Plant polysaccharide data analysis method and device, electronic equipment and storage medium

Abstract

The invention provides a plant polysaccharide data analysis method and device, electronic equipment and a storage medium, and relates to the technical field of biologication.The method comprises the steps that based on a plant polysaccharide database, plant polysaccharide data of target plant polysaccharide is obtained; inputting the plant polysaccharide data of the target plant polysaccharide into a plant polysaccharide analysis model to obtain a recommended extraction mode of the target plant polysaccharide and an extraction parameter predicted value corresponding to the recommended extraction mode; determining an extraction parameter actual value of the target plant polysaccharide in the recommended extraction mode based on the extraction rate actual value and / or the biological activity evaluation index actual value of the extraction experiment of the target plant polysaccharide; the extraction experiment is carried out based on a recommended extraction mode and an extraction parameter prediction value; and storing the recommended extraction mode and the actual value of the extraction parameter into a plant polysaccharide database, and retraining the plant polysaccharide analysis model. According to the method and the device provided by the invention, the extraction efficiency and the development efficiency of the plant polysaccharide are improved.

Description

Technical Field

[0001] The present invention relates to the field of biotechnology, and in particular to a plant polysaccharide data analysis method, device, electronic equipment and storage medium. Background Art

[0002] Plant polysaccharides are an important class of ingredients in natural products and have attracted attention due to their wide applications in food, medicine, cosmetics and other fields. Traditional plant polysaccharide extraction and separation methods usually rely on manual operations in the laboratory, such as solvent extraction, precipitation, centrifugation, etc. Although these methods can obtain certain polysaccharide products, they have shortcomings such as low extraction efficiency, lack of standardization of experimental processes, and biological activity data that mostly rely on experimental testing and are difficult to predict.

[0003] Therefore, how to analyze plant polysaccharide data and improve the extraction efficiency and development efficiency of plant polysaccharides has become a technical problem that needs to be urgently solved in the industry. Summary of the invention

[0004] The present invention provides a plant polysaccharide data analysis method, device, electronic equipment and storage medium, which are used to solve the technical problem of how to analyze plant polysaccharide data and improve the extraction efficiency and development efficiency of plant polysaccharides.

[0005] The present invention provides a plant polysaccharide data analysis method, comprising: Based on the plant polysaccharide database, obtaining plant polysaccharide data of target plant polysaccharides; Inputting the plant polysaccharide data of the target plant polysaccharide into a plant polysaccharide analysis model, and obtaining a recommended extraction method for the target plant polysaccharide and extraction parameter prediction values ​​corresponding to the recommended extraction method output by the plant polysaccharide analysis model; Determine the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide; the extraction experiment of the target plant polysaccharide is carried out based on the recommended extraction method and the predicted value of the extraction parameter; Storing the recommended extraction method and the actual values ​​of the extraction parameters in the plant polysaccharide database, and retraining the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters; The plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods.

[0006] In some embodiments, the plant polysaccharide data includes the polysaccharide name, biological activity, molecular structure, plant source, extraction method and extraction parameters under the extraction method; the extraction parameters include extraction temperature, extraction time, enzyme type and extraction pressure.

[0007] In some embodiments, before inputting the plant polysaccharide data of the target plant polysaccharide into the plant polysaccharide analysis model and obtaining the recommended extraction method of the target plant polysaccharide and the predicted extraction parameter value corresponding to the recommended extraction method output by the plant polysaccharide analysis model, the method further comprises: Determining a current plant polysaccharide analysis model among multiple plant polysaccharide analysis models; Wherein, the plurality of plant polysaccharide analysis models include a first plant polysaccharide analysis model, a second plant polysaccharide analysis model and a third plant polysaccharide analysis model; The first plant polysaccharide analysis model is obtained after training based on the first loss function; the second plant polysaccharide analysis model is obtained after training based on the second loss function; the third plant polysaccharide analysis model is obtained after training based on the first loss function and the second loss function; The first loss function is determined based on the error between the predicted value of the extraction rate of each sample plant polysaccharide under different extraction methods and the actual value of the extraction rate; the second loss function is determined based on the error between the predicted value of the biological activity evaluation index of each sample plant polysaccharide under different extraction methods and the actual value of the biological activity evaluation index.

[0008] In some embodiments, the plant polysaccharide analysis model includes a first prediction layer and a second prediction layer; The first prediction layer includes a plurality of prediction sub-models arranged in parallel, and is used to send prediction results of each prediction sub-model to the second prediction layer; the prediction sub-models include a random forest model, a gradient boosting regression model and a support vector machine model; The second prediction layer is used to fuse the prediction results of each prediction sub-model to generate a recommended extraction method for the target plant polysaccharide and predicted values ​​of extraction parameters corresponding to the recommended extraction method; the second prediction layer is a linear regression model or a neural network model.

[0009] In some embodiments, the plant polysaccharide analysis model is trained based on the following method: Add random noise to the actual values ​​of the extraction parameters of each sample of plant polysaccharide under different extraction methods to obtain the enhanced values ​​of the extraction parameters of each sample of plant polysaccharide under different extraction methods; The plant polysaccharide analysis model is trained based on the plant polysaccharide data samples of each sample plant polysaccharide and the extraction parameter enhancement values ​​of each sample plant polysaccharide under different extraction methods.

[0010] In some embodiments, the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method is determined based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide, including: Based on the error between the actual extraction rate and the predicted extraction rate, and / or the error between the actual value of the biological activity evaluation index and the predicted value of the biological activity evaluation index, the predicted value of the extraction parameter under the recommended extraction method is adjusted to obtain the actual value of the extraction parameter under the recommended extraction method.

[0011] In some embodiments, retraining the plant polysaccharide analysis model based on the recommended extraction method and the actual value of the extraction parameter includes: Based on the recommended extraction method of the target plant polysaccharide and the actual values ​​of the extraction parameters corresponding to the recommended extraction method, generating a retraining sample, and adding the retraining sample to the retraining sample set; When the number of samples in the retraining sample set is greater than a preset number, the plant polysaccharide analysis model is retrained based on the retraining sample set.

