Method for predicting physical properties of polymer and polymer

Abstract

Provided are: a high-precision method for predicting the physical properties of a polymer; and a polymer having excellent physical properties such as selective capture of specific cells and biocompatibility. The invention relates to a method for predicting the physical property of a polymer, which comprises the following steps of: 1, obtaining a monomer structure from the structure of the polymer, 2, further converting the monomer structure into an expression form which can be identified by a computer after converting the monomer structure, and 3, calculating a descriptor based on the expression form which can be identified by the computer, 4, obtaining the physical property of the polymer, taking the physical property as a target variable, taking the descriptor as an explanatory variable, carrying out regression calculation, and making a regression model, 5, calculating the expression form of a new monomer based on the expression form which can be identified by a computer, and 6, calculating the descriptor after the expression form of the new monomer is converted, and 7, applying the descriptor of the new monomer to the regression model, and predicting the physical property of the polymer polymerized by the new monomer.

Description

Technical Field

[0001] The invention relates to a method for predicting physical properties of a polymer and a polymer. Background Art

[0002] A regression model for predicting the physical property values ​​of compounds using machine learning has been proposed, and the physical properties of new compounds are predicted based on the regression model. Furthermore, while synthesizing new compounds, measuring physical properties, and verifying predicted values, new compounds and their actual measured physical property values ​​are added to improve the accuracy of the regression model, and this process is repeated. In addition, useful compounds can be discovered from new compounds and physical property values, and a large number of implementations have been carried out, especially in the field of pharmaceuticals.

[0003] However, in the field of high molecular weight compounds (polymers), such methods cannot be said to be fully utilized. In particular, the prediction of specific physical properties of polymers (for example, water content, amount of intermediate water, amount of bound water, adhesion to specific cells (cancer cells, stem cells) and biocompatibility) has not yet been fully implemented.

[0004] On the other hand, as a device for capturing specific cells in blood and body fluids (blood cells, cancer cells present in blood / body fluids, etc.), a substrate coated with a special polymer has been proposed (see Patent Document 1, etc.).

[0005] However, in such a substrate, not only specific cells such as cancer cells are captured, but also blood cells are sometimes captured at the same time. Therefore, it is desirable to provide a polymer that can selectively capture specific cells such as cancer cells and inhibit the capture of normal cells such as blood cells. Furthermore, it would also be desirable to provide polymers that are more biocompatible. [Prior art literature] [Patent Literature]

[0006] [Patent Document 1] Japanese Patent Publication No. 2005-523981 Summary of the invention Problems to be solved by the invention

[0007] The present invention aims to solve the above-mentioned problems and to provide a highly accurate method for predicting the physical properties of a polymer and a polymer having excellent physical properties such as selective capture of specific cells and biocompatibility.

Methods to solve the problem

[0008] The present invention relates to a method for predicting physical properties of a polymer, comprising: Step 1: Obtain the monomer structure from the polymer structure, Step 2: The above monomer structure itself or after conversion is further converted into an expression form that can be recognized by a computer. Step 3: Calculate the descriptor based on the expression form that the computer can recognize. Step 4: Obtain the physical properties of the polymer and use them as the target variable, use the descriptor as the explanatory variable, perform regression calculations, and create a regression model. Step 5: Based on the expressions that the computer can recognize, calculate the expression form of the new monomer. Step 6: Calculate the descriptor of the new monomer expression itself or after conversion. Step 7: Apply the descriptor of the new monomer to the regression model to predict the physical properties of the polymer polymerized from the new monomer. [Effects of the invention]

[0009] According to the present invention, a highly accurate method for predicting polymer properties can be provided because the method includes steps 1 to 7. Therefore, according to this method, a large number of polymers (candidate polymers) with excellent properties can be rapidly discovered. Furthermore, it is expected that the polymer can provide excellent physical properties such as the ability to selectively capture specific cells such as cancer cells and biocompatibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0010]

Figure 1

[0011] <Polymer property prediction method> The method for predicting the physical properties of a polymer of the present invention comprises step 1: obtaining a monomer structure from the structure of the polymer, step 2: further converting the above-mentioned monomer structure itself or after conversion into an expression form recognizable by a computer, step 3: calculating a descriptor based on the expression form recognizable by a computer, step 4: obtaining the physical properties of the polymer and using it as a target variable, using the descriptor as an explanatory variable, performing regression calculation, and preparing a regression model, step 5: calculating the expression form of a new monomer based on the expression form recognizable by a computer, step 6: calculating the descriptor of the expression form of the above-mentioned new monomer itself or after conversion, and step 7: applying the descriptor of the above-mentioned new monomer to the above-mentioned regression model to predict the physical properties of the polymer polymerized by the above-mentioned new monomer.

[0012] The above-mentioned method for predicting the physical properties of a polymer includes step 1: obtaining a monomer structure from the structure of the polymer. In the above-mentioned step 1, for example, a monomer structure having a carbon-carbon double bond constituting the polymer is obtained from the structure of the polymer. Specifically, when the structure of the polymer is polyacrylic acid, the structure of the monomer constituting the polyacrylic acid, i.e., acrylic acid, is obtained.

[0013] In the above step 1, the above polymer is not particularly limited, and a known polymer can be used as appropriate. Examples of the polymer include homopolymers of one monomer and copolymers of two or more monomers. The polymers may be used alone or in combination of two or more.

[0014] The polymer can be prepared by a known method, for example, by using a solution of monomers constituting the polymer and polymerizing the monomers by a known method. The solvent of the monomer solution is not particularly limited, for example, the following solvents can be used. Among them, toluene and methanol are preferred.

[0015] From the viewpoint of selective capture of specific cells such as cancer cells and biocompatibility, the polymer is preferably a hydrophilic polymer (hydrophilic polymer).

