Enzymatic method for chrysalis peptide

Abstract

This invention provides an enzymatic hydrolysis method for silkworm pupa peptides, belonging to the field of enzymatic hydrolysis technology. The method includes: washing and pulverizing silkworm pupae to obtain silkworm pupa powder; selecting enzyme types based on the silkworm pupa protein structure and preparing a corresponding buffer solution by determining the optimal pH range; constructing a reinforcement graph neural network model to predict the optimal material-to-liquid ratio; mixing the silkworm pupa powder and buffer solution according to the optimal material-to-liquid ratio and heating to a preset temperature for enzymatic hydrolysis to obtain an enzymatic hydrolysate; concentrating the treated enzymatic hydrolysate and drying it to produce powdered or granular products for packaging to obtain silkworm pupa peptides. This invention, by utilizing a graph neural network model, the Red-billed Blue Magpie algorithm, and reinforcement learning strategies, can accurately predict the optimal material-to-liquid ratio, improving production efficiency, reducing costs, and ensuring the stability of the enzymatic hydrolysis reaction.

Description

Technical Field

[0001] This invention relates to the field of enzymatic hydrolysis technology, and in particular to an enzymatic hydrolysis method for silkworm pupa peptides. Background Technology

[0002] Silkworm pupae are rich in protein, with a complete and well-balanced amino acid profile, making them a high-quality protein resource. Proteolytically hydrolyzing silkworm pupae into pupa peptides not only improves their protein properties but also broadens their application areas and increases their resource value, showing great promise in the food and feed industries. For example, pupa peptides possess both antibacterial and antioxidant activities and can be used to prepare aquaculture-related products, enhancing the immunity of aquatic animals.

[0003] However, the utilization of silkworm pupa protein currently faces numerous challenges. On one hand, silkworm pupa protein exhibits poor water solubility, foaming properties, emulsifying ability, and water-holding capacity; off-flavors and pigments are often incompletely removed, limiting its application in the food industry and resulting in a significant waste of high-quality protein resources. On the other hand, existing methods for preparing silkworm pupa peptides generally suffer from low degree of hydrolysis, low yield, and low bioactivity. Traditional intermittent enzymatic hydrolysis methods also suffer from low protease utilization, long production cycles, unstable product quality, and high labor intensity, making them unsuitable for industrial production.

[0004] Furthermore, in the preparation of silkworm pupa peptides, the selection of parameters such as the material-to-liquid ratio, enzymatic hydrolysis temperature, and pH value significantly affects the enzymatic hydrolysis effect. However, traditional methods struggle to accurately determine these parameters, leading to difficulties in guaranteeing hydrolysis efficiency and product quality. Therefore, a more efficient and precise enzymatic hydrolysis method for silkworm pupa peptides is needed to improve the utilization rate of silkworm pupa protein and the quality of silkworm pupa peptides, thereby meeting the demands of industrial production and the market. Summary of the Invention

[0005] This invention provides an enzymatic hydrolysis method for silkworm pupa peptides to overcome the deficiencies in the prior art.

[0006] On one hand, the present invention provides an enzymatic hydrolysis method for silkworm pupa peptides, comprising:

[0007] Silkworm pupae are cleaned and crushed to obtain silkworm pupa powder.

[0008] Select enzyme types based on the protein structure of silkworm pupae, and determine the optimal pH range to prepare the corresponding buffer solutions.

[0009] A graph neural network model was constructed, and the Red-billed Blue Magpie algorithm was used to optimize the graph neural network model to obtain an enhanced graph neural network model, which was then used to predict the optimal material-liquid ratio.

[0010] Silkworm pupa powder and buffer solution are mixed at the optimal material-to-liquid ratio and heated to a preset temperature to carry out enzymatic hydrolysis to obtain enzymatic hydrolysate.

[0011] The enzymatic hydrolysate is pretreated by filtration and then posttreated by dehydration and desalination according to the target requirements.

[0012] The processed enzymatic hydrolysate is concentrated and then dried to produce powdered or granular products, which are then packaged to obtain silkworm pupa peptides.

[0013] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the step of determining the optimal pH range includes:

[0014] The development index of silkworm pupa protein was obtained by analyzing its amino acid composition and structural domain characteristics.

[0015] Based on the silkworm pupa protein development index, select the corresponding endonuclease to generate small molecule peptides, and select exonuclease to assist and perform hydrophobic treatment according to specific needs.

[0016] By observing the molecular weight distribution of different enzymatic hydrolysis products through SDS-PAGE electrophoresis, the enzyme with the highest hydrolysis efficiency can be screened to determine the enzyme type.

[0017] By collecting historical data on enzyme decomposition, the pH range of different enzymes in the enzymatic hydrolysis of silkworm pupa protein was obtained. Different pH gradients were set up to conduct single-factor experiments, and the hydrolysis rate and peptide yield of silkworm pupa protein under different pH values ​​were measured. The pH range with the highest characteristic index was selected as the optimal pH range.