[0012] The present invention provides a plant polysaccharide data analysis device, comprising: An acquisition unit, used for acquiring plant polysaccharide data of a target plant polysaccharide based on a plant polysaccharide database; A prediction unit, used for inputting the plant polysaccharide data of the target plant polysaccharide into a plant polysaccharide analysis model, and obtaining a recommended extraction method of the target plant polysaccharide output by the plant polysaccharide analysis model and a predicted value of an extraction parameter corresponding to the recommended extraction method; An extraction unit, used to determine the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide; the extraction experiment of the target plant polysaccharide is carried out based on the recommended extraction method and the predicted value of the extraction parameter; An optimization unit, used for storing the recommended extraction method and the actual values ​​of the extraction parameters in the plant polysaccharide database, and retraining the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters; The plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods.

[0013] The present invention provides an electronic device, comprising a memory, a processor and a computer program stored in the memory and running on the processor, wherein the plant polysaccharide data analysis method is implemented when the processor executes the computer program.

[0014] The present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, and the computer program implements the plant polysaccharide data analysis method when executed by a processor.

[0015] The plant polysaccharide data analysis method, device, electronic device and storage medium provided by the present invention obtain the plant polysaccharide data of the target plant polysaccharide according to the plant polysaccharide database; input the plant polysaccharide data of the target plant polysaccharide into the plant polysaccharide analysis model to obtain the recommended extraction method of the target plant polysaccharide and the predicted values ​​of extraction parameters corresponding to the recommended extraction method output by the plant polysaccharide analysis model; determine the actual values ​​of the extraction parameters of the target plant polysaccharide under the recommended extraction method according to the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide; the extraction experiment of the target plant polysaccharide is carried out according to the recommended extraction method and the predicted values ​​of the extraction parameters; store the recommended extraction method and the actual values ​​of the extraction parameters in the plant polysaccharide database, and perform the extraction of the plant polysaccharide according to the recommended extraction method and the actual values ​​of the extraction parameters. The analysis model is retrained; since the plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods, it can effectively use the data in the plant polysaccharide database for analysis, and fully utilize the data through data mining; since the extraction parameter prediction values ​​output by the plant polysaccharide analysis model are adjusted by the actual values ​​of the extraction parameters of the extraction experiment of the target plant polysaccharide, and the plant polysaccharide database is updated and the plant polysaccharide analysis model is retrained, the accuracy of the extraction parameter prediction is improved, and by combining model prediction with extraction experiments, a large number of experiments can be reduced, the experimental cost is reduced, and the extraction efficiency and development efficiency of plant polysaccharides are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0017] In order to more clearly illustrate the technical solutions in the present invention or 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 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.

[0018] Figure 1 It is a schematic diagram of the process of the plant polysaccharide data analysis method provided by the present invention.

[0019] Figure 2 It is a schematic diagram of the structure of the plant polysaccharide data analysis device provided by the present invention.

[0020] Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0022] It should be noted that the terms "first", "second", etc. in the present invention are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units or modules is not necessarily limited to those steps or units or modules that are clearly listed, but may include other steps or units or modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] Although the extraction efficiency of some plant polysaccharides has been improved through high-throughput sequencing, ultrasonic extraction and other related technologies, the systematic analysis of biological activity is still insufficient. Researchers often collect data in isolation, forming independent databases, lacking horizontal communication and data integration, so that the potential research value cannot be fully explored. Traditional research methods often focus on the experimental process, but lack in-depth analysis and application of data, which limits the understanding of the relationship between polysaccharide structure and function. There is no effective model to predict the activity of polysaccharides, resulting in a large amount of experimental uncertainty, which in turn wastes scientific research resources.

[0024] In order to solve the shortcomings of related technologies, Figure 1 is a schematic diagram of the process of the plant polysaccharide data analysis method provided by the present invention, such as Figure 1 As shown, the method includes step 110 , step 120 , step 130 and step 140 .

[0025] Step 110: Based on the plant polysaccharide database, obtain plant polysaccharide data of the target plant polysaccharide.

[0026] Specifically, the executor of the plant polysaccharide data analysis method provided in the embodiment of the present invention is a plant polysaccharide data analysis device. The device can be implemented by software, such as a plant polysaccharide data analysis program; or by hardware, such as a processor, computer or server that executes the plant polysaccharide data analysis method.

[0027] Plant polysaccharides, also known as plant polysaccharides, refer to a class of macromolecular compounds in plants that are composed of multiple monosaccharide molecules connected by glycosidic bonds. Plant polysaccharide data include polysaccharide name, biological activity, molecular structure, plant source, extraction method, and extraction parameters under different extraction methods.

[0028] The name of polysaccharide refers to the name of plant polysaccharide, which may include Chinese name, English name, full name or abbreviation, etc. Molecular structure refers to the type of monosaccharide unit, glycosidic bond type, branching structure, molecular weight and chain length in plant polysaccharide. Plant source refers to the plant from which plant polysaccharide can be extracted. Biological activity refers to the effect or function of plant polysaccharide on organisms (such as human body), including antioxidant, anti-aging, enzyme inhibition, biostimulation, immunomodulation and anti-inflammatory, etc. Extraction method refers to the method of extracting plant polysaccharide from plants, including solvent extraction, ultrasonic extraction, enzymatic hydrolysis, etc. Extraction parameters include extraction temperature, extraction time, enzyme type and extraction pressure, etc. Extraction temperature refers to the temperature during the extraction process of plant polysaccharide. Extraction time refers to the duration of the extraction process of plant polysaccharide. Enzyme type refers to the type of enzyme used in the extraction process of plant polysaccharide. Extraction pressure refers to the operating pressure of the extraction process of plant polysaccharide.

[0029] The plant polysaccharide database can be a database specifically used to store plant polysaccharide data. The plant polysaccharide database can be set to store plant polysaccharide data. Three data tables can be set in the plant polysaccharide database, namely the plant information table, the polysaccharide information table and the experimental data table. The plant information table is mainly used to store relevant information about plants, which may include plant names, plant families, Latin names, etc. The polysaccharide information table is mainly used to store relevant information about plant polysaccharides, which may include polysaccharide names, plant sources, molecular structures, biological activities and extraction methods, etc. The experimental data table is mainly used to store relevant information about plant polysaccharide extraction methods, which may include monosaccharide composition, extraction methods and experimental conditions (temperature, humidity and enzyme types), etc. The plant information table and the polysaccharide information table can be associated with the plant name as a foreign key; the polysaccharide information table and the experimental data table can be associated with the polysaccharide name as a foreign key.

