Enzyme kinetics parameter prediction method, and electronic device and storage medium
By reweighting and cross-validating the training samples, the accuracy of the enzyme kinetic parameter prediction model is improved, the problem of low prediction accuracy caused by uneven distribution of data categories is solved, and the prediction ability of high turnover enzymes is enhanced.
Patent Information
- Application Number
- PCT/CN2023/097104
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-30
- Publication Date
- 2024-12-05
AI Technical Summary
The existing enzyme kinetic parameter prediction model has a large error in prediction of high-value parts and a low prediction accuracy rate due to uneven distribution of data categories in the training samples.
By reweighting each type of sample in the training sample, the high turnover number enzyme data is given higher weight, and the accuracy of the prediction model is determined by using the cross-validation method, the model with the highest prediction accuracy is selected for the prediction parameters of enzyme kinetics.
The accuracy of the prediction of high turnover enzymes by the enzyme kinetic parameter prediction model is improved, and the model's learning ability for high-value samples is enhanced.
Smart Images

Figure CN2023097104_05122024_PF_FP_ABST
Abstract
Description
Enzyme kinetic parameter prediction method, electronic device and storage medium Technical Field
[0001] The present application relates to the field of biomedicine technology, and in particular to an enzyme kinetic parameter prediction method, electronic device, and storage medium. Background Art
[0002] In the biomedical field, enzymes are catalysts of cellular metabolism, and enzymes with high kinetic parameters play a key role in cellular metabolism. For example, enzyme kinetic parameters include enzyme turnover number. Enzymes with high turnover numbers exhibit efficient catalysis, accelerating substrate conversion rates and improving the efficiency of cellular metabolism. By studying the expression and activity of high-turnover enzymes, we can understand the regulatory mechanisms of enzymes under different physiological states, thereby further exploring the regulatory mechanisms of cellular metabolism and revealing the essential laws of cellular metabolism.
[0003] When training prediction models for predicting high turnover numbers of enzymes, related technologies often find it difficult to meet the demand for training data when training the prediction models due to the small number of actual high turnover number enzymes. This results in the prediction models being limited by the uneven category distribution of data in the training samples, resulting in large errors in the prediction of high-value parts and low prediction accuracy of the prediction models.
[0004] Summary of the Invention
[0005] In view of the above, it is necessary to provide an enzyme kinetic parameter prediction method, electronic device and storage medium that can solve the problem of low prediction accuracy of the prediction model caused by the uneven distribution of data categories in the training samples.
[0006] The enzyme kinetic parameter prediction method includes: obtaining training samples, using at least one reweighting method to reweight each type of sample in the training samples, wherein the training samples include target samples of the target category; using the reweighted training samples to train a prediction model for the enzyme kinetic parameters, and using a cross-validation method to determine the model prediction accuracy of the prediction model corresponding to each reweighting method; selecting the prediction model with the highest prediction accuracy as the target model, and using the target model to predict the enzyme kinetic parameters of the enzyme to be predicted.
[0007] In one embodiment, the training samples include a data set containing enzyme kinetic parameters, and the target samples include training samples in which the enzyme kinetic parameters are within a preset value range.
[0008] In one embodiment, the enzyme kinetic parameters include the enzyme turnover number k cat , Michaelis constant K m , the ratio of enzyme turnover number to Michaelis constant k cat / K m, where, when the enzyme kinetic parameter is k cat When, the training sample includes the DLKcat dataset.
[0009] In one embodiment, the DLKcat data set includes ternary data, including enzyme sequence-substrate structure-enzyme kinetic parameters, and the enzyme kinetic parameters include enzyme turnover number k cat .
[0010] In one embodiment, the at least one reweighting method includes: a direct modification of sample weights (DMW) method, a cost-sensitive reweighting (CSW) method, a class-balanced reweighting (CBW) method, and a label distribution smoothing (LDS) method.
[0011] In one embodiment, the cross-validation method includes a five-fold cross-validation method.
[0012] In one embodiment, the class-balanced reweighted CBW method includes: using a preset formula to calculate the effective sample number of the target sample, the formula includes: E n =(1-β n )(1-β),
[0013] Among them, E n represents the effective sample number of the target sample, n represents the number of the target samples, and β represents a hyperparameter with a preset value range of (0, 1); the weight of the target sample is updated to the inverse of the effective sample number.
[0014] In one embodiment, the method of using a cross-validation method to determine the model prediction accuracy of the prediction model corresponding to each reweighting method includes: using a cross-validation method to determine the prediction accuracy of the prediction model corresponding to each reweighting method for the enzyme kinetic parameters within the preset numerical range.