[0018] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the step of preparing the corresponding buffer solution includes:

[0019] Select a solution based on the determined optimal pH range, and calculate the required amounts of buffer and acid based on the solution volume and concentration.

[0020] Based on the mass of the substances, weigh the buffer and acid to obtain the required mass.

[0021] Dissolve the required mass in distilled water at the preset mark, and stir until fully dissolved to obtain a buffer solution.

[0022] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the steps for optimizing the obtained enhanced graph neural network model include:

[0023] Using silkworm pupa protein and buffer solution as nodes, edge weights are defined based on the interaction between silkworm pupa protein and buffer solution, and a graph is constructed based on the nodes and edge weights.

[0024] We select a graph convolutional network as the backbone model to obtain a graph neural network model.

[0025] Multiple red-billed blue magpie individuals are randomly generated as the initial population, and each red-billed blue magpie individual represents a set of hyperparameter values ​​of a graph convolutional network.

[0026] Define the fitness function based on the performance metrics of graph convolutional networks.

[0027] Within the perception range of each red-billed blue magpie, a neighboring individual is randomly selected, and the neighboring individual is updated according to the red-billed blue magpie position update formula to obtain the updated position.

[0028] The fitness value of each individual Red-billed Blue Magpie is calculated using the fitness function. The individuals are then sorted from highest to lowest fitness value, and the individuals with the highest fitness value are selected as dominant individuals. For the remaining individuals, one is randomly selected as a candidate individual, and the candidate individual is updated according to the Red-billed Blue Magpie position update formula.

[0029] Calculate the fitness value of the candidate update position, obtain the candidate fitness value, and determine whether the candidate fitness value is less than the individual fitness value. If it is, accept the update position; otherwise, accept the candidate update position.

[0030] After reaching the preset number of iterations, the hyperparameter value represented by the current best red-billed blue magpie individual is output as the optimal hyperparameter value to construct a graph neural network model, thus obtaining the enhanced graph neural network model.

[0031] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the formula for updating the position is expressed as follows:

[0032]

[0033] In the formula, It updates the location. They are individual neighbors. It is a red-billed blue magpie. It is the step scaling factor.

[0034] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the step of predicting the optimal material-liquid ratio includes:

[0035] The material-to-liquid ratio optimization is modeled, with raw material properties and the current material-to-liquid ratio as states, and the discrete adjustment of the material-to-liquid ratio as actions. A reward function is designed based on the predicted product quality.

[0036] Attribute nodes are determined based on the properties of raw materials and liquids, and the weights of the action edges are defined based on the solubility and reaction rates of the raw materials and liquids.

[0037] By incorporating the attribute nodes and the weights of the action edges into the reinforcement graph neural network model, an initial liquid-to-material ratio strategy is randomly generated.

[0038] A reinforced graph neural network model trained using historical enzyme degradation data is used to predict reaction results, expressed by the following formula:

[0039]

[0040] In the formula, These are actual product indicators. It is an enhanced graph neural network model prediction. It is the result of the reaction. It is the number of data points used in loss calculation.

[0041] The reinforcement graph neural network model is updated using the policy gradient method, and the liquid-to-material ratio strategy is iteratively optimized.

[0042] After a preset number of iterations, the liquid-to-liquid ratio strategy with the fastest dissolution rate and reaction rate is selected as the optimal liquid-to-liquid ratio from the liquid-to-liquid ratio strategies.

[0043] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the steps of updating the enhanced graph neural network model and iteratively optimizing the liquid-to-material ratio strategy include:

[0044] The current liquid-to-material ratio strategy is executed in the enhanced graph neural network model to generate trajectory data, aggregate attribute node information, and perform node updates and global feature extraction to obtain the strategy network decision.

[0045] The enzymatic hydrolysis effect corresponding to the policy network decision is used as the reward signal. The policy distribution is output according to the attribute node characteristics, and the cumulative reward of each liquid-to-material ratio policy is calculated.

[0046] Calculate the policy gradient based on the cumulative reward and policy distribution.

[0047] The reinforcement graph neural network model is updated based on the policy gradient, and the input liquid-to-material ratio policy is adjusted to move the policy distribution towards the high-reward region for updating.

[0048] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the strategy gradient is expressed as follows:

[0049]

[0050] In the formula, It is the sampling trajectory. It is the first The state of the step, It is the first Step strategy, It is a policy function. It's a cumulative reward. It is the policy gradient. It is the gradient operator. It is the total length of the trajectory.

[0051] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the steps of performing the enzymatic hydrolysis reaction to obtain the enzymatic hydrolysate include:

[0052] Weigh the silkworm pupa powder and buffer solution according to the optimal material-to-liquid ratio, adjust the pH of the system to the preset state, add the enzyme preparation and stir evenly.

