A method and system for predicting antimicrobial peptides in marine biofilms based on data analysis

Abstract

The present invention relates to the field of antimicrobial peptide prediction and discloses a method and system for predicting marine biofilm antimicrobial peptides based on data analysis. The method comprises the following steps: S1, obtaining relevant features of antimicrobial peptides in marine biofilms by a joint extraction method of sequence features and structural features; S2, extracting sequence features and structural features based on a trained neural network model combining a convolutional neural network and a recurrent neural network; S3, respectively calculating a first activity evaluation coefficient and a second activity evaluation coefficient based on the sequence features and structural features extracted in S2, and constructing an activity prediction model for marine biofilm antimicrobial peptides using the first activity evaluation coefficient and the second activity evaluation coefficient; and S4, predicting whether there is an antimicrobial peptide with qualified activity based on a comparison result of an activity prediction value output by the marine biofilm antimicrobial peptide activity prediction model with a preset activity threshold.

Description

Technical Field

[0001] The present invention relates to the field of antimicrobial peptide prediction, and in particular to a method and system for predicting marine biofilm antimicrobial peptides based on data analysis. Background Art

[0002] With the advancement of science and technology, research on marine biological resources has deepened. Scientists have discovered a large number of peptides with antimicrobial activity in the membranes of marine microorganisms. These antimicrobial peptides, due to their unique bioactivity and ability to inhibit a wide range of pathogens, have shown great potential for application in the medical field. However, the diversity and complexity of marine life make traditional experimental screening methods inefficient for the discovery of antimicrobial peptides, making them unable to meet the urgent needs of current research.

[0003] Although some predictive models have been used to screen antimicrobial peptides, most of these models are limited to linear analysis of antimicrobial peptide sequences and fail to delve into the complex relationship between antimicrobial peptides and the structure of marine biofilms. As an important component of marine organisms, the structural characteristics of the biofilm have a profound impact on the function of antimicrobial peptides.

[0004] Therefore, how to accurately predict whether there are qualified active antimicrobial peptides in marine biofilms has become a hot topic and difficulty in current research. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for predicting antimicrobial peptides in marine biofilms based on data analysis to solve the above technical problems.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A method for predicting antimicrobial peptides in marine biofilms based on data analysis comprises the following steps:

[0008] S1. Obtain relevant features of antimicrobial peptides in marine biofilms through a combined extraction method of sequence features and structural features;

[0009] S2. Extracting sequence features and structural features based on a trained neural network model combining a convolutional neural network and a recurrent neural network. The sequence features include: the number and ratio of various amino acids in the antimicrobial peptide, the order of amino acids in the antimicrobial peptide, the number and distribution of positively and negatively charged amino acids in the antimicrobial peptide, and the hydrophobicity of the entire peptide chain of the antimicrobial peptide; the structural features include secondary structure;

[0010] S3. Based on the sequence features and structural features extracted in S2, respectively, calculating a first activity evaluation coefficient and a second activity evaluation coefficient; and constructing an activity prediction model for the marine biofilm antimicrobial peptide using the first activity evaluation coefficient and the second activity evaluation coefficient;

[0011] S4. Compare the activity prediction value output by the activity prediction model of the marine biofilm antimicrobial peptide with a pre-set activity threshold; if the activity prediction value exceeds the pre-set activity threshold, it is predicted that the marine biofilm contains antimicrobial peptides with qualified activity; otherwise, it is determined that the marine biofilm does not contain antimicrobial peptides with qualified activity.

[0012] As a further technical solution, the process of obtaining the first activity evaluation coefficient is:

[0013] By formula:

[0014]

[0015] The first activity evaluation coefficient AC1 is calculated;

[0016] P i is the ratio of the i-th amino acid in the antimicrobial peptide, ω i is the weight factor, n is the total number of amino acids, Δδ is the reference value, S z is the charge distribution index, S H is the hydrophobicity index; ρ1, ρ2, ρ3 are the exponential coefficients.

