of alleles in the candidate individual.

[0077] Maximum peak height of matching alleles: the maximum value of the peak height corresponding to the allele among the matching alleles at all sites.

[0078] Minimum peak height of matched alleles: the minimum peak height corresponding to the allele among all the matched alleles at all sites.

[0079] Matching allele peak height ratio: the sum of matching allele peak heights at all sites divided by the sum of mixed map allele peak heights.

[0080] Matching allele p-value: The negative value of the probability of observing a better result than the matching allele number at all sites, the formula is:

[0081]

[0082] where c m represents the actual number of matching alleles at site m, C m It represents the theoretical number of matching alleles observed at site m, and can be 0, 1, or 2. p(C m ≥c m ) means C m Greater than c m The probability of M indicates that there are M sites in the mixed DNA map.

[0083] Frequency product: The negative value of the product of the frequency of the matched allele and the frequency of the unmatched allele at all loci. The formula is:

[0084]

[0085] where p m represents the frequency of the first matching allele at site m, q m represents the frequency of the second matching allele at site m; if there is no matching allele, the result for this site is 1.

[0086] Peak height expectation, peak height variation coefficient, peak height degradation parameter: all are parameter values ​​of the gamma model. The gamma model is used to fit the mixed spectrum peak height data, and the value is obtained after maximum likelihood estimation. The gamma model formula is as follows:

[0087]

[0088] Where Y represents the mean peak height of the mixed DNA profile, μ, ω, and σ represent the expected peak height, the coefficient of variation of the peak height, and the peak height degradation parameter, respectively; It means that under the given conditions of μ, ω, σ, the peak height of the mixed DNA spectrum is Probability; It represents the mean peak height of alleles at site m on the mixed DNA map; It represents the mean molecular weight of the allele at site m on the mixed DNA map; M means there are M sites in total on the mixed DNA map.

[0089] Step 6: Obtain a dataset of features and labels, and divide the dataset into a training dataset and a test dataset.

[0090] In this embodiment, a feature and label data set is obtained, and the data set is divided into 80% training data set and 20% test data set; some data sets are as follows Figure 2 The characteristic values ​​in each group of features are normalized by Z-Score, and the formula is z=(x i -μ) / σ, z represents the standardized eigenvalue, x i represents the i-th eigenvalue of the group, μ represents the mean of the eigenvalues ​​of the group, and σ represents the variance of the eigenvalues ​​of the group.

[0091] Step 7: Use the regression model to train the regression model using the training data set, and perform hyperparameter optimization and feature selection during the training process.

[0092] In this embodiment, ridge regression, Lasso regression, K-nearest neighbor regression, random forest regression, support vector regression, elastic network regression and other models in machine learning are used, and the above regression models are trained respectively using the aforementioned training data sets, and the training process adopts the grid search method to find the optimal hyperparameter combination. Recursive feature elimination is used to select features.

[0093] Step 8: Use the test data set as the input of all trained models, obtain the predicted values ​​of all models, compare the predicted values ​​with the true values, judge the fitting effect, and obtain the target model based on the fitting effect.

[0094] In this embodiment, the test data set is used as the input of all trained models to obtain the predicted values ​​of all models. The predicted values ​​are compared with the true values, and the model with the highest fitting effect and the highest prediction accuracy is determined based on the comparison results. The model with the highest fitting effect and the highest prediction accuracy is selected as the target model. The model fitting effect is evaluated by the determination coefficient. R 2The larger the MSE, the better the model fitting effect. The prediction accuracy is evaluated by the mean square error. The smaller the MSE, the higher the model prediction accuracy. 2 、MSE calculation formula is as follows:

[0095]

[0096]

[0097] Among them, y i represents the true value, represents the predicted value, represents the mean of the true value, and m is the number of data in the prediction set. The R2 and MSE of each model in the training set and test set are shown in Table 2. 2 The maximum value and the minimum MSE value are used to analyze the log 10 After feature selection, the model uses the proportion of alleles not missing, the sum of peak heights of alleles in the mixed map, the proportion of peak heights of matching alleles, the p value of matching alleles, the total number of alleles in the mixed map, and the coefficient of variation of peak height as features.

[0098] Table 2 R of six models in training and test sets 2 and MSE

[0099] Model <![CDATA[Training set R 2 > <![CDATA[Test set R 2 > MSE Ridge Regression 0.68 0.71 18.05 Lasso Return 0.68 0.71 18.04 K-nearest neighbor regression 0.43 0.19 49.51 Random Forest Regression 0.98 0.85 6.32 Support Vector Regression 0.88 0.84 16.53 Elastic Net Regression 0.3 0.6 24.57

[0100] Step 9: Obtain the mixed DNA map from the scene, traverse each candidate individual in the DNA database in turn, calculate the corresponding feature value of each candidate individual, and perform Log 10 (LR) prediction, get the target individual.

[0101] In this embodiment, the simulated mixed map E' is a real case, consisting of two contributors. Using Caucasian population data, two alleles are randomly extracted from the population data at each site to simulate unrelated individuals. Four DNA databases are simulated, each containing 10,000, 100,000, 1,000,000, and 5,000,000 unrelated candidate individuals. The eigenvalues ​​corresponding to the two real contributors and the eigenvalues ​​corresponding to each candidate individual in each database are calculated, and the model in step eight is used to perform Log 10 (LR) prediction.