[0053] The preset temperature is determined based on the enzyme's optimal temperature. A magnetic stirrer is used to maintain low-speed stirring. The reaction time is determined based on the enzyme activity and the target degree of hydrolysis.

[0054] The temperature is checked at preset time intervals and adjusted according to the real-time temperature.

[0055] Once the reaction time is reached, place the container in an ice bath to cool to room temperature to terminate enzyme activity. Transfer the liquid to a centrifuge tube and collect the supernatant as the enzyme hydrolysate.

[0056] According to the enzymatic hydrolysis method for silkworm pupa peptides provided by the present invention, the steps of pretreatment and posttreatment of the enzymatic hydrolysate include:

[0057] Using filter paper to filter the enzymatic hydrolysate, centrifugation can be used to remove fine suspended particles.

[0058] Depending on the specific scenario, three methods—vacuum concentration, membrane separation concentration, and freeze drying—are selected for dehydration.

[0059] Depending on the desired outcome, dialysis, ion exchange chromatography, or membrane separation can be used for desalination.

[0060] This invention provides an enzymatic hydrolysis method for silkworm pupa peptides. Through in-depth analysis of the silkworm pupa protein structure and scientific enzyme selection, combined with the determination of the optimal pH range, it can significantly improve the hydrolysis efficiency of silkworm pupa proteins, generating more small molecule peptides, thereby increasing the yield and quality of silkworm pupa peptides and meeting the quality requirements of silkworm pupa peptides in different application scenarios. Furthermore, by utilizing a graph neural network model, the Red-billed Blue Magpie algorithm, and reinforcement learning strategies, the optimal material-to-liquid ratio can be accurately predicted, reducing the number of experiments and resource consumption, improving production efficiency, lowering costs, and ensuring the stability and reproducibility of the enzymatic hydrolysis reaction.

[0061] This invention provides an enzymatic hydrolysis method for silkworm pupa peptides. By modeling the optimal material-to-liquid ratio, defining states, actions, and reward functions, and learning a material-to-liquid ratio adjustment strategy through a reinforcement graph neural network model, the method updates the model using a policy gradient method, shifting the policy distribution towards higher reward regions to obtain the optimal material-to-liquid ratio. The introduction of intelligent techniques such as graph neural networks and reinforcement learning fully utilizes historical data and complex interactions to achieve intelligent modeling, prediction, and optimization of the enzymatic hydrolysis process, improving the automation and intelligence level of production and providing strong technical support for the large-scale production of silkworm pupa peptides. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0063] Figure 1 This is one of the flowcharts illustrating an enzymatic hydrolysis method for silkworm pupa peptides provided in an embodiment of the present invention;

[0064] Figure 2 This is a second schematic flowchart of an enzymatic hydrolysis method for silkworm pupa peptides provided in an embodiment of the present invention;

[0065] Figure 3 This is the third flowchart of an enzymatic hydrolysis method for silkworm pupa peptides provided in this embodiment of the invention. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0067] The following is combined with Figures 1-3 This invention describes an enzymatic hydrolysis method for silkworm pupa peptides.

[0068] like Figure 1 As shown in the embodiment of the present invention, an enzymatic hydrolysis method for silkworm pupa peptides includes:

[0069] Silkworm pupae are cleaned and crushed to obtain silkworm pupa powder.

[0070] Physical cleaning removes impurities (cocoon fragments, dirt, insect remains), microorganisms (bacteria, mold spores), and some odorous substances (volatile fatty acids, aldehydes) from the surface of silkworm pupae, improving the purity and safety of the silkworm pupa powder. The silkworm pupae are then placed on a vibrating screen, and a gas separator removes light impurities. Utilizing the density difference between silkworm pupae and sand / gravel, heavy impurities are separated using water flotation (silkworm pupae float, sand / gravel settle) or a gravity separator.

[0071] Mechanical crushing disrupts the tissue structure (epidermis, protein matrix) of silkworm pupae, releasing proteins and increasing specific surface area (improving subsequent enzymatic hydrolysis efficiency). Simultaneously, powder particle size is controlled to meet different application requirements. For whole silkworm pupae, a shear crusher can be used to break them into suitable particle sizes, followed by low-temperature pulverization using a hammer mill or air jet mill, and finally, a vibrating screen separates out unqualified coarse particles.

[0072] The steps to determine the optimal pH range include:

[0073] The development index of silkworm pupa protein is obtained by analyzing its amino acid composition and domain characteristics. Amino acid composition analysis can be performed using an automated amino acid analyzer to determine the content and ratio of essential amino acids (such as lysine and methionine) and non-essential amino acids in silkworm pupa protein, and to calculate nutritional evaluation indicators such as amino acid score (AAS) and chemical score (CS) to assess the protein's nutritional value.