[0017] As a further technical solution, the charge distribution index S z The way to obtain is:

[0018] By formula group:

[0019] and The charge distribution index S is calculated z ;

[0020] Where Z is the charge distribution coefficient, Z min 、Z max are the maximum and minimum values ​​of the charge distribution coefficient, m + is the number of positively charged amino acids, m - is the number of negatively charged amino acids, q j+ is the charge of the jth positively charged amino acid, q j- is the charge of the jth negatively charged amino acid.

[0021] As a further technical solution, the hydrophobicity index S is obtained. H The process is:

[0022] By formula:

[0023] and Calculate the hydrophobicity index S H ;

[0024] Where H is the hydrophobicity coefficient, H max 、H min are the maximum and minimum values ​​of the hydrophobicity coefficient, is the weight coefficient, h i is the hydrophobicity value of the ith amino acid.

[0025] As a further technical solution, the expression of the activity prediction model of marine biofilm antimicrobial peptides is:

[0026] texAC=AC1*μ1+AC2*μ2;

[0027] Among them, texAC is the activity prediction value, μ1 and μ2 are the preset proportional coefficients;

[0028] Among them, N a is the number of amino acids in the a-helix structure, n is the total number of amino acids, N a0 is the standard value;

[0029] As a further technical solution, the method further includes S5, verifying the result of predicting the presence of qualified active antimicrobial peptides in the marine biofilm;

[0030] If the verification result is true, the activity prediction value is output normally;

[0031] If the verification result is false, the output of the activity prediction value is stopped and the activity prediction model of the marine biofilm antimicrobial peptide is optimized.

[0032] As a further technical solution, the process of verifying the results of predicting the presence of qualified active antimicrobial peptides in the marine biofilm is as follows:

[0033] Get the IC50 value of the current antimicrobial peptide and the concentration of the antimicrobial peptide C onc And the diameter K of the antibacterial circle one ;

[0034] Substituting into the formula:

[0035]

[0036] Calculate the evaluation index S x ;

[0037] Among them, σ1, σ2, σ3 are preset proportional coefficients, IC50 max is the maximum value of antimicrobial peptide, C oncth is the standard concentration value of antimicrobial peptide, K oneth is the standard value of the diameter of the antibacterial circle;

[0038] The calculated evaluation index S xAnd the evaluation index warning value S x0 Make comparisons;

[0039] If S x ≥S x0 , the verification result is true;

[0040] If S x x0 , the verification result is false.

[0041] A marine biofilm antimicrobial peptide prediction system based on data analysis, comprising:

[0042] The data extraction module obtains the relevant features of antimicrobial peptides in marine biofilms through the combined extraction of sequence features and structural features;

[0043] The feature acquisition module extracts sequence features and structural features based on a neural network model that combines a trained convolutional neural network with a recurrent neural network. The sequence features include: the number and ratio of various amino acids in the antimicrobial peptide, the order of amino acids in the antimicrobial peptide, the number and distribution of positively and negatively charged amino acids in the antimicrobial peptide, and the hydrophobicity of the entire peptide chain; the structural features include secondary structure;

[0044] A prediction model construction module calculates a first activity evaluation coefficient and a second activity evaluation coefficient based on the extracted sequence features and structural features, and constructs an activity prediction model for marine biofilm antimicrobial peptides using the first activity evaluation coefficient and the second activity evaluation coefficient;

[0045] The prediction and analysis module compares the activity prediction value output by the activity prediction model of the marine biofilm antimicrobial peptide with a pre-set activity threshold; if the activity prediction value exceeds the pre-set activity threshold, it is predicted that the marine biofilm contains antimicrobial peptides with qualified activity; otherwise, it is determined that the marine biofilm does not contain antimicrobial peptides with qualified activity.