[0030] The target plant polysaccharide is the plant polysaccharide to be searched. Keywords can be determined based on at least one of the polysaccharide name, molecular structure, plant source, biological activity and extraction method of the target plant polysaccharide, and searched in the plant polysaccharide database to obtain more comprehensive plant polysaccharide data of the target plant polysaccharide.

[0031] Step 120, input the plant polysaccharide data of the target plant polysaccharide into the plant polysaccharide analysis model, and obtain the recommended extraction method of the target plant polysaccharide output by the plant polysaccharide analysis model and the extraction parameter prediction value corresponding to the recommended extraction method. The plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of the extraction parameters of each sample plant polysaccharide under different extraction methods.

[0032] Specifically, for each plant polysaccharide, there can be a variety of extraction methods, such as solvent extraction, ultrasonic extraction, enzymatic hydrolysis, etc. Under each extraction method, there are corresponding different extraction parameters.

[0033] A plant polysaccharide analysis model can be constructed through machine learning methods to predict the extraction method of the target plant polysaccharide, obtain the recommended extraction method, and the predicted values ​​of the extraction parameters under the recommended extraction method. For example, the plant polysaccharide analysis model can determine one of the enzymatic hydrolysis-hot water extraction method and the ultrasonic-assisted extraction method as the recommended extraction method, and predict the relevant extraction parameters.

[0034] A neural network can be used as an initial model to collect a large number of plant polysaccharide data samples of sample plant polysaccharides and actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods from a plant polysaccharide database. The initial model is trained based on the plant polysaccharide data samples, extraction methods and actual values ​​of extraction parameters of each sample plant polysaccharide to improve the model's prediction ability for extraction methods and extraction parameters, and ultimately obtain a plant polysaccharide analysis model.

[0035] The plant polysaccharide data of the target plant polysaccharide can be input into the plant polysaccharide analysis model, and the plant polysaccharide analysis model analyzes and processes the plant polysaccharide data of the target plant polysaccharide to obtain the recommended extraction method of the target plant polysaccharide and the extraction parameter prediction value corresponding to the recommended extraction method.

[0036] Step 130, determining the actual value of the extraction parameters of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the target plant polysaccharide extraction experiment; the extraction experiment of the target plant polysaccharide is carried out based on the recommended extraction method and the predicted value of the extraction parameter.

[0037] Specifically, the extraction experiment of the target plant polysaccharide can be carried out according to the recommended extraction method and the predicted value of the extraction parameter. In the extraction experiment, the extraction rate and the biological activity evaluation index are relatively important extraction indicators.

[0038] The extraction rate is a quantitative indicator for evaluating the extraction effect of plant polysaccharides. It characterizes the efficiency of the extraction process, and the higher the better. The biological activity evaluation index is a quantitative indicator for evaluating the biological activity of plant polysaccharides. The IC50 value (half-maximal inhibition concentration, Inhibitory Concentration 50%) can be selected as the biological activity evaluation index. The IC50 value is an important quantitative indicator for evaluating biological activity, especially inhibitory activity. It provides researchers with a standardized way to compare and evaluate the effects of compounds at different concentrations, and is widely used in drug development, enzymology research, cell biology and other fields.

[0039] The predicted value is obtained by the plant polysaccharide analysis model after reasoning and prediction based on the plant polysaccharide data, while the actual value is determined based on the extraction experiment. There may be errors between the two. In the extraction experiment, the predicted value of the extraction parameter can be adjusted according to the actual value of the extraction rate and / or the actual value of the biological activity evaluation index, and finally the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method is obtained.

[0040] Step 140: store the recommended extraction method and the actual values ​​of the extraction parameters in the plant polysaccharide database, and retrain the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters.

[0041] Specifically, the recommended extraction method and the actual values ​​of the extraction parameters are stored in the plant polysaccharide database, and the data records related to the extraction method in the plant polysaccharide data of the target plant polysaccharide can be updated. In addition, the recommended extraction method and the actual values ​​of the extraction parameters can be used as new learning samples to retrain the plant polysaccharide analysis model to improve the accuracy of the model prediction.

[0042] The plant polysaccharide data analysis method provided in the embodiment of the present invention obtains the plant polysaccharide data of the target plant polysaccharide according to the plant polysaccharide database; inputs the plant polysaccharide data of the target plant polysaccharide into the plant polysaccharide analysis model to obtain the recommended extraction method of the target plant polysaccharide and the predicted values ​​of extraction parameters corresponding to the recommended extraction method output by the plant polysaccharide analysis model; determines the actual values ​​of extraction parameters of the target plant polysaccharide under the recommended extraction method according to the actual value of extraction rate and / or the actual value of biological activity evaluation index of the extraction experiment of the target plant polysaccharide; the extraction experiment of the target plant polysaccharide is carried out according to the recommended extraction method and the predicted values ​​of extraction parameters; stores the recommended extraction method and the actual values ​​of extraction parameters in the plant polysaccharide database, and performs the plant polysaccharide analysis model according to the recommended extraction method and the actual values ​​of extraction parameters. retraining; since the plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods, it can effectively utilize the data in the plant polysaccharide database for analysis, and fully utilize the data through data mining; since the extraction parameter prediction values ​​output by the plant polysaccharide analysis model are adjusted by the actual values ​​of the extraction parameters of the extraction experiment of the target plant polysaccharides, and the plant polysaccharide database is updated and the plant polysaccharide analysis model is retrained, the accuracy of the extraction parameter prediction is improved, and through the combination of model prediction and extraction experiment, a large number of experiments can be reduced, the experimental cost is reduced, and the extraction efficiency and development efficiency of plant polysaccharides are improved.

[0043] It should be noted that each implementation of the present invention can be freely combined, the order can be changed, or it can be executed separately, and does not need to rely on or depend on a fixed execution order.

[0044] In some embodiments, before inputting the plant polysaccharide data of the target plant polysaccharide into the plant polysaccharide analysis model and obtaining the recommended extraction method of the target plant polysaccharide and the extraction parameter prediction value corresponding to the recommended extraction method output by the plant polysaccharide analysis model, the method further comprises: Determining a current plant polysaccharide analysis model among multiple plant polysaccharide analysis models; Wherein, the plurality of plant polysaccharide analysis models include a first plant polysaccharide analysis model, a second plant polysaccharide analysis model and a third plant polysaccharide analysis model; The first plant polysaccharide analysis model is obtained after training based on the first loss function; the second plant polysaccharide analysis model is obtained after training based on the second loss function; the third plant polysaccharide analysis model is obtained after training based on the first loss function and the second loss function; The first loss function is determined based on the error between the predicted value of the extraction rate of each sample plant polysaccharide under different extraction methods and the actual value of the extraction rate; the second loss function is determined based on the error between the predicted value of the biological activity evaluation index of each sample plant polysaccharide under different extraction methods and the actual value of the biological activity evaluation index.