[0074] Domain characteristics analysis: The secondary structure (proportion of α-helices, β-sheets, and random coils) of silkworm pupa proteins was determined by circular dichroism (CD) and Fourier transform infrared spectroscopy (FT-IR) to analyze its spatial structural stability. The adaptability of the protein to enzymatic hydrolysis was quantitatively evaluated from both amino acid composition and domain characteristics to obtain the silkworm pupa protein development index.

[0075] Based on the silkworm pupa protein development index, select the corresponding endonuclease to generate small molecule peptides, and select exonuclease to assist and perform hydrophobic treatment according to specific needs.

[0076] By observing the molecular weight distribution of different enzymatic hydrolysis products through SDS-PAGE electrophoresis, the enzyme with the highest hydrolysis efficiency can be screened to determine the enzyme type. This process may include: using a 12% separating gel + 5% stacking gel, loading 10 μL (enzyme digest diluted 50 times), maintaining a constant electrophoresis voltage of 80V (stacking gel) → 120V (separating gel), and running for approximately 90 minutes.

[0077] Staining solution: Coomassie Brilliant Blue R-250 staining solution (0.1% methanol: acetic acid: water = 5:1:4), destaining solution: methanol: acetic acid: water = 5:1:4, until the background is clear.

[0078] Molecular weight distribution analysis:

[0079] By comparing the electrophoretic bands of different enzymatic digestion products, we focus on the band intensity in the 1kDa, 1~3kDa, and 3~10kDa ranges and calculate the area ratio of each range.

[0080] Screening criteria: Enzymes with <3kDa peptides accounting for >60% and uniform band distribution were selected. Historical enzyme decomposition data were collected to obtain the pH range of different enzymes when hydrolyzing silkworm pupa protein. Different pH gradients were set to conduct single-factor experiments to determine the hydrolysis rate and peptide yield of silkworm pupa protein at different pH values. The pH range with the highest characteristic index was selected as the optimal pH range.

[0081] The steps for preparing the appropriate buffer solution include:

[0082] Select a solution based on the determined optimal pH range, and calculate the required amounts of buffer and acid based on the solution volume and concentration.

[0083] Based on the mass of the substances, weigh the buffer and acid to obtain the required mass.

[0084] Dissolve the required mass in distilled water at the preset mark, and stir until fully dissolved to obtain a buffer solution.

[0085] A graph neural network model was constructed, and the Red-billed Blue Magpie algorithm was used to optimize the graph neural network model to obtain an enhanced graph neural network model, which was then used to predict the optimal material-liquid ratio.

[0086] The steps to optimize and obtain the enhanced graph neural network model include:

[0087] Using silkworm pupa protein and buffer solution as nodes, edge weights are defined based on the interaction between silkworm pupa protein and buffer solution, and a graph is constructed based on the nodes and edge weights.

[0088] Choosing a graph convolutional network as the backbone model yields a graph neural network model, expressed by the formula:

[0089]

[0090] In the formula, It is an activation function. It is the normalized adjacency matrix. It is the first The node feature matrix of the layer It is a learnable weight matrix. It is the first The node feature matrix of the layer.

[0091] Multiple red-billed blue magpie individuals are randomly generated as the initial population, and each red-billed blue magpie individual represents a set of hyperparameter values ​​of a graph convolutional network.

[0092] The fitness function is defined based on the performance metrics of graph convolutional networks. The fitness function is expressed by the following formula:

[0093]

[0094] In the formula, This is the actual performance difference value. It is based on the individual and related factors The calculated difference in predicted performance, It refers to the number of samples.

[0095] Within the perception range of each red-billed blue magpie, a neighboring individual is randomly selected, and the neighboring individual is updated according to the red-billed blue magpie position update formula to obtain the updated position.

[0096] The formula for updating the position is expressed as follows:

[0097]

[0098] In the formula, It updates the location. They are individual neighbors. It is a red-billed blue magpie. It is the step scaling factor.

[0099] The fitness value of each individual Red-billed Blue Magpie is calculated using the fitness function. The individuals are then sorted from highest to lowest fitness value, and the individuals with the highest fitness value are selected as dominant individuals. For the remaining individuals, one is randomly selected as a candidate individual, and the candidate individual is updated according to the Red-billed Blue Magpie position update formula.

[0100] Calculate the fitness value of the candidate update position, obtain the candidate fitness value, and determine whether the candidate fitness value is less than the individual fitness value. If it is, accept the update position; otherwise, accept the candidate update position.

[0101] After reaching the preset number of iterations, the hyperparameter value represented by the current best red-billed blue magpie individual is output as the optimal hyperparameter value to construct a graph neural network model, thus obtaining the enhanced graph neural network model.

[0102] Silkworm pupa powder and buffer solution are mixed at the optimal material-to-liquid ratio and heated to a preset temperature to carry out enzymatic hydrolysis to obtain enzymatic hydrolysate.