[0046] Beneficial effects of the present invention:

[0047] The present invention jointly extracts sequence and structural features and utilizes a neural network model combining a convolutional neural network (CNN) and a recurrent neural network (RNN) for feature extraction and activity assessment. This method can more comprehensively capture key information about antimicrobial peptides, thereby improving the accuracy of predicting antimicrobial peptide activity. Furthermore, compared with traditional experimental methods, this data-driven prediction method can significantly shorten the prediction cycle and improve prediction efficiency. Through data analysis and machine learning techniques, the activity of antimicrobial peptides can be more accurately predicted, thereby enhancing the ability to discover antimicrobial peptides in marine biofilms, providing new ideas and methods for new drug development, food safety, agricultural disease control, and other fields. BRIEF DESCRIPTION OF THE DRAWINGS​

[0048] The present invention will be further described below with reference to the accompanying drawings.

[0049] Figure 1 A diagram showing the steps of the method of the present invention;

[0050] Figure 2 This is a system structure diagram of the present invention. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making any creative efforts shall fall within the scope of protection of the present invention.

[0052] See also Figure 1-Figure 2 As shown, the present invention is a method for predicting antimicrobial peptides in marine biofilms based on data analysis, comprising the following steps:

[0053] S1. Obtain relevant features of antimicrobial peptides in marine biofilms through a combined extraction method of sequence features and structural features;

[0054] S2. Extracting sequence features and structural features based on a trained neural network model combining a convolutional neural network and a recurrent neural network. The sequence features include: the number and ratio of various amino acids in the antimicrobial peptide, the order of amino acids in the antimicrobial peptide, the number and distribution of positively and negatively charged amino acids in the antimicrobial peptide, and the hydrophobicity of the entire peptide chain of the antimicrobial peptide; the structural features include secondary structure;

[0055] S3. Based on the sequence features and structural features extracted in S2, respectively, calculating a first activity evaluation coefficient and a second activity evaluation coefficient; and constructing an activity prediction model for the marine biofilm antimicrobial peptide using the first activity evaluation coefficient and the second activity evaluation coefficient;

[0056] S4. Compare the activity prediction value output by the activity prediction model of the marine biofilm antimicrobial peptide with a pre-set activity threshold; if the activity prediction value exceeds the pre-set activity threshold, it is predicted that the marine biofilm contains antimicrobial peptides with qualified activity; otherwise, it is determined that the marine biofilm does not contain antimicrobial peptides with qualified activity.

[0057] In this implementation, first, a combined extraction method is used to obtain the sequence characteristics and structural characteristics of antimicrobial peptides; these characteristics include the number and ratio of various amino acids in the antimicrobial peptides, the order of amino acids, the number and distribution of positive and negative charged amino acids, the hydrophobicity of the peptide chain, and the secondary structure, which can fully reflect the biological properties and potential activity of the antimicrobial peptides; then, a neural network model combining a trained convolutional neural network (CNN) and a recurrent neural network (RNN) is used to further process and extract the extracted sequence characteristics and structural characteristics, which can dig out deeper features related to the activity of the antimicrobial peptides from the original features. The information is used to train a model so that it can accurately identify and predict the activity of antimicrobial peptides. Subsequently, based on the extracted features and the trained model, a first activity evaluation coefficient and a second activity evaluation coefficient are calculated. These two coefficients respectively reflect the effects of the sequence characteristics and structural characteristics of the antimicrobial peptide on its activity. By constructing an activity prediction model, these two coefficients are combined to obtain the activity prediction value of the antimicrobial peptide. Finally, the activity prediction value is compared with a pre-set activity threshold. If the predicted value exceeds the threshold, it is determined that the marine biofilm contains an antimicrobial peptide with qualified activity; otherwise, it is determined that the antimicrobial peptide with qualified activity does not exist.