[0045] Specifically, the loss function measures the gap between the model's predicted value and the actual value. Different loss functions will affect the model's learning method, optimization process, and final performance. During the model training process, different loss functions can be set to make the trained plant polysaccharide analysis model have better performance under different optimization objectives.

[0046] The optimization objectives of the model may include having a higher extraction rate, having a higher bioactivity evaluation index (IC50 value), or achieving a balance between the extraction rate and the bioactivity evaluation index. Therefore, a first loss function and a second loss function can be determined. The first loss function is determined based on the error between the predicted value of the extraction rate and the actual value of the extraction rate of each sample plant polysaccharide under different extraction methods, for example, the first loss function is determined by the mean squared error (MSE) between the predicted value of the extraction rate and the actual value of the extraction rate. The second loss function is determined based on the error between the predicted value of the bioactivity evaluation index and the actual value of the bioactivity evaluation index under different extraction methods of each sample plant polysaccharide, for example, the second loss function is determined by the mean squared error between the predicted value of the bioactivity evaluation index and the actual value of the bioactivity evaluation index.

[0047] By using the first loss function for optimization training, a first plant polysaccharide analysis model can be obtained, which has a higher extraction rate; by using the second loss function for optimization training, a second plant polysaccharide analysis model can be obtained, which has a higher bioactivity evaluation index; by combining the first loss function and the second loss function for optimization training, a third plant polysaccharide analysis model can be obtained, which has a balance between the extraction rate and the bioactivity evaluation index. The specific combination method can be: different weights are set for the first loss function and the second loss function respectively, and based on the weights corresponding to the two loss functions, the two loss functions are weighted summed to obtain the final loss function.

[0048] When using the model for reasoning, the current plant polysaccharide analysis model can be determined from the first plant polysaccharide analysis model, the second plant polysaccharide analysis model and the third plant polysaccharide analysis model according to actual needs, and the recommended method and extraction parameters are predicted using the current plant polysaccharide analysis model.

[0049] The plant polysaccharide data analysis method provided by the embodiment of the present invention obtains plant polysaccharide analysis models with different optimization objectives according to different loss function training, which can reduce a large number of experiments, reduce experimental costs, and improve the extraction efficiency of plant polysaccharides.

[0050] In some embodiments, the plant polysaccharide analysis model includes a first prediction layer and a second prediction layer; The first prediction layer includes a plurality of prediction sub-models arranged in parallel, and is used to send the prediction results of each prediction sub-model to the second prediction layer; the prediction sub-models include a random forest model, a gradient boosting regression model, and a support vector machine model; The second prediction layer is used to fuse the prediction results of each prediction sub-model to generate a recommended extraction method for the target plant polysaccharide and the predicted values ​​of the extraction parameters corresponding to the recommended extraction method; the second prediction layer is a linear regression model or a neural network model.

[0051] Specifically, the model structure of the plant polysaccharide analysis model may include a first prediction layer and a second prediction layer connected in sequence.

[0052] The first prediction layer may include multiple prediction sub-models set in parallel. These prediction sub-models can independently process the input sample data, obtain their own prediction results, and then send the prediction results to the second prediction layer. The type and number of prediction sub-models can be set as needed, for example, including random forest model, gradient boosting regression model and support vector machine model.

[0053] The second prediction layer is used to fuse the prediction results of each prediction sub-model to generate the final prediction result, that is, the recommended extraction method of the target plant polysaccharide and the predicted value of the extraction parameter corresponding to the recommended extraction method. The second prediction layer can select a linear regression model or a neural network model.

[0054] The plant polysaccharide data analysis method provided by the embodiment of the present invention, the plant polysaccharide analysis model includes a first prediction layer and a second prediction layer, the first prediction layer performs basic prediction, and can use different models to learn different features and perform predictions; the second prediction layer performs fusion prediction, and can learn the error patterns of different basic models to further optimize the results; through the combination of the two layers, the accuracy of the extraction parameter prediction is improved.

[0055] In some embodiments, the plant polysaccharide analysis model is trained based on the following method: Add random noise to the actual values ​​of the extraction parameters of each sample of plant polysaccharide under different extraction methods to obtain the enhanced values ​​of the extraction parameters of each sample of plant polysaccharide under different extraction methods; The plant polysaccharide analysis model is trained based on the plant polysaccharide data samples of each sample plant polysaccharide and the extraction parameter enhancement values ​​of each sample plant polysaccharide under different extraction methods.

[0056] Specifically, during the training process of the plant polysaccharide analysis model, a large number of samples can be collected, including plant polysaccharide data samples of each sample plant polysaccharide and actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods.

[0057] These samples can be divided into training set, validation set and test set according to a certain quantitative ratio, such as 7:2:1. In the process of training, validating and testing the model, stratified sampling method can be used to ensure that the distribution of target variables in the training set, validation set and test set is consistent with the distribution of target variables in the entire data set.

[0058] Data enhancement is a technique that improves the generalization ability of machine learning models by artificially increasing the number and diversity of training samples. In practical applications, data usually has randomness due to various reasons (such as measurement errors, sensor noise, fluctuations in experimental conditions, etc.). Adding noise helps to make the model more robust, avoid overfitting, and better handle uncertainties in the real world. In an embodiment of the present invention, random noise can be added to the actual values ​​of the extraction parameters of each sample plant polysaccharide under different extraction methods. For example, a random noise of ±5% is added to the actual value of the extraction parameter to obtain an enhanced value of the extraction parameter. The value of the random noise can be randomly selected from [-5%*actual value of the extraction parameter, 5%*actual value of the extraction parameter]. The enhanced value of the extraction parameter is the sum of the actual value of the extraction parameter and the random noise.

[0059] Finally, the plant polysaccharide analysis model is trained based on the plant polysaccharide data samples of each sample of plant polysaccharides and the extraction parameter enhancement values ​​of each sample of plant polysaccharides under different extraction methods.