[0015] An embodiment of the present application provides an enzyme kinetic parameter prediction device, which includes: a reweighting module for obtaining training samples and reweighting each type of sample in the training samples using at least one reweighting method, wherein the training samples include target samples of the target category; a training module for training an enzyme kinetic parameter prediction model using the reweighted training samples, and using a cross-validation method to determine the model prediction accuracy of the prediction model corresponding to each reweighting method; a prediction module for selecting the prediction model with the highest prediction accuracy as the target model, and using the target model to predict the enzyme kinetic parameters of the enzyme to be predicted.
[0016] An embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the enzyme kinetic parameter prediction method or the enzyme kinetic parameter prediction method is implemented.
[0017] An embodiment of the present application provides an electronic device, comprising a memory and at least one processor, wherein the memory stores at least one instruction, and when the at least one instruction is executed by the at least one processor, the enzyme kinetic parameter prediction method is implemented.
[0018] Compared with the related art, the enzyme kinetic parameter prediction method provided in the embodiment of the present application reweights each type of sample in the training sample to give a higher weight to the high-turnover enzyme data as the target sample, so that the target sample receives more attention in the model training; the prediction model is trained using the reweighted training samples and the cross-validation algorithm to improve the prediction accuracy of the prediction model for high-turnover enzymes. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.
[0020] FIG1 is an architecture diagram of an electronic device provided in an embodiment of the present application.
[0021] FIG2 is a flow chart of an enzyme kinetic parameter prediction method provided in one embodiment of the present application.
[0022] FIG3 is a structural diagram of an enzyme kinetic parameter prediction device provided in one embodiment of the present application.
[0023] The following specific implementation methods will further illustrate the present application in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION
[0024] In order to more clearly understand the above-mentioned objectives, features and advantages of the present application, the present application is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features therein can be combined with each other in the absence of conflict.
[0025] The following description sets forth many specific details to facilitate a full understanding of the present application. The embodiments described are only a portion of the embodiments of the present application, not all of the embodiments. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present application without creative effort are intended to fall within the scope of protection of the present application.
[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application pertains. The terms used herein in the specification of this application are for the purpose of describing specific embodiments only and are not intended to limit this application.
[0027] In one embodiment, in the field of biomedical technology, enzymes are catalysts of cellular metabolism, and enzymes with high enzyme kinetic parameters play a key role in the process of cellular metabolism. For example, enzyme kinetic parameters include enzyme turnover number, and enzymes with high turnover number have efficient catalysis, which can accelerate the conversion rate of substrates and improve the efficiency of cellular metabolism. By studying the expression and activity of high turnover number enzymes, we can understand the regulatory mechanism of enzymes under different physiological states, thereby deeply exploring the regulatory mechanism of cellular metabolism and revealing the essential laws of cellular metabolism.
[0028] Furthermore, the research on high-turnover enzymes has broad application value in other technological fields. For example, in the food industry, enzymes can be used to improve the efficiency and quality of food processing; in the pharmaceutical industry, enzymes can be used to synthesize drugs or remove harmful chemicals; and in environmental protection, enzymes can be used to remove harmful substances from wastewater or exhaust gases. The research and prediction of high-turnover enzymes can provide strong technical support and support for these technical fields.
[0029] When training prediction models for predicting high turnover numbers of enzymes, related technologies often find it difficult to meet the demand for training data when training the prediction models due to the small number of actual high turnover number enzymes. This results in the prediction models being limited by the uneven distribution of data categories in the data set, resulting in large errors in the prediction of high-value parts and low prediction accuracy of the prediction models.
[0030] In order to solve the above problems, the enzyme kinetic parameter prediction method provided in the embodiments of the present application reweights each type of sample in the training samples to give a higher weight to the high-turnover enzyme data as the target sample, so that the target sample receives more attention in the model training; the prediction model is trained using the reweighted training samples and the cross-validation algorithm to improve the prediction accuracy of the prediction model for high-turnover enzymes.
[0031] For example, FIG1 is a block diagram of an electronic device provided in one embodiment of the present application. The enzyme kinetic parameter prediction method provided in this embodiment of the present application is performed by an electronic device, which can be a computer, server, laptop computer, mobile phone, or other device. The electronic device 1 includes a memory 11, at least one processor 12, at least one communication bus 13, and a transceiver 14.
