Quantitative optimization method and system for anticancer drug candidate compound molecules

Abstract

The present invention belongs to the technical field of data processing of anti-cancer candidate drug molecular descriptors, and provides a quantitative optimization method and system for anti-cancer candidate drug compound molecules. The method includes obtaining candidate drug molecular descriptors and target biological activity data; screening out several key molecular descriptors according to the correlation between the candidate drug molecular descriptors and the target biological activity data; encoding each key molecular descriptor and converting it into a corresponding vector representation; based on the vector representation of each key molecular descriptor, minimizing the cancer target biological activity value as the objective function, under the conditions of the value constraints of each key molecular descriptor and the ADMET property constraints of the compound, screening out the molecular descriptor category and content value range of the anti-cancer candidate drug.

Description

Technical Field

[0001] The invention belongs to the technical field of anti-cancer candidate drug molecule descriptor data processing, and in particular relates to a quantitative optimization method and system for anti-cancer candidate drug compound molecules. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Molecular biology is involved in the drug development process, and the analysis of the characteristics and properties of compounds has important application value. How to select the optimal drug compound combination molecules and their quantity in combination with the target characteristics to promote the inhibition of target biological activity and improve the ADMET characteristics of the drug is an inevitable requirement for studying highly effective drugs that inhibit cancer development. By establishing an effective model for inhibiting target protein activity, potential drug compounds can be screened and new compound molecules with better inhibitory target biological activity can be predicted. At the same time, in addition to paying attention to the biological activity of candidate drugs, it is also necessary to conduct in-depth discussions on the impact of the compound's ADMET properties on human health.

[0004] The inventors have found that in the current screening technology for anticancer candidate drug compound molecules, due to the high data dimensionality of the compounds and their molecular descriptors, the limited data amount and the similarity of the compound structures, the biological activity of antagonizing cancer targets and the constrained ADMET properties are not considered simultaneously, resulting in poor accuracy of the constructed prediction model, and thus the results and value range of the molecular screening of anticancer candidate drug compounds cannot be accurately predicted, ultimately leading to poor anticancer effects. Summary of the invention

[0005] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a quantitative optimization method and system for anticancer candidate drug compound molecules, which can simultaneously consider the biological activity of antagonizing cancer targets and constrained ADMET properties, improve the prediction accuracy of the molecular descriptor category and the content value range of the anticancer candidate drug, and thus improve the anticancer effect of the drug.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] The first aspect of the present invention provides a method for quantitative optimization of anticancer candidate drug compound molecules, comprising:

[0008] Obtain candidate drug molecular descriptors and target bioactivity data;

[0009] According to the correlation between the molecular descriptors of candidate drugs and the target bioactivity data, several key molecular descriptors were screened out;

[0010] Each key molecular descriptor is encoded and converted into a corresponding vector representation;

[0011] Based on the vector representation of each key molecular descriptor, with minimizing the bioactivity value of the cancer target as the objective function, under the constraints of the value of each key molecular descriptor and the ADMET property of the compound, the molecular descriptor categories and their content value ranges of candidate anticancer drugs are screened out.

[0012] The second aspect of the present invention provides a quantitative optimization system for anticancer candidate drug compound molecules, comprising:

[0013] A candidate drug information acquisition module, which is used to obtain candidate drug molecular descriptors and target biological activity data;

[0014] A key molecular descriptor screening module, which is used to screen out several key molecular descriptors based on the correlation between the molecular descriptors of candidate drugs and the target biological activity data;

[0015] A key molecule vector representation module, which is used to encode each key molecule descriptor and convert it into a corresponding vector representation;

[0016] The anticancer molecular descriptor quantification module is used to screen out the molecular descriptor categories and content value ranges of anticancer candidate drugs based on the vector representation of each key molecular descriptor, with minimizing the bioactivity value of the cancer target as the objective function, under the value constraints of each key molecular descriptor and the ADMET property constraints of the compound.

[0017] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps in ... as described above when the program is executed by a processor.

[0018] The fourth aspect of the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the quantitative optimization method for the anticancer candidate drug compound molecules as described above are implemented.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] The present invention provides a quantitative optimization method and system for anticancer candidate drug compound molecules, which screens out several key molecular descriptors based on the correlation between the candidate drug molecular descriptors and the target biological activity data, not only considering multiple different drug compound molecular descriptors, but also taking into account the chemical structure information of drug compounds from the perspective of "the structure of a substance determines its properties" in biomolecular science, and adding compound structure similarity analysis variables;

[0021] Based on the vector representation of each key molecular descriptor, with minimizing the bioactivity value of the cancer target as the objective function, the selection of decision variables and constraints is increased. Under the conditions of the value constraints of each key molecular descriptor and the ADMET property constraints of the compound, the prediction accuracy of the molecular descriptor category and the content value range of the anticancer candidate drug is effectively improved, thereby improving the anticancer effect of the drug.