[0060] The plant polysaccharide data analysis method provided by the embodiment of the present invention adds random noise to the actual value of the extraction parameter, which can simulate experimental errors and effectively increase the diversity of data, thereby improving the generalization ability of the plant polysaccharide analysis model.

[0061] In some embodiments, determining the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the target plant polysaccharide extraction experiment includes: Based on the error between the actual extraction rate and the predicted extraction rate, and / or the error between the actual value of the biological activity evaluation index and the predicted value of the biological activity evaluation index, the predicted value of the extraction parameter under the recommended extraction method is adjusted to obtain the actual value of the extraction parameter under the recommended extraction method.

[0062] Specifically, when performing an extraction experiment of a target plant polysaccharide, the predicted value of the extraction parameter under the recommended extraction method can be adjusted according to the error between the actual value of the extraction rate and the predicted value of the extraction rate to obtain the actual value of the extraction parameter under the recommended extraction method. The actual value of the extraction parameter can maximize the extraction rate of the extraction experiment.

[0063] The predicted value of the extraction parameter under the recommended extraction method can also be adjusted according to the error between the actual value of the biological activity evaluation index and the predicted value of the biological activity evaluation index to obtain the actual value of the extraction parameter under the recommended extraction method. The actual value of the extraction parameter can minimize the biological activity evaluation index of the extraction experiment.

[0064] The error between the actual extraction rate and the predicted extraction rate, as well as the error between the actual value of the biological activity evaluation index and the predicted value of the biological activity evaluation index, can also be considered at the same time, and the predicted value of the extraction parameter under the recommended extraction method can be adjusted to obtain the actual value of the extraction parameter under the recommended extraction method. The actual value of the extraction parameter can balance the extraction rate and the biological activity evaluation index of the extraction experiment.

[0065] The plant polysaccharide data analysis method provided in an embodiment of the present invention adjusts the extraction parameter prediction value according to the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the target plant polysaccharide extraction experiment, obtains the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method, stores the adjusted actual value of the extraction parameter in a plant polysaccharide database, and retrains the plant polysaccharide analysis model, which can further improve the accuracy of the extraction parameter prediction.

[0066] In some embodiments, retraining the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters includes: Based on the recommended extraction method of the target plant polysaccharide and the actual values ​​of the extraction parameters corresponding to the recommended extraction method, a retraining sample is generated, and the retraining sample is added to the retraining sample set; When the number of samples in the retraining sample set is greater than the preset number, the plant polysaccharide analysis model is retrained based on the retraining sample set.

[0067] Specifically, the retraining process of the plant polysaccharide analysis model can be triggered by the number of samples. The recommended extraction methods of the target plant polysaccharides and the actual values ​​of the extraction parameters corresponding to the recommended extraction methods can be organized as retraining samples and added to the retraining sample set.

[0068] The preset number can be set as needed, for example, 20. When the number of samples in the retraining sample set is greater than the preset number, the plant polysaccharide analysis model can be retrained based on the samples in the retraining sample set.

[0069] The plant polysaccharide data analysis method provided in an embodiment of the present invention retrains the plant polysaccharide analysis model when the number of samples in the retraining sample set is greater than a preset number. On the one hand, this can improve the prediction accuracy of the plant polysaccharide analysis model and maintain dynamic updates; on the other hand, it also avoids frequent retraining of the plant polysaccharide analysis model, which wastes computing resources and avoids increasing training costs.

[0070] Based on the above embodiments, the present invention provides two specific embodiments to illustrate the above method.

[0071] The first specific embodiment is to extract longan polysaccharide.

[0072] The extraction of longan polysaccharides has attracted wide attention due to its antioxidant properties. This example uses a machine learning model to predict the optimal extraction parameters (including enzyme type, enzymolysis temperature, enzymolysis time, and hot water extraction temperature), and optimizes the experimental process based on the model output to verify the extraction rate and biological activity. In this example, the extraction method is limited to enzymolysis-hot water extraction.

[0073] Random Forest and Neural Network can be used to build a plant polysaccharide data analysis model. The number of trees (n_estimators) of the random forest model is 200, and the maximum depth (max_depth) is 10. The neural network includes an input layer, a hidden layer, and an output layer. The hidden layer can include 3 network layers with 128, 64, and 32 nodes respectively. The ReLU activation function is used. The output layer includes 2 nodes, which predict the extraction rate and IC50 value respectively. Loss function The joint mean square error (MSE) can be used. . and are weights respectively.

[0074] The input features of the plant polysaccharide data analysis model are mainly extraction parameters, including enzyme type (cellulase / pectinase / no enzyme treatment, One-Hot coding), enzymatic hydrolysis temperature: 45°C-65°C, enzymatic hydrolysis time: 2-4 hours and hot water extraction temperature: 80°C-90°C. The goal of the plant polysaccharide data analysis model is to maximize the extraction rate and minimize the IC50 value.

[0075] 500 sets of experimental data can be prepared, with 70% of the training set, 15% of the validation set, and 15% of the test set. ±5% noise is added to the experimental parameters to improve the robustness of the model. During the training process, the random forest optimizes the hyperparameters through cross-validation, and the neural network uses the Adam optimizer (learning rate 0.001), trains for 50 epochs (rounds), and the batch size is 32.

[0076] In terms of extraction rate prediction, the MSE was 3.5% and the goodness of fit (R 2 ) was 0.92. For IC50 prediction, the MSE was 5.2 μg / mL (micrograms per milliliter) and the goodness of fit (R 2 ) is 0.88.

[0077] After model calculation, the optimal combination of extraction parameters under the enzymatic hydrolysis-hot water extraction method is: enzyme type is cellulase; enzymatic hydrolysis temperature is 55°C; enzymatic hydrolysis time is 4 hours; hot water extraction temperature is 90°C. The extraction rate is expected to be 70%. The IC50 value is expected to be 60 μg / mL.

[0078] Next, the extraction experiment was carried out. The experimental optimization goal was to maximize the extraction rate and minimize the IC50 value. The optimization directions included: combining the enzymatic hydrolysis and extraction processes into a continuous system in the process to reduce the residence time; using precise temperature control equipment to ensure that the temperature is stable within ±1°C. In resource optimization, the amount of solvent was optimized, and 20% of water resources were saved by adjusting the material-liquid ratio; the solvent concentration was fixed to pure water to reduce the interference of solvent types.