[0032] The structure of the electronic device shown in FIG1 does not constitute a limitation of the embodiments of the present application, and can be either a bus structure or a star structure. The electronic device 1 can also include more or less other hardware or software than shown in the figure, or a different component arrangement.
[0033] In some embodiments, the electronic device 1 is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, a microprocessor, an application-specific integrated circuit, a programmable gate array, a digital processor, and an embedded device. The electronic device 1 may also include other external devices, such as input and output devices such as a keyboard, a mouse, a remote control, a display, a touchpad, or a voice control device.
[0034] It should be noted that the electronic device 1 is only an example. Other existing or future electronic products that are suitable for this application should also be included in the scope of protection of this application and included here by reference.
[0035] FIG2 is a flow chart of an enzyme kinetic parameter prediction method provided in one embodiment of the present application. The enzyme kinetic parameter prediction method is applied to an electronic device, such as the electronic device 1 in FIG1 , and specifically includes the following steps. The order of the steps in the flow chart may be changed, and some steps may be omitted, depending on different requirements.
[0036] Step S21: obtain training samples, and reweight each type of samples in the training samples using at least one reweighting method.
[0037] In one embodiment, the training samples include target samples of a target category.
[0038] Specifically, the training sample includes a data set containing enzyme kinetic parameters, including but not limited to the enzyme turnover number k cat , Michaelis constant K of enzyme m , the ratio of the enzyme turnover number to the enzyme's Michaelis constant k cat / K m The target samples include training samples whose enzyme kinetic parameters are within a preset value range.
[0039] For example, when the enzyme kinetic parameter is k catWhen the training sample includes the DLKcat dataset, the target sample includes k cat DLKcat data within a preset numerical range. The DLKcat data set includes ternary data, which includes enzyme sequence-substrate structure-enzyme kinetic parameters, and the enzyme kinetic parameters include enzyme turnover number k cat .
[0040] In one embodiment, the enzyme sequence in the ternary data represents the protein sequence of the enzyme, and the ternary data represents the enzyme kinetic parameters of the enzyme protein sequence in a specific substrate structure, and the enzyme kinetic parameters in the ternary data are real enzyme kinetic parameters obtained from experiments. For example, when the ternary data is enzyme sequence-substrate structure-enzyme turnover number k cat When , it represents the k of the enzyme protein sequence for a specific substrate structure cat In addition, the enzyme kinetic parameters may also include other parameters, such as the Michaelis constant K m wait.
[0041] In one embodiment, the training samples can be divided into multiple categories according to the value of the enzyme kinetic parameter, and the category with the largest value in the value range of the enzyme kinetic parameter (for example, k cat More than 10,000s -1 ) as the target category, and the samples of the target category are taken as target samples. For example, according to k cat The value of k is used to divide the DLKcat data in the same value range into one category, where k cat The DLKcat data with the largest value is taken as the target sample. In the following examples, the enzyme kinetic parameter k cat For example, the DLKcat data of the enzyme with high turnover number in the target sample is called a high-value sample.
[0042] In one embodiment, when using training samples to train a prediction model, the related art typically directly uses the proportion of samples of each class in the training samples as the weight of each class. However, due to the small number of high-value samples, if the prediction model is directly trained using related techniques, the proportion of high-value samples is too small, resulting in insufficient learning of the prediction model from the DLKcat data of high-turnover enzymes, and thus inability to accurately predict the expression level of high-turnover enzymes.
[0043] In one embodiment, in order to solve the above problem, the method provided in the present application reweights each type of sample in the training sample to give high-value samples a higher weight so that they receive more attention in model training, thereby improving the model's predictive ability for high-turnover enzymes.
[0044] In one embodiment, the at least one reweighting method includes but is not limited to: a direct modification of sample weights (DMW) method, a cost-sensitive reweighting (CSW) method, a category-balanced reweighting (CBW) method, and a label distribution smoothing (LDS) method.
[0045] In one embodiment, a directly modified sample weight (DMW) method can directly increase the weight of high-value samples to make the model pay more attention to high-value samples and improve their prediction accuracy or precision. Specifically, the DMW method includes: (1) taking the proportion of high-value samples to all training samples as the original weight, and setting a weight coefficient, i.e., a multiplier, for the high-value samples; (2) for each high-value sample, updating its weight to the product of the original weight and the weight coefficient, to the weight of the high-value sample. For example, when the weight coefficient is set to 10, the weight of the high-value sample will increase 10 times.