[0103] like Figure 2 As shown, the steps for predicting the optimal feed-liquid ratio include:

[0104] The feed-to-liquid ratio optimization is modeled, with raw material attributes and the current feed-to-liquid ratio as states, and the discrete adjustment of the feed-to-liquid ratio as actions. A reward function is designed based on the predicted product quality, expressed by the formula:

[0105]

[0106] In the formula, It is the degree of hydrolysis. It's the yield. It's the cost of energy consumption. , and These are weighting coefficients. It is a reward function.

[0107] Attribute nodes are determined based on the properties of the raw materials and liquids, and the weights of the action edges are defined based on the solubility and reaction rates of the raw materials and liquids. The formula is expressed as follows:

[0108]

[0109] In the formula, and It is the node feature vector. It is a feature mapping function. It is the weight of the applied edge.

[0110] By incorporating the attribute nodes and the weights of the action edges into the reinforcement graph neural network model, an initial liquid-to-material ratio strategy is randomly generated.

[0111] A reinforced graph neural network model trained using historical enzyme degradation data is used to predict reaction results, expressed by the following formula:

[0112]

[0113] In the formula, These are actual product indicators. It is an enhanced graph neural network model prediction. It is the result of the reaction. It is the number of data points used in loss calculation.

[0114] like Figure 3 As shown, the reinforcement graph neural network model is updated according to the policy gradient method, and the liquid-to-material ratio policy is iteratively optimized.

[0115] The current liquid-to-material ratio strategy is executed in the enhanced graph neural network model to generate trajectory data, aggregate attribute node information, and perform node updates and global feature extraction to obtain the strategy network decision.

[0116] The enzymatic hydrolysis effect corresponding to the policy network decision is used as the reward signal. The policy distribution is output according to the attribute node characteristics, and the cumulative reward of each liquid-to-material ratio policy is calculated.

[0117] Based on the cumulative reward and policy distribution, the policy gradient is calculated using the following formula:

[0118]

[0119] In the formula, It is the sampling trajectory. It is the first The state of the step, It is the first Step strategy, It is a policy function. It's a cumulative reward. It is the policy gradient. It is the gradient operator. It is the total length of the trajectory.

[0120] The reinforcement graph neural network model is updated based on the policy gradient, and the input liquid-to-material ratio policy is adjusted to move the policy distribution towards the high-reward region for updating.

[0121] After a preset number of iterations, the liquid-to-liquid ratio strategy with the fastest dissolution rate and reaction rate is selected as the optimal liquid-to-liquid ratio from the liquid-to-liquid ratio strategies.

[0122] The steps for obtaining the enzymatic hydrolysate through enzymatic hydrolysis include:

[0123] Weigh the silkworm pupa powder and buffer solution according to the optimal material-to-liquid ratio, adjust the pH of the system to the preset state, add the enzyme preparation and stir evenly.

[0124] The preset temperature is determined based on the enzyme's optimal temperature. A magnetic stirrer is used to maintain low-speed stirring. The reaction time is determined based on the enzyme activity and the target degree of hydrolysis.

[0125] The temperature is checked at preset time intervals and adjusted according to the real-time temperature.

[0126] Once the reaction time is reached, place the container in an ice bath to cool to room temperature to terminate enzyme activity. Transfer the liquid to a centrifuge tube and collect the supernatant as the enzyme hydrolysate.

[0127] The enzymatic hydrolysate is pretreated by filtration and then posttreated by dehydration and desalination according to the target requirements.

[0128] Using filter paper to filter the enzymatic hydrolysate, centrifugation can be used to remove fine suspended particles.

[0129] Depending on the specific scenario, three methods—vacuum concentration, membrane separation concentration, and freeze drying—are selected for dehydration.

[0130] Vacuum concentration principle: Lowering the boiling point of the solvent under negative pressure to avoid high temperature damaging the activity of the product.

[0131] Operating steps:

[0132] Transfer the filtrate to the distillation flask of the rotary evaporator, adding no more than 2 / 3 of the flask's volume.

[0133] Turn on the vacuum pump, adjust the vacuum level to 0.08~0.1MPa, and set the water bath temperature (40~60℃, adjust according to the thermal stability of the product).

[0134] Rotary distillation flasks heat the liquid evenly until it is concentrated to the target concentration.

[0135] After concentration is complete, slowly release the vacuum and remove the concentrate.

[0136] Applicable scenarios: Concentration of heat-sensitive products (such as bioactive peptides and amino acids).

[0137] 2. Membrane separation and concentration principle: The solvent and solute are separated by the molecular weight cutoff of the membrane, while the target product is retained.

[0138] Operating steps:

[0139] Select a membrane with a suitable molecular weight cutoff (e.g., an ultrafiltration membrane with a molecular weight cutoff of 1000~10000 Da) and install it into the membrane separation equipment.

[0140] The filtrate is pumped into the membrane module, and pressure (0.1~0.5MPa) is applied to allow the solvent (water) to permeate through the membrane, while the target product (such as a peptide) is retained.