[0079] According to the model recommendation, experimental tests of key variables were performed. Enzyme hydrolysis temperature: 45°C-65°C. Enzyme hydrolysis time: 2-4 hours. Hot water extraction temperature: 80°C-90°C. The results verified that the actual extraction rate was 68% (deviation -2%). The actual IC50 value: 63 μg / mL (deviation +3 μg / mL).

[0080] The experimental results can be uploaded to the database, and the model can be dynamically updated to improve the prediction performance.

[0081] The second specific embodiment is to extract dandelion polysaccharides.

[0082] Dandelion polysaccharides have become a research focus due to their significant anti-tumor and immunomodulatory functions. This example optimizes the ultrasonic extraction conditions (ultrasonic power, extraction time, and solvent concentration) and combines the machine learning model to recommend the optimal solution to improve the polysaccharide extraction efficiency and biological activity. In this example, the extraction method is limited to ultrasonic-assisted extraction.

[0083] Random Forest and Neural Network can be used to construct a plant polysaccharide data analysis model. The number of trees (n_estimators) of the random forest model is 200, and the maximum depth (max_depth) is 12. The neural network includes an input layer, a hidden layer, and an output layer. The hidden layer can include 3 network layers with 128, 64, and 32 nodes, respectively. The ReLU activation function is used. The output layer includes 2 nodes, which predict the extraction rate and IC50 value, respectively. The loss function L can use the joint mean square error (MSE). Ensemble learning is also used, and the model combination is random forest, gradient boosting regression, and support vector regression, with weights of 0.5, 0.3, and 0.2, respectively.

[0084] The input features of the plant polysaccharide data analysis model are mainly extraction parameters, including ultrasonic power (200W, 400W, 600W; W is the power unit: watt), extraction time (15min, 30min), solvent concentration (ethanol concentration: 40%, 60%, 80%). The goal of the plant polysaccharide data analysis model is to maximize the extraction rate and minimize the IC50 value.

[0085] 400 sets of experimental data can be prepared, with 70% for training, 15% for validation, and 15% for testing. ±5% random noise is added to the solvent concentration and extraction time to improve the robustness of the model. During the training process, random forest and ensemble learning optimize hyperparameters through cross-validation, and the neural network uses the Adam optimizer (learning rate 0.001) for 50 epochs (rounds) with a batch size of 32.

[0086] For the prediction of extraction rate, the MSE was 3.8% and the goodness of fit (R2) was 0.91. For the prediction of IC50, the MSE was 5.0 μg / mL (micrograms per milliliter) and the goodness of fit (R2) was 0.89.

[0087] After model calculation, the optimal combination of extraction parameters under ultrasonic-assisted extraction is: ultrasonic power: 600W. Extraction time: 30min. Solvent concentration: 60% ethanol. The expected extraction rate is 72%. The expected IC50 value is 65 μg / mL.

[0088] Next, the extraction experiment was carried out. The experimental optimization goal was to maximize the extraction rate and minimize the IC50 value. The optimization directions included: using dynamic power regulation technology in process optimization to keep the ultrasonic output within the range of ±10W to ensure the stability of the extraction process; controlling the extraction time in stages (such as starting at 400W for the first 15 minutes and increasing to 600W for the next 15 minutes). In resource optimization, the material-liquid ratio was optimized to 1:15 (g / mL), reducing ethanol consumption by 10%; using a recovery solvent system to achieve solvent recycling.

[0089] According to the model recommendation, experimental tests of key variables were carried out, and some experimental data tables were obtained, as shown in Table 1.

[0090] Table 1 Experimental data of ultrasonic assisted extraction method

[0091] The results showed that the actual extraction rate was 70% (deviation -2%) and the actual IC50 value was 67 μg / mL (deviation +2 μg / mL).

[0092] The experimental results can be uploaded to the database, and the model can be dynamically updated to improve the prediction performance. When more than 20 groups of new data are added, the model can be retrained and updated, such as dynamically adjusting the number of hidden layer nodes of the neural network and re-optimizing the tree depth and sample split number of the random forest.

[0093] The following describes an apparatus provided by an embodiment of the present invention. The apparatus described below and the method described above can refer to each other.

[0094] Figure 2 is a schematic diagram of the structure of the plant polysaccharide data analysis device provided by the present invention, such as Figure 2 As shown, the device comprises: An acquisition unit 210 is used to acquire plant polysaccharide data of a target plant polysaccharide based on a plant polysaccharide database; A prediction unit 220, for inputting plant polysaccharide data of a target plant polysaccharide into a plant polysaccharide analysis model, and obtaining a recommended extraction method of the target plant polysaccharide output by the plant polysaccharide analysis model and extraction parameter prediction values ​​corresponding to the recommended extraction method; An extraction unit 230 is used to determine the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the target plant polysaccharide extraction experiment; the extraction experiment of the target plant polysaccharide is performed based on the recommended extraction method and the predicted value of the extraction parameter; An optimization unit 240, for storing the recommended extraction method and the actual values ​​of the extraction parameters in the plant polysaccharide database, and retraining the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters; The plant polysaccharide analysis model is obtained after training based on plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods.

[0095] The plant polysaccharide data analysis device provided in the embodiment of the present invention obtains the plant polysaccharide data of the target plant polysaccharide according to the plant polysaccharide database; inputs the plant polysaccharide data of the target plant polysaccharide into the plant polysaccharide analysis model to obtain the recommended extraction method of the target plant polysaccharide and the predicted values ​​of extraction parameters corresponding to the recommended extraction method output by the plant polysaccharide analysis model; determines the actual values ​​of extraction parameters of the target plant polysaccharide under the recommended extraction method according to the actual value of extraction rate and / or the actual value of biological activity evaluation index of the extraction experiment of the target plant polysaccharide; the extraction experiment of the target plant polysaccharide is carried out according to the recommended extraction method and the predicted values ​​of extraction parameters; stores the recommended extraction method and the actual values ​​of extraction parameters in the plant polysaccharide database, and performs the plant polysaccharide analysis model according to the recommended extraction method and the actual values ​​of extraction parameters. retraining; since the plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods, it can effectively utilize the data in the plant polysaccharide database for analysis, and fully utilize the data through data mining; since the extraction parameter prediction values ​​output by the plant polysaccharide analysis model are adjusted by the actual values ​​of the extraction parameters of the extraction experiment of the target plant polysaccharides, and the plant polysaccharide database is updated and the plant polysaccharide analysis model is retrained, the accuracy of the extraction parameter prediction is improved, and through the combination of model prediction and extraction experiment, a large number of experiments can be reduced, the experimental cost is reduced, and the extraction efficiency and development efficiency of plant polysaccharides are improved.