[0022] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0024] Figure 1 is a flow chart of a quantitative optimization method for anticancer candidate drug compound molecules according to an embodiment of the present invention;

[0025] Figure 2 It is a schematic diagram of the structure of a quantitative optimization system for anticancer candidate drug compound molecules according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0027] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0028] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0029] Embodiment 1

[0030] like Figure 1 As shown, this embodiment provides a quantitative optimization method for anticancer candidate drug compound molecules, which specifically includes the following steps:

[0031] S101: Obtain candidate drug molecular descriptors and target biological activity data.

[0032] S102: Screening out several key molecular descriptors based on the correlation between the candidate drug molecular descriptors and the target biological activity data.

[0033] Specifically, in step S102, the process of screening out several key molecular descriptors is as follows:

[0034] S1021: Pearson correlation coefficient was used to calculate the correlation matrix, and linear correlation analysis was performed on each molecular descriptor to obtain the correlation matrix;

[0035] S1022: According to the correlation magnitudes of the variables in the correlation matrix, a set number of molecular descriptors with relatively high correlations are obtained and used as key molecular descriptors.

[0036] It should be noted here that the quantitative optimization method of the anti-cancer candidate drug compound molecules of this embodiment is not only applicable to breast cancer, but also to other cancers.

[0037] The data in this embodiment are all based on breast cancer as an example:

[0038] According to the Pearson correlation coefficient and the size of the variable correlation of the correlation matrix, several molecular descriptors with high correlation were obtained, and the data with low correlation were discarded. Several (for example, the top 20) molecular descriptors with the most significant impact on the target biological activity were screened out, as shown in Table 1, namely "MDEC-23", "MLogP", "LipoaffinityIndex", "maxsOH", "minsOH", "nC", "nT6Ring", "n6Ring", "minsssN", "BCUTp-1h", "C2SP2", "hmin", "AMR", "SwHBa", "maxsssN", "MDEC-22", "SP-5", "SaaCH", "CrippenLogP", "maxHsOH".

[0039] Table 1 Top 20 molecular descriptors and their correlation coefficients

[0040] Molecular descriptors Correlation coefficient Molecular descriptors Correlation coefficient MDEC-23 0.538048 C2SP2 0.426632 MLogP 0.529321 hmin 0.426365 LipoaffinityIndex 0.491855 AMR 0.425149 maxsOH 0.466621 oeLh 0.423176 minsOH 0.466127 maxsssN 0.420674 nC 0.459549 MDEC-22 0.420095 nT6Ring 0.440443 SP-5 0.419318 n6Ring 0.432108 SaaCH 0.418828 minsssN 0.430690 CrippenLogP 0.412300 BCUTp-1h 0.428526 maxHsOH 0.408761

[0041] S103: Encode each key molecular descriptor and convert it into a corresponding vector representation.

[0042] S104: Based on the vector representation of each key molecular descriptor, minimizing the bioactivity value of the cancer target is used as the objective function. Under the conditions of the value constraints of each key molecular descriptor and the ADMET property constraints of the compound, the molecular descriptor categories and content value ranges of the candidate anticancer drugs are screened.

[0043] The ADMET properties include properties of Caco-2, CYP3A4, hERG, HOB and MN; wherein Caco-2 indicates the permeability of the compound to the small intestinal epithelial cells, CYP3A4 indicates whether the compound can be metabolized by CYP3A4, hERG indicates whether the compound has cardiotoxicity, HOB indicates the oral bioavailability of the compound, and MN indicates whether the compound has genotoxicity.

[0044] According to the molecular descriptor data of multiple drug compounds and the ADMET properties of compounds, prediction models for compounds Caco-2, CYP3A4, hERG, HOB, and MN based on machine learning support vector machine (SVM) were constructed, and the relevant features extracted from the five property prediction models were analyzed. Through the synergistic analysis of the five types of ADMET properties of compounds, the same molecular descriptors were selected while taking into account the target biological activity.