[0058] By jointly extracting sequence and structural features in the aforementioned technical solution and utilizing a neural network model combining a convolutional neural network (CNN) and a recurrent neural network (RNN) for feature extraction and activity assessment, the present invention can more comprehensively capture key information about antimicrobial peptides, thereby improving the accuracy of predicting antimicrobial peptide activity. Furthermore, compared to traditional experimental methods, this data-driven prediction approach can significantly shorten the prediction cycle and improve prediction efficiency. Through data analysis and machine learning techniques, it can more accurately predict the activity of antimicrobial peptides, thereby enhancing the ability to discover antimicrobial peptides in marine biofilms and providing new insights and approaches for new drug development, food safety, agricultural disease control, and other fields.

[0059] The process of obtaining the first activity evaluation coefficient is:

[0060] By formula:

[0061]

[0062] The first activity evaluation coefficient AC1 is calculated;

[0063] P i is the ratio of the i-th amino acid in the antimicrobial peptide, ω i is the weight factor, n is the total number of amino acids, Δδ is the reference value, S z is the charge distribution index, S His the hydrophobicity index; ρ1, ρ2, and ρ3 are exponential coefficients, all of which are equal to 1 by default. The contribution of each actual parameter item to the first activity evaluation coefficient can be dynamically adjusted.

[0064] The charge distribution index S z The way to obtain is:

[0065] By formula group:

[0066] and The charge distribution index S is calculated z ;

[0067] Where Z is the charge distribution coefficient, Z min 、Z max are the maximum and minimum values ​​of the charge distribution coefficient, m + is the number of positively charged amino acids, m - is the number of negatively charged amino acids, q j+ is the charge of the jth positively charged amino acid, q j- is the charge of the jth negatively charged amino acid, where the charge can be assigned based on experimental data or historical data.

[0068] Obtain the hydrophobicity index S H The process is:

[0069] By formula:

[0070] and Calculate the hydrophobicity index S H ;

[0071] Where H is the hydrophobicity coefficient, H max 、H min are the maximum and minimum values ​​of the hydrophobicity coefficient, is the weight coefficient, which is determined based on historical data, h i is the hydrophobicity value of the ith amino acid.

[0072] In this embodiment, a method for obtaining the first activity evaluation coefficient AC1 is provided. Specifically, the hydrophobicity index S is first obtained. H , charge distribution index S z And the ratio of various amino acids in antimicrobial peptides, and then substitute it into the formula The first activity evaluation coefficient AC1 was calculated. Obviously, if the ratio between the ratio of amino acids and the reference value is larger, it means that the ratio of amino acids related to the activity of antimicrobial peptides is larger, and therefore the activity of antimicrobial peptides in the marine biofilm is stronger. H , charge distribution index Sz The larger the value is, the stronger the activity of the antimicrobial peptide is. Therefore, a more accurate prediction model can be formed by jointly communicating the above multiple parameters in time series.

[0073] The expression of the activity prediction model of marine biofilm antimicrobial peptides is:

[0074] texAC=AC1*μ1+AC2*μ2;

[0075] Where texAC is the activity prediction value, μ1 and μ2 are preset proportional coefficients, and AC2 is the first activity evaluation coefficient;

[0076] Among them, N a is the number of amino acids in the a-helix structure, n is the total number of amino acids, N a0 is the standard value; obviously, the smaller the difference between the number of amino acids in the a-helix structure and the standard value, the closer it is to the standard value, and therefore the stronger the predicted activity;

[0077] This example provides a specific expression for the prediction model. The expression texAC = AC1*μ1 + AC2*μ2 indicates that a larger first activity evaluation coefficient and a larger second activity evaluation coefficient indicate a larger activity prediction value. Therefore, a more accurate antimicrobial peptide activity prediction model can be established by integrating sequence and structural features in this manner.

[0078] The method further comprises S5, verifying the result of predicting the presence of qualified active antimicrobial peptides in the marine biofilm;

[0079] If the verification result is true, the activity prediction value is output normally;

[0080] If the verification result is false, the output of the activity prediction value is stopped and the activity prediction model of the marine biofilm antimicrobial peptide is optimized.