[0096] On the basis of the above-mentioned embodiments, the embodiments of the present invention further optimize the data processing method, model design, experiment recommendation and feedback mechanism to improve the accuracy, applicability and efficiency of the system.

[0097] 1. Data enhancement and multi-dimensional feature expansion In terms of data enhancement: introduce ±5% random noise in experimental conditions (such as temperature and time) to simulate experimental errors. Increase diversified samples to improve the robustness of the model to abnormal data. Use SMOTE (SyntheticMinority Oversampling Technique) to generate virtual data points under rare conditions (such as low temperature and long time conditions) to balance data distribution.

[0098] In terms of feature expansion: In addition to basic experimental conditions and chemical features, the following important features are added. Dynamic solvent concentration: simulate the effect of solvent concentration changes on extraction efficiency at different time points. Environmental conditions: external factors such as humidity and air pressure. Polysaccharide morphological characteristics: molecular structure information extracted through spectra (such as infrared spectra or nuclear magnetic resonance spectra).

[0099] 2. Model fusion and automatic tuning In terms of model fusion: multi-level integrated model based on random forest, gradient boosting regression and neural network. First-layer model: independently trained random forest, gradient boosting regression, support vector machine. Second-layer model: use linear regression or neural network to fuse the output of the first layer to generate the final prediction value. The first-layer model is a basic prediction model that learns the independent relationship of different features. The second-layer model is a fusion model that learns the error patterns of different basic models to further optimize the results.

[0100] In terms of automatic tuning: Introduce Bayesian Optimization. Optimization goal: Find a balance between extraction rate and biological activity. Optimization parameters: Number of trees and depth of random forests; learning rate and number of hidden nodes of neural networks; kernel coefficient of support vector machines. Use tools such as Optuna or Hyperopt to dynamically adjust hyperparameter combinations based on cross-validation results.

[0101] 3. Multi-objective optimization of experimental recommendations Optimization goals include maximizing extraction yield, minimizing IC50 value, and balancing resource costs (such as time and solvent usage).

[0102] The Pareto Front algorithm is introduced to generate a series of experimental conditions to ensure the optimal solution in multiple dimensions such as efficiency, activity, and resource consumption.

[0103] Recommendations include: Scheme A: Prioritizes extraction efficiency, with recommended parameters of temperature 70°C, time 45 min, ethanol solvent, and an expected extraction rate of 72%.

[0104] Scheme B: Priority activity, recommended parameters are temperature 55°C, time 60 min, pectinase treatment, and expected IC50 value 60 μg / mL.

[0105] Scheme C: Comprehensive balance, recommended parameters are temperature 65°C, time 50 min, mixed solvent, expected extraction rate 70%, IC50 value 65 μg / mL.

[0106] Based on the multi-objective optimization output, a prioritized list of experimental conditions is generated and presented through dynamic tables or visualization tools.

[0107] Recommended order: Prioritize experiments with the smallest deviation from the target variable.

[0108] Parameter adjustment: Allows users to fine-tune recommendation conditions based on actual needs.

[0109] 4. Intelligent feedback loop In terms of experimental result feedback: through online data collection, the actual experimental results are uploaded to the system in real time: recording the extraction rate, IC50 value and abnormal data during the experiment.

[0110] Comparison of experimental data with model predictions: Deviation analysis: Calculate the relative error between the predicted value and the actual value.

[0111] Data quality assessment: Automatically label noisy or anomalous experimental data.

[0112] In terms of model dynamics during training: Automatically trigger model updates: When the amount of new data exceeds a certain threshold (e.g. 20 groups), the system starts the retraining process. Data weight adjustment: Give new data a higher weight to improve the model's adaptability to the latest experimental scenarios.

[0113] Dynamic optimization goal: optimize the weight distribution of the loss function to make the model more in line with actual experimental needs.

[0114] The technical effects of the above technical solution include: 1. Improve prediction performance Data augmentation and feature expansion significantly improve the model's predictive ability for complex experimental conditions: The extraction rate prediction MSE was reduced from 3.5% to 2.8%. The bioactivity prediction MSE was reduced from 5.2 μg / mL to 4.3 μg / mL. Model fusion made the overall prediction more accurate, and the R2 score was increased to above 0.95.

[0115] 2. Improve experimental efficiency The experimental recommendation conditions generated by multi-objective optimization reduced the number of experiments by more than 50%: The traditional method requires 10 experiments to optimize an experimental condition. The optimal solution recommends that only 4 experiments are needed to achieve the optimization goal.

[0116] 3. Save resources and costs Save solvent usage: Reduce the average solvent usage by 20% by optimizing the solvent concentration. Reduce experimental time: The average experimental time is shortened from 60 minutes to 40 minutes.

[0117] 4. Improve adaptability and robustness After adding polysaccharide characteristics and environmental parameters, the model can adapt to more types of plant extraction tasks. The intelligent feedback mechanism ensures that the model is continuously optimized during data iteration and maintains high efficiency in the long term.

[0118] Figure 3 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 3As shown, the electronic device may include: a processor (Processor) 310, a communication interface (Communications Interface) 320, a memory (Memory) 330 and a communication bus (Communications Bus) 340, wherein the processor 310, the communication interface 320, and the memory 330 communicate with each other through the communication bus 340. The processor 310 may call the logic command in the memory 330 to execute the method described in the above embodiment, for example: Based on a plant polysaccharide database, plant polysaccharide data of a target plant polysaccharide are obtained; the plant polysaccharide data of the target plant polysaccharide are input into a plant polysaccharide analysis model to obtain a recommended extraction method for the target plant polysaccharide and predicted values ​​of extraction parameters corresponding to the recommended extraction method output by the plant polysaccharide analysis model; based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide, the actual values ​​of the extraction parameters of the target plant polysaccharide under the recommended extraction method are determined; the extraction experiment of the target plant polysaccharide is carried out based on the recommended extraction method and the predicted values ​​of extraction parameters; the recommended extraction method and the actual values ​​of extraction parameters are stored in the plant polysaccharide database, and the plant polysaccharide analysis model is retrained based on the recommended extraction method and the actual values ​​of extraction parameters; wherein the plant polysaccharide analysis model is obtained after training based on plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods.