[0141] The solution is circulated and concentrated to the required concentration. A small amount of deionized water can be added during the process (for rinsing) to improve the recovery rate.

[0142] Advantages: Low-temperature operation, no phase change, low energy consumption, suitable for heat-sensitive substances.

[0143] 3. Freeze-drying principle: First, the solution is frozen into a solid state, and then the water is removed by sublimation under vacuum to obtain freeze-dried powder.

[0144] Operating steps:

[0145] The filtrate was dispensed into lyophilization bottles and pre-frozen to -40 to -20°C to form solid samples.

[0146] Place the contents into the freeze dryer chamber, turn on the vacuum pump, adjust the vacuum level to 10~50Pa, and raise the temperature to 20~40℃ to sublimate the water.

[0147] The freeze-drying time is usually 12 to 48 hours until the sample reaches a constant weight.

[0148] Applicable scenarios: Dehydration of high-value-added, easily oxidized or heat-sensitive products, with the product being a loose powder.

[0149] Depending on the desired outcome, dialysis, ion exchange chromatography, or membrane separation can be used for desalination.

[0150] The principle of dialysis is to use a semipermeable membrane to retain large molecules (such as peptides) while allowing small salt molecules to pass through.

[0151] Operating steps:

[0152] Select a dialysis bag with a suitable molecular weight cutoff (e.g., a molecular weight cutoff of 3500 Da) and activate it by boiling it in distilled water for 10 minutes.

[0153] Fill the dialysis bag with the concentrate, leaving 1 / 3 space to prevent expansion and breakage. After sealing, place it in a container filled with deionized water.

[0154] Stir the dialysate magnetically and change the dialysate every 4 to 6 hours (3 to 4 times) until the conductivity drops to the target value.

[0155] Remove the dialysis bag and collect the desalted solution by squeezing or centrifugation.

[0156] 2. Principle of ion exchange chromatography: Ion exchange resin adsorbs salt ions and releases the target product.

[0157] Operating steps:

[0158] Select the appropriate type of resin (such as cation exchange resin to remove cation salts, and anion exchange resin to remove anion salts), pretreat it, and then pack it into the chromatography column.

[0159] The enzymatic hydrolysate is slowly passed through the chromatography column, and the flow rate is controlled (e.g., 1-2 mL / min) to allow the salt ions to bind with the resin.

[0160] Elute the column with deionized water or low-salt buffer, collect the eluent until no salt is detected (e.g., silver nitrate titration for Cl⁻ detection).

[0161] 3. Membrane separation method. Principle: Nanofiltration membranes have a molecular weight cutoff of 200~1000 Da, which can separate small molecule salts and medium molecular weight products (such as oligopeptides).

[0162] Operating steps:

[0163] Select a nanofiltration membrane (which retains NaCl but allows water and small molecules to pass through) and pump the enzymatic hydrolysate into the membrane device.

[0164] Applying pressure (0.5~1.5MPa) allows salt ions and water to permeate through the membrane, while the target product (such as a polypeptide) is retained.

[0165] The filtrate is circulated until its conductivity meets the requirements, and the retentate is the concentrated solution after desalination.

[0166] The processed enzymatic hydrolysate is concentrated and then dried to produce powdered or granular products, which are then packaged to obtain silkworm pupa peptides.

[0167] To ensure the solid content (total solute mass fraction) of the concentrate meets the requirements of the drying equipment, avoiding low drying efficiency or poor product quality, the drying method can be spray drying or freeze drying. The volume of the concentrate is adjusted by vacuum concentration or membrane concentration, and the solid content is measured using a refractometer or gravimetric method until the target is met.

[0168] For industrial applications, spray drying can be selected. The principle is to atomize the concentrated liquid into fine droplets, which are then instantly dried into powder upon contact with hot air.

[0169] Operating steps:

[0170] Equipment preheating: Turn on the spray dryer and set the inlet air temperature to 150~180℃ and the exhaust air temperature to 70~90℃ (adjust according to the thermal stability of the product; for silkworm pupa peptides, set the inlet air temperature to 160℃ and the exhaust air temperature to 80℃).

[0171] Feed atomization: Pour the concentrate into the feed tank and deliver it to the atomizer (centrifugal or pressure type) at a flow rate of 30~50mL / min using a peristaltic pump. The atomized particle size is controlled at 50~100μm.

[0172] Drying and collection: The atomized droplets come into countercurrent contact with hot air inside the drying tower. The drying time is about 5 to 10 seconds. The powder is collected from the bottom outlet, and the fine powder is recovered through a cyclone separator.

[0173] For highly active products, freeze drying can be selected. The principle is to first freeze the concentrate into a solid state, and then sublimate and dehydrate it under vacuum.