[0081] The process for validating the results of the prediction of the presence of qualified active antimicrobial peptides in the marine biofilm was as follows:

[0082] Get the IC50 value of the current antimicrobial peptide and the concentration of the antimicrobial peptide C onc And the diameter K of the antibacterial circle one ;

[0083] Substituting into the formula:

[0084]

[0085] Calculate the evaluation index S x ;

[0086] Among them, σ1, σ2, σ3 are preset proportional coefficients, IC50max is the maximum value of antimicrobial peptide, C oncth is the standard concentration value of antimicrobial peptide, K oneth is the standard value of the diameter of the antibacterial circle;

[0087] The calculated evaluation index S x And the evaluation index warning value S x0 Make comparisons;

[0088] If S x ≥S x0 , the verification result is true;

[0089] If S x x0 , the verification result is false.

[0090] In this embodiment, a verification step is added after outputting the activity prediction value, so that the formula Calculate the evaluation index S x , by calculating the evaluation index S x And the evaluation index warning value S x0 For comparison, if S x ≥S x0 , then the verification result is true; if S x x0 , then the verification result is false, inaccurate activity prediction values ​​are eliminated, and the activity prediction values ​​are output after the corresponding model optimization, so that the final activity prediction results are more accurate and the errors caused by inaccurate activity prediction results are avoided.

[0091] A marine biofilm antimicrobial peptide prediction system based on data analysis, comprising:

[0092] The data extraction module obtains the relevant features of antimicrobial peptides in marine biofilms through the combined extraction of sequence features and structural features;

[0093] The feature acquisition module extracts sequence features and structural features based on a neural network model that combines a trained convolutional neural network with a recurrent neural network. The sequence features include: the number and ratio of various amino acids in the antimicrobial peptide, the order of amino acids in the antimicrobial peptide, the number and distribution of positively and negatively charged amino acids in the antimicrobial peptide, and the hydrophobicity of the entire peptide chain; the structural features include secondary structure;

[0094] A prediction model construction module calculates a first activity evaluation coefficient and a second activity evaluation coefficient based on the extracted sequence features and structural features, and constructs an activity prediction model for marine biofilm antimicrobial peptides using the first activity evaluation coefficient and the second activity evaluation coefficient;

[0095] ​​The prediction and analysis module compares the activity prediction value output by the activity prediction model of the marine biofilm antimicrobial peptide with a pre-set activity threshold; if the activity prediction value exceeds the pre-set activity threshold, it is predicted that the marine biofilm contains antimicrobial peptides with qualified activity; otherwise, it is determined that the marine biofilm does not contain antimicrobial peptides with qualified activity.

[0096] It should be noted that the calculation formulas and various parameters involved in the calculations in the present invention have been dimensionally processed in advance, and the process of dimensionless processing is well known in the industry and will not be described here.