[0045] In this embodiment, in the process of calculating the extracted molecular descriptors, the molecular descriptors and the value range are optimized as a complex linear programming model, and the optimization goal is to achieve both biological activity and ADMET properties. The objective function is defined as minimizing the target biological activity value, and its expression is as follows:

[0046]

[0047] Among them, x i Represents different molecular descriptors, C i Indicates the biological activity value corresponding to the sample, is the conversion factor in optimization, which can be determined by the biological activity value, S i For optimization.

[0048] The optimized direction formula is expressed as:

[0049]

[0050] Among them, RON before represents the Erα bioactivity value before optimizing the operating variables, RON after Represents the target bioactivity value after optimizing the operating variables. The calculation formula of the target bioactivity value before and after optimizing the operating variables is as follows:

[0051] ΔRON before =RON-RON before

[0052] ΔRON after =RON-RON after

[0053] ΔRON after The calculation formula is as follows:

[0054] RON after =f RON (x1, x2, ... x5)

[0055] Among them, x i (i=1,2…5) represents the main operating variables, f RON Represents the target bioactivity value prediction model.

[0056] In the specific implementation process, the objective function is solved by using a particle swarm optimization algorithm.

[0057] The decision variables are the main operators, and each set of values ​​represents a corresponding main variable operation plan.

[0058] According to the value ranges of the operator variables given in this embodiment, the value constraints of the main operation variables are constructed:

[0059] x i min ≤x i ≤x i max

[0060] Among them, x i min represents the minimum value that each operating variable can take, x i max Indicates the maximum value that each manipulated variable can take.

[0061] While paying attention to the biological activity of the compound, its ADMET properties also need to be considered. The compound ADMET property constraints at least meet three types of properties among Caco-2, CYP3A4, hERG, HOB and MN properties. Therefore, while screening molecular descriptors that inhibit the biological activity of the compound, the ADMET properties should be as good as possible. Therefore, the preliminary constraints can be designed as follows:

[0062]

[0063] The S after-Caco-2 It represents the ability of the compound to penetrate the intestinal epithelial cells after the planned operation variables, where '1' represents that the compound has good intestinal epithelial cell penetration and '0' represents that the compound has poor intestinal epithelial cell penetration.

[0064] The H after-CYP3A4 Indicates the ability of the compound to be metabolized by CYP3A4 after planning the operating variables, where '1' means that the compound can be metabolized by CYP3A4, and '0' means that the compound cannot be metabolized by CYP3A4.

[0065] The W after-hERGIt indicates the cardiotoxicity of the compound after the planned operation variable, where '1' means that the compound is cardiotoxic and '0' means that the compound is not cardiotoxic.

[0066] After planning the operating variables, the calculation formula for the biological properties of the compound is as follows: s Represents the prediction model of biological properties of compounds:

[0067] S after =f s (x1, x2, ... x 16 )

[0068] According to the requirements of this embodiment, the minimum value of target biological activity f(x) is used as the objective function, and the constraints are the five types of ADMET properties of the drug compound. According to the requirements, at least three of the five types of compound properties must be met, and the corresponding values ​​are provided by using binary classification. Therefore, the complex linear programming model initially established is as follows:

[0069]

[0070]

[0071] Particle swarm optimization (PSO) is an evolutionary computing method, the basic idea of ​​which is to find the optimal solution. First, the parameters of the particle swarm are initialized, and then evolutionary iterative calculations are continuously performed until the optimal solution is found. The swarm consists of i particles, each of which carries D-dimensional information. The current position of the particle is regarded as a feasible solution to the optimization problem. The movement of the particle is only related to the position and speed, which are respectively expressed as and Represent the position and velocity of a particle:

[0072]

[0073] Among them, L d and U d are the upper and lower limits of the feasible solution range of the particle. min and v min is the upper and lower limits of the particle search speed. In order to obtain the best solution for optimization, all particles need to continuously iterate to update the position and speed of the population. The update formulas for speed and position are as follows:

[0074]

[0075]

[0076] Among them, w is the inertia factor used to balance the relationship between local optimization ability and global optimization ability; C and C2 are acceleration terms representing individual learning factors and social learning factors respectively; R1 and R2 are randomly selected numbers between [0, 1]; is the iteration speed of the ith particle in the dth dimension at time t+1; v id is the iteration speed of the ith particle in the dth dimension at time t.

[0077] In order to better calculate the planning tasks of this embodiment, the "probability jump characteristic" is used on the basis of the traditional particle swarm optimization algorithm to enable the algorithm to jump out of the local extreme value. This embodiment introduces an elimination mechanism. If the range of the population searching for new solutions is always limited to a certain area, this solution will be eliminated and other particles will be selected to supplement the population. The diversity of the population is maintained by expanding the search range of new solutions.