[0046] In addition, after increasing the weight of high-value samples, in order to avoid categories with too large or too small weights in the training sample data, the weights of all categories of data in the reweighted training samples can be normalized so that the sum of the weights corresponding to all categories of data is 1, thereby avoiding interference caused by the weight value itself.
[0047] The DMW method can adjust sample weights through simple mathematical operations, with low computational cost and no additional computational burden. However, the DMW method requires the user to manually set the weight coefficients, which may require multiple attempts and adjustments to achieve the optimal result.
[0048] In one embodiment, the cost-sensitive re-weighting method (CSW) can improve the performance of the model on an unbalanced data set by adjusting the weights of high-value samples so that the model pays more attention to high-value samples with fewer samples. CSW also includes variants such as CSW with square root and CSW with square. When the sample data is a continuous label value, the CSW method divides the sample data into discrete boxes according to the numerical range to which it belongs, and the weight of each sample is set to the inverse of the number of samples in the box to which it belongs. In addition, the CSW with square root and CSW with square method respectively reset the weight of each sample based on the weight obtained by CSW by taking the square root and squaring it.
[0049] In one embodiment, the Class-Balanced Reweighting (CBW) method can associate the weight of each sample with its class. For example, if there are few samples in the target class, the samples in the target class will be given a larger weight, so that the model pays more attention to the target samples in the target class. Specifically, the CBW method recognizes that the value of adding new data points decreases as the size of the dataset increases.
[0050] The class-balanced reweighted CBW method includes: using a preset formula to calculate the effective sample number of the target sample, the formula includes: E n =(1-β n )(1-β), where E n The target sample weight is updated to the inverse of the effective sample number. In addition to reweighting the target sample based on the effective sample number, the CBW method also performs the same process for each class of samples to complete the reweighting of the entire training sample set.
[0051] In one embodiment, the Label Distribution Smoothing (LDS) method solves the problem of unbalanced data sets by utilizing the similarity of adjacent label (or category) spaces, which can reduce the prediction error of the model for sample data with a small number of samples. Specifically, the LDS method uses a symmetric kernel to convolve with the empirical density distribution to generate a kernel-smoothed effective density distribution, and uses the effective density as the reweighted weight of each class of labels. The formula used by the LDS method can be expressed as: P(y′)=∫k(y, y′)p(y)dy,
[0052] where p(y) represents the number of times label y appears in the training samples (e.g., the number of target samples), P(y′) represents the effective density of label y′, and k(y, y′) represents the symmetric kernel.
[0053] Step S22: using the reweighted training samples to train a prediction model for enzyme kinetic parameters, and using a cross-validation method to determine the model prediction accuracy of the prediction model corresponding to each reweighting method.
[0054] In one embodiment, the model architecture of the prediction model corresponds to the category of the enzyme kinetic parameter, and the prediction model is used to predict the enzyme kinetic parameter. catWhen the prediction model includes but is not limited to the DLKcat model, the DLKcat model is a deep learning model provided by the research team of Chalmers University of Technology, which can be used to perform high-throughput k-scaling of metabolic enzymes from any organism based on substrate structure and protein sequence. cat The DLKcat model captures the k of the mutant enzyme. cat Change and identify k cat The amino acid residues that have a strong influence on the value of cat The embodiment of the present application utilizes the reweighted training samples and uses the model parameters and architecture of the DLKcat model to train a prediction model for enzyme kinetic parameters.
[0055] In one embodiment, the method of using a cross-validation method to determine the model prediction accuracy of the prediction model corresponding to each reweighting method includes: using a cross-validation method to determine the prediction accuracy of the prediction model corresponding to each reweighting method for the enzyme kinetic parameters within the preset value range (e.g., a value greater than 10000s -1 K cat ) prediction accuracy.
[0056] In one embodiment, the cross-validation method includes but is not limited to a five-fold cross-validation method, which is a K-fold cross-validation method with K=5. Specifically, the five-fold cross-validation method can be used to weight the target sample (e.g., k cat The value is greater than 10000s -1 The DLKcat data is divided into five parts of roughly equal size, each part is used as one fold of data; the first fold of data is used as the test set, and the other (second to fifth) folds are used as training sets to train the first model, and the first prediction accuracy of the first model is obtained based on the test set; the second fold of data is used as the test set, and the other (first, third to fifth) folds are used as training sets to train the second model, and the second prediction accuracy of the second model is obtained based on the test set; and so on, a total of five prediction accuracies are obtained, and the average of the five prediction accuracies is used to obtain the prediction accuracy of the model.