[0097] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for predicting antimicrobial peptides in marine biofilms based on data analysis, characterized in that: The following steps are involved: S1. Obtain relevant features of antimicrobial peptides in marine biofilms through a combined extraction method of sequence features and structural features; S2. Extracting sequence features and structural features based on a trained neural network model combining a convolutional neural network and a recurrent neural network. The sequence features include: the number and ratio of various amino acids in the antimicrobial peptide, the order of amino acids in the antimicrobial peptide, the number and distribution of positively and negatively charged amino acids in the antimicrobial peptide, and the hydrophobicity of the entire peptide chain of the antimicrobial peptide; the structural features include secondary structure; S3. Based on the sequence features and structural features extracted in S2, respectively, a first activity evaluation coefficient and a second activity evaluation coefficient are calculated, and an activity prediction model of the marine biofilm antimicrobial peptide is constructed using the first activity evaluation coefficient and the second activity evaluation coefficient; S4. Comparing the activity prediction value output by the marine biofilm antimicrobial peptide activity prediction model with a pre-set activity threshold; if the activity prediction value exceeds the pre-set activity threshold, then predicting the presence of qualified active antimicrobial peptides in the marine biofilm; otherwise, determining that the marine biofilm does not contain qualified active antimicrobial peptides; The process of obtaining the first activity evaluation coefficient is: By formula: Calculate the first activity evaluation coefficient ; For the The ratio of amino acids in antimicrobial peptides, is the weight factor, is the total number of amino acids, is the reference value, is the charge distribution index, is the hydrophobicity index; 、 、 is the exponential coefficient; The charge distribution index The way to obtain is: By formula group: and Calculate the charge distribution index ; in, is the charge distribution coefficient, 、 are the maximum and minimum values ​​of the charge distribution coefficient, respectively. is the number of positively charged amino acids, is the number of negatively charged amino acids, For the The charge of the positively charged amino acids, For the The charge of each negatively charged amino acid; Obtain the hydrophobicity index The process is: By formula: and Calculate the hydrophobicity index ; in, is the hydrophobicity coefficient, 、 are the maximum and minimum values ​​of the hydrophobicity coefficient, is the weight coefficient, For the The hydrophobicity value of each amino acid; The expression of the activity prediction model of marine biofilm antimicrobial peptides is: ； in, is the activity prediction value, 、 is the preset scale factor, is the second activity evaluation coefficient; ;in, for - the number of amino acids in the helical structure, is the total number of amino acids, is the standard value.

2. The method for predicting marine biofilm antimicrobial peptides based on data analysis according to claim 1, characterized in that: The method further comprises S5, verifying the result of predicting the presence of qualified active antimicrobial peptides in the marine biofilm; If the verification result is true, the activity prediction value is output normally; If the verification result is false, the output of the activity prediction value is stopped and the activity prediction model of the marine biofilm antimicrobial peptide is optimized.

3. The method for predicting marine biofilm antimicrobial peptides based on data analysis according to claim 2, characterized in that: The process for validating the results of predicting the presence of qualified antimicrobial peptides in the marine biofilm was as follows: Obtain current antimicrobial peptides Value, concentration of antimicrobial peptide And the diameter of the antibacterial circle ; Substituting into the formula: Calculate the evaluation index ; in, 、 、 is the preset scale factor, is the maximum value of antimicrobial peptides, is the standard concentration value of antimicrobial peptides, is the standard value of the diameter of the antibacterial circle; The calculated evaluation index and early warning values ​​of evaluation indicators Make comparisons; like , the verification result is true; like , the verification result is false.

4. A marine biofilm antimicrobial peptide prediction system based on data analysis, characterized in that: The system is applicable to the method for predicting marine biofilm antimicrobial peptides based on data analysis according to any one of claims 1 to 3, comprising: The data extraction module obtains the relevant features of antimicrobial peptides in marine biofilms through the combined extraction of sequence features and structural features; The feature acquisition module extracts sequence features and structural features based on a neural network model that combines a trained convolutional neural network with a recurrent neural network. The sequence features include: the number and ratio of various amino acids in the antimicrobial peptide, the order of amino acids in the antimicrobial peptide, the number and distribution of positively and negatively charged amino acids in the antimicrobial peptide, and the hydrophobicity of the entire peptide chain; the structural features include secondary structure; A prediction model construction module calculates a first activity evaluation coefficient and a second activity evaluation coefficient based on the extracted sequence features and structural features, and constructs an activity prediction model for marine biofilm antimicrobial peptides using the first activity evaluation coefficient and the second activity evaluation coefficient; The prediction and analysis module compares the activity prediction value output by the activity prediction model of the marine biofilm antimicrobial peptide with a pre-set activity threshold; if the activity prediction value exceeds the pre-set activity threshold, it is predicted that the marine biofilm contains antimicrobial peptides with qualified activity; otherwise, it is determined that the marine biofilm does not contain antimicrobial peptides with qualified activity.

Images