[0078] This embodiment takes into account the value range of the molecular descriptors of constrained drug compounds. In order to avoid falling into the local optimal solution and the ability to jump out of the local extreme value, the search range of new solutions is expanded and the population diversity is protected. Based on the particle swarm optimization theory, a "probability jump characteristic" is introduced, and a new elimination mechanism is set by eliminating invalid solutions, thereby improving the global optimal performance of the algorithm and the accuracy of the model.

[0079] In this embodiment, the relevant feature descriptors are determined as "ETA_Eta_R_L, VAdjMatmaxsOH, SP-3, nC", and examples of the properties of the anticancer drug compounds screened are shown in Table 2.

[0080] The SMILES refers to the chemical structure information of drug compounds;

[0081] The IC50 refers to the concentration of the drug at which it achieves 50% of its maximum clinical efficacy (inhibition).

[0082] The pIC50 refers to the negative logarithm of 10 of the IC50 value.

[0083] The Caco-2 refers to the permeability of a compound to the small intestinal epithelial cells, wherein '1' represents good permeability of the compound to the small intestinal epithelial cells, and '0' represents poor permeability of the compound to the small intestinal epithelial cells.

[0084] The CYP3A4 refers to whether the compound can be metabolized by CYP3A4, wherein '1' represents that the compound can be metabolized by CYP3A4, and '0' represents that the compound cannot be metabolized by CYP3A4.

[0085] The hERG refers to whether the compound has cardiotoxicity, '1' represents that the compound has cardiotoxicity, and '0' represents that the compound does not have cardiotoxicity.

[0086] The HOB refers to the oral bioavailability of a compound, wherein '1' represents that the oral bioavailability of the compound is good, and '0' represents that the oral bioavailability of the compound is poor.

[0087] The MN refers to whether the compound has genotoxicity, wherein '1' represents that the compound has genotoxicity, and '0' represents that the compound does not have genotoxicity.

[0088] Table 2 Examples of properties of anticancer drug compounds screened

[0089]

[0090] From the experimental results, it can be seen that the biological activity is maintained above 7.4, and at the same time, more than three of the five properties of the compound Caco-2, CYP3A4, hERG, HOB, and MN are met. The value ranges of the five key molecular descriptors are shown in Table 3 below:

[0091] Table 3 Corresponding molecular descriptors and their value ranges

[0092] Molecular descriptors Exact value range (percentage) ETA_Eta_R_L 10.593～16.996 VAdjMat 5.321～6.169 maxsOH 8.702～10.370 SP-3 8.677～12.14 nC 20～26

[0093] Among them, the value range of "ETA_Eta_R_L" is between 10-17, the value range of "VAdjMat" is between 5-7, the value range of "maxsOH" is between 8-11, the value range of "SP-3" is between 8-12, and the value range of "SP-3" is between 20-26.

[0094] Accuracy is usually used as a measure of classification effect, but accuracy is not applicable to the evaluation of all classification situations. This problem is obvious in unbalanced classification. Take medical testing as an example. A large number of people in the hospital test samples are disease-free, and only a small number of people are sick. If accuracy is used as the basis for the test results, it may be found that the accuracy of the test results is as high as 99% or more, but this 99% accuracy shows that the patients are disease-free, and those who are sick may also be tested as disease-free. Such test results cause immeasurable losses to patients.

[0095] In this embodiment, the quantitative analysis of the number of molecular descriptors of candidate anti-breast cancer drugs under constraints is also an unbalanced classification problem, so the accuracy is not used as the evaluation index of classification. This embodiment uses confusion matrix, precision, recall rate, F1 _scores The quantitative analysis of molecular descriptors of the screened effective anti-breast cancer drug candidates was evaluated by indicators such as PCR and PCR.

[0096] Embodiment 2

[0097] like Figure 2 As shown, this embodiment provides a quantitative optimization system for anticancer candidate drug compound molecules, which specifically includes the following modules:

[0098] A candidate drug information acquisition module, which is used to obtain candidate drug molecular descriptors and target biological activity data;

[0099] A key molecular descriptor screening module, which is used to screen out several key molecular descriptors based on the correlation between the molecular descriptors of candidate drugs and the target biological activity data;

[0100] A key molecule vector representation module, which is used to encode each key molecule descriptor and convert it into a corresponding vector representation;

[0101] The anticancer molecular descriptor quantification module is used to screen out the molecular descriptor categories and content value ranges of anticancer candidate drugs based on the vector representation of each key molecular descriptor, with minimizing the bioactivity value of the cancer target as the objective function, under the value constraints of each key molecular descriptor and the ADMET property constraints of the compound.