[0057] Step S23 , selecting the prediction model with the highest prediction accuracy as the target model, and using the target model to predict enzyme kinetic parameters of the enzyme to be predicted.
[0058] In one embodiment, the present application uses the cross-validation method to verify the prediction accuracy of the prediction model corresponding to each reweighting method, and proves that the prediction accuracy of the prediction model corresponding to each reweighting method for high-value samples is better than the initial model (initial DLKcat model), among which the prediction model corresponding to the CBW method has the highest prediction accuracy for high-value samples.
[0059] In one embodiment, the present application takes into account the distribution characteristics of samples in different value intervals, uses a sample reweighting strategy, adjusts the weights of samples in different intervals, and effectively improves the prediction performance of the prediction model for enzyme kinetic parameters in a specific interval.
[0060] In one embodiment, the method provided in the embodiment of the present application can also select other reweighting strategies to improve the prediction of enzyme kinetic parameters in a specific interval; in addition to increasing the weight of the high-value interval, the low-value interval or other specific value intervals can also be selected for targeted optimization; in addition, in addition to selecting the enzyme turnover number as the enzyme kinetic parameter, it is also possible to reweight the data sets corresponding to the enzyme kinetic parameters such as the Michaelis constant of the enzyme and test the prediction model.
[0061] In one embodiment, the enzyme kinetic parameter prediction method provided in the embodiment of the present application reweights each type of sample in the training sample to give a higher weight to the high turnover enzyme data as the target sample, so that the target sample receives more attention in the model training; the prediction model is trained using the reweighted training samples and the cross-validation algorithm to improve the prediction accuracy of the prediction model for high turnover enzymes.
[0062] FIG3 is a structural diagram of an enzyme kinetic parameter prediction device provided in an embodiment of the present application.
[0063] In some embodiments, the enzyme kinetic parameter prediction device 40 may include multiple functional modules composed of computer program segments. The computer program of each program segment in the enzyme kinetic parameter prediction device 40 may be stored in a memory of an electronic device and executed by at least one processor to perform the enzyme kinetic parameter prediction function (see FIG. 2 for details).
[0064] In the present embodiment, the enzyme kinetic parameter prediction device 40 can be divided into multiple functional modules according to the functions performed by it. The functional modules may include: a reweighting module 401, a training module 402, and a prediction module 403. The module referred to in this application refers to a series of computer program segments that can be executed by at least one processor and can complete fixed functions, which are stored in a memory. In the present embodiment, the functional implementation of each module in the enzyme kinetic parameter prediction device 40 can be found in the above definition of the enzyme kinetic parameter prediction method, and no further description is given here.
[0065] The reweighting module 401 is configured to obtain training samples and reweight each type of samples in the training samples using at least one reweighting method, wherein the training samples include target samples of a target category.
[0066] The training module 402 is used to train the prediction model of enzyme kinetic parameters using the reweighted training samples, and use a cross-validation method to determine the model prediction accuracy of the prediction model corresponding to each reweighting method.
[0067] The prediction module 403 is used to select the prediction model with the highest prediction accuracy as the target model, and use the target model to predict enzyme kinetic parameters of the enzyme to be predicted.
[0068] 1 , the memory 11 stores a computer program that, when executed by the at least one processor 12, implements all or part of the steps in the enzyme kinetic parameter prediction method. The memory 11 includes a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.
[0069] Furthermore, the computer-readable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the blockchain node, etc.
[0070] In one embodiment of the present application, a computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor 12 , the process shown in FIG. 2 is implemented.
[0071] In some embodiments, the at least one processor 12 is the control core (Control Unit) of the electronic device 1, which uses various interfaces and lines to connect the various components of the entire electronic device 1, and executes the various functions and processes data of the electronic device 1 by running or executing the programs or modules stored in the memory 11, and calling the data stored in the memory 11. For example, when the at least one processor 12 executes the computer program stored in the memory, it implements all or part of the steps of the enzyme kinetic parameter prediction method described in the embodiment of the present application; or implements all or part of the functions of the enzyme kinetic parameter prediction device. The at least one processor 12 can be composed of an integrated circuit, for example, it can be composed of a single packaged integrated circuit, or it can be composed of multiple integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors and a combination of various control chips.
[0072] In some embodiments, the at least one communication bus 13 is configured to implement connection and communication between the memory 11 and the at least one processor 12 .