[0174] Operating steps:

[0175] Pre-freezing: Dispense the concentrate into freeze-drying bottles (thickness ≤1cm), place them in the pre-freezing chamber of the freeze dryer, and freeze at -40℃ for 2~4 hours until completely solidified.

[0176] Vacuum drying: Turn on the vacuum pump, adjust the vacuum to 10~30Pa, raise the temperature to 20~30℃ (raise the temperature slowly to avoid spraying the bottle), and maintain for 24~48 hours until constant weight.

[0177] Crushing and sieving: The freeze-dried product is in loose block form. It is crushed with a mortar and pestle or a grinder and then passed through an 80-120 mesh sieve to obtain powder.

[0178] In the laboratory, a vacuum oven can be used for drying. The principle is to heat and evaporate moisture under negative pressure.

[0179] Operating steps:

[0180] Pour the concentrate into a shallow dish (thickness ≤ 0.5 cm), place it in a vacuum oven, set the temperature to 50~60℃ and the vacuum degree to -0.08 MPa.

[0181] Dry for 12-24 hours until completely cured, then grind in a mortar and pestle and pass through a 60-100 mesh sieve.

[0182] This embodiment provides an enzymatic hydrolysis method for silkworm pupa peptides. It determines the silkworm pupa protein development index through amino acid composition and domain analysis, and targets the selection of combinations of endonucleases and exonucleases (e.g., adding exonucleases based on hydrophobicity requirements), avoiding a "one-size-fits-all" approach to enzyme selection. Furthermore, it utilizes the Red-beaked Blue Magpie algorithm to optimize and construct a reinforcement graph neural network model, treating the material-to-liquid ratio as a reinforcement learning problem. It iterative optimization is performed using states (raw material properties, current ratio), actions (adjusting the ratio), and rewards (product quality) to achieve a globally optimal solution search. A graph convolutional network is used to capture the interaction between raw materials and solvent, enhancing the model's ability to represent complex relationships. Magnetic stirring maintains low-speed, uniform mixing, and temperature fluctuations are monitored in real time to ensure the enzyme remains within its optimal activity range. The reaction time is determined dynamically based on enzyme activity and the target degree of hydrolysis, rather than a fixed time, improving product yield. This significantly improves the efficiency and accuracy of enzymatic hydrolysis and material-to-liquid ratio optimization, increases product quality, reduces energy consumption and cost, and optimizes resource utilization.

[0183] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0184] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0185] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for enzymatic hydrolysis of silkworm pupa peptides, characterized in that, include: Silkworm pupae are cleaned and crushed to obtain silkworm pupa powder; Select enzyme types based on the protein structure of silkworm pupae, and determine the optimal pH range to prepare the corresponding buffer solutions; A graph neural network model is constructed, and the graph neural network model is optimized using the Red-billed Blue Magpie algorithm to obtain an enhanced graph neural network model, which is used to predict the optimal material-liquid ratio. The silkworm pupa powder and the buffer solution are mixed according to the optimal material-liquid ratio, and heated to a preset temperature to carry out an enzymatic hydrolysis reaction to obtain an enzymatic hydrolysate; The enzymatic hydrolysate is pretreated by filtration and then posttreated by dehydration and desalination according to the target requirements. The processed enzymatic hydrolysate is concentrated and then dried to produce powdered or granular products, which are then packaged to obtain silkworm pupa peptides.

2. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 1, characterized in that, The steps for determining the optimal pH range include: The development index of silkworm pupa protein was obtained by analyzing its amino acid composition and structural domain characteristics. Based on the silkworm pupa protein development index, select the corresponding endonuclease to generate small molecule peptides, and select exonuclease to assist and perform hydrophobic treatment according to specific needs; The molecular weight distribution of different enzymatic hydrolysis products was observed by SDS-PAGE electrophoresis to screen for enzymes with the highest hydrolysis efficiency, thereby determining the enzyme type. By collecting historical data on enzyme decomposition, the pH range of different enzymes in the enzymatic hydrolysis of silkworm pupa protein was obtained. Different pH gradients were set, and single-factor experiments were conducted to determine the hydrolysis rate and peptide yield of silkworm pupa protein under different pH values. The pH range with the highest value of the characteristic index was selected as the optimal pH range.

3. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 1, characterized in that, The steps for preparing the appropriate buffer solution include: Select a solution based on the determined optimal pH range, and calculate the required amounts of buffer and acid based on the solution volume and concentration. Based on the mass of the substances, weigh the buffer and acid to obtain the required mass; Dissolve the required mass in distilled water at a preset volume and stir until fully dissolved to obtain the buffer solution.

4. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 1, characterized in that, The steps to optimize and obtain the enhanced graph neural network model include: Using the silkworm pupa protein and the buffer solution as nodes, edge weights are defined based on the interaction between the silkworm pupa protein and the buffer solution, and a graph is constructed based on the nodes and the edge weights. We select a graph convolutional network as the backbone model to obtain a graph neural network model; Multiple red-billed blue magpie individuals are randomly generated as the initial population, and each red-billed blue magpie individual represents a set of hyperparameter values ​​of a graph convolutional network. Define the fitness function based on the performance metrics of the graph convolutional network; Within the perception range of each individual red-billed blue magpie, a neighboring individual is randomly selected, and the neighboring individual is updated according to the red-billed blue magpie position update formula to obtain the updated position; The fitness value of each individual Red-billed Blue Magpie is calculated according to the fitness function. The fitness values ​​are sorted from high to low. Individuals with high fitness values ​​are selected as dominant individuals. For the remaining individuals, one is randomly selected as a candidate individual. The candidate individual is updated according to the Red-billed Blue Magpie position update formula. Calculate the fitness value of the candidate update position to obtain the candidate fitness value, and determine whether the candidate fitness value is less than the individual fitness value. If it is, accept the update position; otherwise, accept the candidate update position. After reaching the preset number of iterations, the current optimal hyperparameter value represented by the red-billed blue magpie individual is output as the optimal hyperparameter value to construct the graph neural network model, thus obtaining the enhanced graph neural network model.

5. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 4, characterized in that, The formula for updating the position is expressed as follows: ； In the formula, It updates the location. They are individual neighbors. It is a red-billed blue magpie. It is the step scaling factor.

6. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 2, characterized in that, The step of predicting the optimal feed-liquid ratio includes: Model the liquid-to-material ratio optimization, taking the raw material properties and the current liquid-to-material ratio as the state, and the discrete adjustment of the liquid-to-material ratio as the action, and designing a reward function based on the predicted product quality. Attribute nodes are determined based on the properties of raw materials and liquids, and the weights of the action edges are defined based on the solubility and reaction rates of the raw materials and liquids. The attribute nodes and the weights of the active edges are fed into the enhanced graph neural network model to randomly generate an initial liquid-to-material ratio strategy. The enhanced graph neural network model is trained using the historical data of enzyme degradation to predict the reaction result, expressed by the following formula: ； In the formula, These are actual product indicators. It is an enhanced graph neural network model prediction. This is the result of the reaction, where N is the number of data points involved in the loss calculation; The enhanced graph neural network model is updated according to the policy gradient method, and the liquid-to-material ratio strategy is iteratively optimized. After a preset number of iterations, the liquid-to-liquid ratio strategy with the fastest dissolution rate and reaction rate is selected as the optimal liquid-to-liquid ratio from the liquid-to-liquid ratio strategies.

7. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 6, characterized in that, The steps of updating the enhanced graph neural network model and iteratively optimizing the liquid-to-material ratio strategy include: The current liquid-to-material ratio strategy is executed in the enhanced graph neural network model to generate trajectory data, aggregate the attribute node information, and perform node updates and global feature extraction to obtain the strategy network decision. The enzymatic hydrolysis effect corresponding to the strategy network decision is used as the reward signal. The strategy distribution is output according to the attribute node characteristics, and the cumulative reward for each liquid-to-material ratio strategy is calculated. Calculate the policy gradient based on the cumulative reward and the policy distribution; The reinforcement graph neural network model is updated based on the policy gradient, and the liquid-to-material ratio policy is adjusted to move the policy distribution towards the high-reward region for updating.

8. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 7, characterized in that, The formula for the policy gradient is expressed as follows: ； In the formula, It is the sampling trajectory. This is the state at step t. This is the strategy for step t. It is a policy function. It's a cumulative reward. It is the policy gradient. It is the gradient operator, and T is the total length of the trajectory.

9. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 1, characterized in that, The steps for obtaining the enzymatic hydrolysate by performing an enzymatic hydrolysis reaction include: Weigh the silkworm pupa powder and the buffer solution according to the optimal material-liquid ratio, adjust the pH value of the system to the preset state, add the enzyme preparation and stir evenly; Determine the preset temperature based on the enzyme's optimal temperature, maintain low-speed stirring with a magnetic stirrer, and determine the reaction time based on enzyme activity and target degree of hydrolysis. Check the temperature at preset time intervals and adjust accordingly based on the real-time temperature; Once the reaction time is reached, place the mixture in an ice bath to cool to room temperature to terminate enzyme activity. Transfer the liquid to a centrifuge tube and collect the supernatant as the enzyme hydrolysate.

10. The enzymatic hydrolysis method for silkworm pupa peptides according to claim 1, characterized in that, The steps for pretreatment and posttreatment of the enzymatic hydrolysate include: The enzymatic hydrolysate can be filtered using filter paper, and fine suspended particles can be removed by centrifugal filtration. Depending on the specific scenario, three methods—vacuum concentration, membrane separation concentration, and freeze drying—are selected for dehydration. Depending on the desired outcome, dialysis, ion exchange chromatography, or membrane separation can be used for desalination.

Images