[0119] In addition, the logic commands in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several commands to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0120] The processor in the electronic device provided in the embodiment of the present invention can call the logic instructions in the memory to implement the above method. Its specific implementation method is consistent with the implementation method of the aforementioned method and can achieve the same beneficial effects, which will not be repeated here.

[0121] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method provided in the above embodiments is implemented.

[0122] Its specific implementation is consistent with the aforementioned method implementation and can achieve the same beneficial effects, so it will not be repeated here.

[0123] An embodiment of the present invention provides a computer program product, including a computer program. When the computer program is executed by a processor, the method described above is implemented.

[0124] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0125] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A plant polysaccharide data analysis method, characterized in that: include: Based on the plant polysaccharide database, obtaining plant polysaccharide data of target plant polysaccharides; Inputting the plant polysaccharide data of the target plant polysaccharide into a plant polysaccharide analysis model, and obtaining a recommended extraction method for the target plant polysaccharide and extraction parameter prediction values ​​corresponding to the recommended extraction method output by the plant polysaccharide analysis model; Determine the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide; the extraction experiment of the target plant polysaccharide is carried out based on the recommended extraction method and the predicted value of the extraction parameter; Storing the recommended extraction method and the actual values ​​of the extraction parameters in the plant polysaccharide database, and retraining the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters; The plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods.

2. The plant polysaccharide data analysis method according to claim 1, characterized in that: The plant polysaccharide data include the polysaccharide name, biological activity, molecular structure, plant source, extraction method and extraction parameters under the extraction method; the extraction parameters include extraction temperature, extraction time, enzyme type and extraction pressure.

3. The plant polysaccharide data analysis method according to claim 1, characterized in that: Before inputting the plant polysaccharide data of the target plant polysaccharide into the plant polysaccharide analysis model and obtaining the recommended extraction method of the target plant polysaccharide and the predicted value of the extraction parameter corresponding to the recommended extraction method output by the plant polysaccharide analysis model, the method further comprises: Determining a current plant polysaccharide analysis model among multiple plant polysaccharide analysis models; Wherein, the plurality of plant polysaccharide analysis models include a first plant polysaccharide analysis model, a second plant polysaccharide analysis model and a third plant polysaccharide analysis model; The first plant polysaccharide analysis model is obtained after training based on the first loss function; the second plant polysaccharide analysis model is obtained after training based on the second loss function; the third plant polysaccharide analysis model is obtained after training based on the first loss function and the second loss function; The first loss function is determined based on the error between the predicted value of the extraction rate of each sample plant polysaccharide under different extraction methods and the actual value of the extraction rate; the second loss function is determined based on the error between the predicted value of the biological activity evaluation index of each sample plant polysaccharide under different extraction methods and the actual value of the biological activity evaluation index.

4. The plant polysaccharide data analysis method according to claim 1, characterized in that: The plant polysaccharide analysis model includes a first prediction layer and a second prediction layer; The first prediction layer includes a plurality of prediction sub-models arranged in parallel, and is used to send prediction results of each prediction sub-model to the second prediction layer; the prediction sub-models include a random forest model, a gradient boosting regression model and a support vector machine model; The second prediction layer is used to fuse the prediction results of each prediction sub-model to generate a recommended extraction method for the target plant polysaccharide and predicted values ​​of extraction parameters corresponding to the recommended extraction method; the second prediction layer is a linear regression model or a neural network model.

5. The plant polysaccharide data analysis method according to claim 1, characterized in that: The plant polysaccharide analysis model is trained based on the following method: Add random noise to the actual values ​​of the extraction parameters of each sample of plant polysaccharide under different extraction methods to obtain the enhanced values ​​of the extraction parameters of each sample of plant polysaccharide under different extraction methods; The plant polysaccharide analysis model is trained based on the plant polysaccharide data samples of each sample plant polysaccharide and the extraction parameter enhancement values ​​of each sample plant polysaccharide under different extraction methods.

6. The plant polysaccharide data analysis method according to claim 1, characterized in that: The step of determining the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide comprises: Based on the error between the actual extraction rate and the predicted extraction rate, and / or the error between the actual value of the biological activity evaluation index and the predicted value of the biological activity evaluation index, the predicted value of the extraction parameter under the recommended extraction method is adjusted to obtain the actual value of the extraction parameter under the recommended extraction method.

7. The plant polysaccharide data analysis method according to claim 1, characterized in that: The retraining of the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters comprises: Based on the recommended extraction method of the target plant polysaccharide and the actual values ​​of the extraction parameters corresponding to the recommended extraction method, generating a retraining sample, and adding the retraining sample to the retraining sample set; When the number of samples in the retraining sample set is greater than a preset number, the plant polysaccharide analysis model is retrained based on the retraining sample set.

8. A plant polysaccharide data analysis device, characterized in that: include: An acquisition unit, used for acquiring plant polysaccharide data of a target plant polysaccharide based on a plant polysaccharide database; A prediction unit, used for inputting the plant polysaccharide data of the target plant polysaccharide into a plant polysaccharide analysis model, and obtaining a recommended extraction method of the target plant polysaccharide output by the plant polysaccharide analysis model and a predicted value of an extraction parameter corresponding to the recommended extraction method; An extraction unit, used to determine the actual value of the extraction parameter of the target plant polysaccharide under the recommended extraction method based on the actual value of the extraction rate and / or the actual value of the biological activity evaluation index of the extraction experiment of the target plant polysaccharide; the extraction experiment of the target plant polysaccharide is carried out based on the recommended extraction method and the predicted value of the extraction parameter; An optimization unit, used for storing the recommended extraction method and the actual values ​​of the extraction parameters in the plant polysaccharide database, and retraining the plant polysaccharide analysis model based on the recommended extraction method and the actual values ​​of the extraction parameters; The plant polysaccharide analysis model is obtained after training based on the plant polysaccharide data samples of each sample plant polysaccharide in the plant polysaccharide database and the actual values ​​of extraction parameters of each sample plant polysaccharide under different extraction methods.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the plant polysaccharide data analysis method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the plant polysaccharide data analysis method according to any one of claims 1 to 7 is implemented.

Images