[0073] Although not shown, the electronic device 1 may also include a power source (such as a battery) to power various components. Preferably, the power source may be logically connected to the at least one processor 12 via a power management device, thereby enabling the power management device to manage charging, discharging, and power consumption. The power source may also include one or more DC or AC power sources, a recharging device, a power failure detection circuit, a power converter or inverter, a power status indicator, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, cameras, etc., which will not be described in detail here.
[0074] The above-mentioned integrated unit implemented in the form of a software function module can be stored in a computer-readable storage medium. The above-mentioned software function module is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, electronic device, or network device, etc.) or a processor to execute part of the method described in each embodiment of the present application.
[0075] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the module division is only a logical function division, and other division methods may be used in actual implementation.
[0076] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical units, and may be located in one place or distributed across multiple network elements. Some or all of the modules may be selected to achieve the purpose of this embodiment based on actual needs.
[0077] In addition, the functional modules in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0078] It is obvious to those skilled in the art that the present application is not limited to the details of the above-mentioned exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the spirit or basic characteristics of the present application. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended that all changes that fall within the meaning and scope of the equivalent elements of the claims are included in the present application. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "including" does not exclude other units or, and the singular does not exclude the plural. Multiple units or devices stated in the specification may also be implemented by one unit or device through software or hardware. Words such as first, second, etc. are used to indicate names and do not indicate any particular order.
[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application.
Claims
1. A method for predicting enzyme kinetic parameters, characterized in that: The method comprises: Acquire training samples, and reweight each type of samples in the training samples using at least one reweighting method, wherein the training samples include target samples of the target category; The prediction model of enzyme kinetic parameters is trained using the reweighted training samples, and the model prediction accuracy of the prediction model corresponding to each reweighting method is determined using a cross-validation method; The prediction model with the highest prediction accuracy is selected as the target model, and the target model is used to predict the enzyme kinetic parameters of the enzyme to be predicted.
2. The method for predicting enzyme kinetic parameters according to claim 1, characterized in that: The training samples include a data set containing enzyme kinetic parameters, and the target samples include training samples whose enzyme kinetic parameters are within a preset value range.
3. The method for predicting enzyme kinetic parameters according to claim 2, characterized in that: The enzyme kinetic parameters include the enzyme turnover number k cat 、Michaelis constant K m , the ratio of enzyme turnover number to Michaelis constant k cat / K m , where, when the enzyme kinetic parameter is k cat When, the training samples include the DLKcat dataset.
4. The method for predicting enzyme kinetic parameters according to claim 3, characterized in that: The DLKcat data set includes ternary data, which includes enzyme sequence-substrate structure-enzyme kinetic parameters, and the enzyme kinetic parameters include enzyme turnover number k cat .
5. The method for predicting enzyme kinetic parameters according to claim 1, characterized in that: The at least one reweighting method includes: a direct modification of sample weights DMW method, a cost-sensitive reweighting CSW method, a category-balanced reweighting CBW method, and a label distribution smoothing LDS method.
6. The method for predicting enzyme kinetic parameters according to claim 1, characterized in that: The cross-validation method includes a five-fold cross-validation method.
7. The method for predicting enzyme kinetic parameters according to claim 4, characterized in that: The class-balanced weighted CBW method includes: The effective sample number of the target sample is calculated using a preset formula, the formula comprising: E n =(1-β n )(1-β), Among them, E n represents the effective sample number of the target sample, n represents the number of the target samples, and β represents a hyperparameter with a preset value range of (0,1); The weight of the target sample is updated to the inverse of the number of valid samples.
8. The method for predicting enzyme kinetic parameters according to claim 2, characterized in that: The method of using the cross-validation method to determine the model prediction accuracy of the prediction model corresponding to each re-weighting method includes: The cross-validation method is used to determine the prediction accuracy of the prediction model corresponding to each re-weighted method for the enzyme kinetic parameters within the preset numerical range.
9. An electronic device, characterized in that: The electronic device comprises a processor and a memory, and the processor is used to implement the enzyme kinetic parameter prediction method according to any one of claims 1 to 8 when executing the computer program stored in the memory.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting enzyme kinetic parameters according to any one of claims 1 to 8 is implemented.
Citation Information
Patent Citations
Protein structure class prediction method based on weighted composition position vectors and support vector machine
CN105205343A
Zinc-binding protein action site prediction method based on integrated learning in non-equilibrium mode
CN109326329A
Classifier, neural network model training method, data processing device, and medium
CN109902722A
Classification network training method and device, computer equipment and storage medium
CN111160394A
Network training method based on reweighting strategy
CN114463576A