[0102] Prediction Log 10 (LR) value>0, the corresponding candidate individual is included in the target individual range, and the predicted Log 10If the (LR) value is less than 0, the corresponding candidate individual is excluded from the target individual range. Finally, N target individuals are obtained from the simulated database. The number of target individuals and the calculation speed of each simulated database are shown in Table 2. 10 The (LR) predicted values ​​are 39.82 and 12.43, both of which are within the target individual range. The corresponding true values ​​are 41.52 and 10.44, which are slightly different from the predicted values. The CPU used in the calculation process is i7-13700k.

[0103] Table 3. Number of target individuals screened by simulation database and calculation speed

[0104]

[0105] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the contents disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.

[0106] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. A DNA database retrieval method based on a machine learning algorithm, characterized in that: The steps include: Collecting large-scale STR-DNA profile data set information; wherein the data set includes single-source profiles and mixed DNA profiles, each profile includes allele data and peak height data, and the mixed DNA profile is composed of DNA from at least two contributors; The single-source atlas in the dataset is used to calculate the LR prior parameter θ; For each mixed DNA profile in the data set, simulate the individuals with parent-child relationship, individuals with full sibling relationship, and unrelated individuals of each known contributor; For each mixed DNA profile in the data set, the mixed DNA profile consists of K known contributors, traverse each candidate individual, set a mutually exclusive hypothesis proposition H p , H d ; Calculate the probability ratio LR under the two hypotheses, and convert Log 10 (LR) as the label value; where H p The mixed DNA map consists of candidate individuals and K-1 unknown unrelated individuals; H d The mixed DNA profile consists of K unknown unrelated individuals; Compare each mixed DNA profile with the typing of each candidate individual to obtain characteristic values, and perform normalization on the characteristic values; Obtain a dataset of features and labels, and divide the dataset into a training dataset and a test dataset; Use the regression model, train the regression model using the training data set, and perform hyperparameter optimization and feature selection during the training process; Use the test data set as the input of all trained models, obtain the predicted values ​​of all models, and judge the target model; Obtain the mixed DNA map from the scene, traverse each candidate individual in the DNA database in turn, calculate the corresponding feature value of each candidate individual, and perform Log 10 (LR) prediction, get the target individual.

2. A DNA database retrieval method based on a machine learning algorithm as claimed in claim 1, characterized in that: In the step of simulating the parent-child relationship, full-sibling relationship, and unrelated individuals of each known contributor for each mixed DNA profile in the data set, the following definitions are made: The candidate individuals include: K known contributors, K individuals with parent-child relationships with known contributors, K individuals with full sibling relationships with known contributors, and K unrelated individuals; The known contributor is typed as g, and the genotype at site m is g m ={a1,a2}, where a1 and a2 represent the two alleles at site m respectively.

3. A DNA database retrieval method based on a machine learning algorithm as claimed in claim 2, characterized in that: In the step of simulating the parent-child relationship, full-sibling relationship, and unrelated individuals of each known contributor for each mixed DNA profile in the data set, the simulation process is as follows: Randomly select an allele a1 at the site, and randomly select an allele a3 at the site from the population frequency data. Then the simulated individual with a parent-child relationship with the contributor at site m is typed as g m_PO ={a1,a3}; Get the simulated parent of this contributor m_F ={a1,a3}, randomly select an allele a4 at the site from the population frequency data, then the simulated maternal g of the contributor m_M ={a2,a4}; randomly select one allele from each of the simulated father and simulated mother typing to form the typing g of the simulated individual with a full sibling relationship with the contributor at site m m_FS ; Randomly select two alleles at this site from the population frequency data to form the typing g of unrelated individuals at site m. m_UN ; Traverse all sites and obtain the simulated individual typing g with parent-child relationship with the contributor PO , the simulated individual typing g with full sibling relationship FS , irrelevant simulated individual typing g UN .

4. A DNA database retrieval method based on a machine learning algorithm as claimed in claim 3, characterized in that: For each mixed DNA profile in the data set, the mixed DNA profile consists of K known contributors, traverse each candidate individual, set the mutually exclusive hypothesis proposition H p , H d In the steps: The probability ratio LR formula is:

5. A DNA database retrieval method based on a machine learning algorithm as claimed in claim 4, characterized in that: In the step of comparing each mixed DNA profile with the typing of each candidate individual, obtaining characteristic values, and normalizing the characteristic values: The characteristic values ​​include: total number of mixed map alleles, sum of mixed map allele peak heights, total number of matching alleles, proportion of alleles not lost, maximum matching allele peak height, minimum matching allele peak height, proportion of matching allele peak height, matching allele p-value, frequency product, peak height expectation, peak height coefficient of variation, and peak height degradation parameter.

6. A DNA database retrieval method based on a machine learning algorithm as claimed in claim 5, characterized in that: In the step of using the test dataset as input to all trained models and obtaining the predicted values ​​of all models: Compare the predicted value with the true value, judge the fitting effect, and obtain the target model based on the fitting effect.

7. A DNA database retrieval method based on a machine learning algorithm as claimed in claim 6, characterized in that: After obtaining the mixed DNA map from the scene, we traverse each candidate individual in the DNA database in turn, calculate the corresponding feature value of each candidate individual, and perform Log 10 (LR) prediction, the steps to obtain the target individual: When predicting Log 10 If the (LR) value is > 0, the corresponding candidate individual is included in the target individual range; When predicting Log 10 If the (LR) value is <0, the corresponding candidate individual is excluded from the target individual range.

Images