[0102] Among them, in the key molecular descriptor screening module, the process of screening out several key molecular descriptors is:

[0103] The Pearson correlation coefficient was used to calculate the correlation matrix, and linear correlation analysis was performed on each molecular descriptor to obtain the correlation matrix;

[0104] According to the correlation size of the variables in the correlation matrix, a set number of molecular descriptors with higher correlation are obtained and used as key molecular descriptors.

[0105] It should be noted here that each module in this embodiment corresponds to each step in Example 1 one by one, and the specific implementation process is the same, which will not be repeated here.

[0106] Embodiment 3

[0107] This embodiment provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the quantitative optimization method for anticancer candidate drug compound molecules as described above.

[0108] Embodiment 4

[0109] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the quantitative optimization method for anticancer candidate drug compound molecules as described above are implemented.

[0110] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products of the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of the processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0111] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A quantitative optimization method for anticancer candidate drug compound molecules, characterized in that: include: Obtain candidate drug molecular descriptors and target bioactivity data; According to the correlation between the molecular descriptors of candidate drugs and the target bioactivity data, several key molecular descriptors were screened out; Each key molecular descriptor is encoded and converted into a corresponding vector representation; Based on the vector representation of each key molecular descriptor, with minimizing the bioactivity value of the cancer target as the objective function, under the constraints of the value of each key molecular descriptor and the ADMET property of the compound, the molecular descriptor category and its content value range of the anticancer candidate drug are screened; The process of screening out several key molecular descriptors is as follows: using the Pearson correlation coefficient to calculate the correlation matrix, performing linear correlation analysis on each molecular descriptor to obtain the correlation matrix; according to the correlation size of the variables in the correlation matrix, obtaining a set number of molecular descriptors with higher correlation and using them as key molecular descriptors; The ADMET properties include properties of Caco-2, CYP3A4, hERG, HOB and MN; wherein Caco-2 indicates the permeability of the compound to the small intestinal epithelial cells, CYP3A4 indicates whether the compound can be metabolized by CYP3A4, hERG indicates whether the compound has cardiotoxicity, HOB indicates the oral bioavailability of the compound, and MN indicates whether the compound has genotoxicity; The compound ADMET property constraints must satisfy at least three of the properties of Caco-2, CYP3A4, hERG, HOB, and MN; The objective function is solved by using a particle swarm optimization algorithm.

2. The method for quantitative optimization of anticancer candidate drug compound molecules according to claim 1, characterized in that: In the process of solving the objective function based on the particle swarm optimization algorithm, an elimination mechanism is introduced: if the range of the population's search for new solutions is always limited to a certain area, this solution will be eliminated, and other particles will be selected to supplement the population, so as to maintain the diversity of the population by expanding the search range of new solutions.

3. A quantitative optimization system for anticancer candidate drug compound molecules, characterized in that: include: A candidate drug information acquisition module, which is used to obtain candidate drug molecular descriptors and target biological activity data; A key molecular descriptor screening module, which is used to screen out several key molecular descriptors based on the correlation between the molecular descriptors of candidate drugs and the target biological activity data; A key molecule vector representation module, which is used to encode each key molecule descriptor and convert it into a corresponding vector representation; The anticancer molecular descriptor quantification module is used to screen out the molecular descriptor categories and content value ranges of anticancer candidate drugs based on the vector representation of each key molecular descriptor, with minimizing the bioactivity value of the cancer target as the objective function, under the constraints of the value of each key molecular descriptor and the ADMET property of the compound; In the key molecular descriptor screening module, the process of screening out several key molecular descriptors is as follows: The Pearson correlation coefficient was used to calculate the correlation matrix, and linear correlation analysis was performed on each molecular descriptor to obtain the correlation matrix; According to the correlation size of the variables in the correlation matrix, a set number of molecular descriptors with higher correlation are obtained and used as key molecular descriptors.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the quantitative optimization method of the anti-cancer candidate drug compound molecule as described in any one of claims 1 to 2 are implemented.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the method for quantitative optimization of anticancer candidate drug compound molecules according to any one of claims 1 to 2 are implemented.

Images