[0016] Examples of the hydrophilic polymer include homopolymers and copolymers of one or more hydrophilic monomers, copolymers of one or more hydrophilic monomers and one or more other monomers, etc. These may be used alone or in combination of two or more.

[0017] The hydrophilic monomer is not particularly limited, and various monomers having a hydrophilic group can be used. Examples of the hydrophilic group include well-known hydrophilic groups such as amide group, sulfate group, sulfonic acid group, carboxyl group, hydroxyl group, amino group, amide group, and oxyethylene group.

[0018] Specific examples of the hydrophilic monomer include (meth)acrylic acid, (meth)acrylates ((meth)acrylate alkoxyalkyl esters such as methoxyethyl (meth)acrylate, (meth)acrylate hydroxyalkyl esters such as hydroxyethyl (meth)acrylate), (meth)acrylamide, (meth)acrylamide derivatives having a cyclic group ((meth)acryloylmorpholine, etc.). From the viewpoint of selective capture of specific cells such as cancer cells and biocompatibility, (meth)acrylic acid, (meth)acrylates, and (meth)acrylate alkoxyalkyl esters are preferred, (meth)acrylate alkoxyalkyl esters are more preferred, and 2-methoxyethyl acrylate is particularly preferred. These may be used alone or in combination of two or more.

[0019] Other monomers other than the above can be appropriately selected within the range that does not impair the effect of the hydrophilic polymer. Specific examples of the above other monomers include aromatic monomers such as styrene, vinyl acetate, N-isopropylacrylamide that can impart temperature responsiveness, etc. These can be used alone or in combination of two or more.

[0020] As the above homopolymer and copolymer, specifically, there can be mentioned homopolymers composed of one hydrophilic monomer such as polyacrylic acid, polyacrylate, polymethacrylic acid, polymethacrylate, polyacryloylmorpholine, polymethacryloylmorpholine, polyacrylamide, polymethacrylamide, polyalkoxyalkyl acrylate, polyalkoxyalkyl methacrylate, etc.; copolymers composed of two or more hydrophilic monomers exemplified above; copolymers composed of one or more hydrophilic monomers exemplified above and one or more other monomers exemplified above, etc. The above hydrophilic polymer can be used alone or in combination of two or more.

[0021] Specific examples of the homopolymer and copolymer of the hydrophilic monomer include polyacrylic acid, polyacrylate, polymethacrylic acid, polymethacrylate, polyacryloylmorpholine, polymethacryloylmorpholine, polyacrylamide, polymethacrylamide, polyalkoxyalkyl acrylate, polyalkoxyalkyl methacrylate, and the like.

[0022] From the viewpoint of selective capture of specific cells such as cancer cells and biocompatibility, the hydrophilic polymer is preferably a polymer represented by the following formula (I). These may be used alone or in combination of two or more.

Chemical formula 1

[0023] From the viewpoint of selective capture of specific cells such as cancer cells and biocompatibility, the polymer represented by the above formula (I) is preferably a polymer represented by the following formula (I-1), for example. These may be used alone or in combination of two or more.

Chemical formula 2

[0024] R 52 The number of carbon atoms in the alkyl group is preferably 1 to 10, more preferably 1 to 5. 52Particularly preferred is a methyl group or an ethyl group. p is preferably 1 to 5, and more preferably 1 to 3. m is preferably 1 to 3. n (number of repeating units) is preferably 15 to 1500, and more preferably 40 to 1200.

[0025] From the viewpoint of selective capture of specific cells such as cancer cells and biocompatibility, the hydrophilic polymer is preferably a copolymer of a compound represented by the following formula (II) and other monomers. These may be used alone or in combination of two or more.

[0026]

Chemical formula 3

[0027] From the viewpoint of selective capture of specific cells such as cancer cells and biocompatibility, the compound represented by the above formula (II) is preferably, for example, a compound represented by the following formula (II-1). These may be used alone or in combination of two or more.

Chemical formula 4

[0028] Among the above-mentioned hydrophilic polymers, from the viewpoint of selective capture of specific cells such as cancer cells and biocompatibility, the hydrophilic polymer represented by the above-mentioned formula (I) is preferred, the hydrophilic polymer represented by the above-mentioned formula (I-1) is more preferred, and polyalkoxyalkyl acrylates and polyalkoxyalkyl methacrylates are further preferred.

[0029] From the viewpoint of obtaining a better effect, the number average molecular weight (Mn) of the polymer is preferably 8,000 to 150,000, more preferably 10,000 to 60,000, and even more preferably 10,000 to 39,000. When the polymer is a hydrophilic polymer, the same number average molecular weight (Mn) is also preferred. In addition, in this specification, the number average molecular weight (Mn) and the weight average molecular weight (Mw) can be calculated by converting the measured values ​​by gel permeation chromatography (GPC) (GPC-8000 series manufactured by Tosoh Corporation, detector: differential refractometer, column: TSKGELSUPERMALTPORE HZ-M manufactured by Tosoh Corporation) into standard polystyrene.

[0030] In the above-mentioned method for predicting physical properties of a polymer, in the above-mentioned step 1, for example, when the structure of the polymer is poly(2-methoxyethyl acrylate), 2-methoxyethyl acrylate (monomer structure) is obtained.

[0031] In the above-mentioned method for predicting the physical properties of a polymer, after the above-mentioned step 1, step 2 is performed: the structure of the monomer obtained in the above-mentioned step 1 is converted, and further converted into an expression form that can be recognized by a computer. That is, in the above-mentioned step 2, the compound of the monomer structure itself obtained in the above-mentioned step 1 (the obtained monomer structure itself), or the compound after the monomer structure is converted, is converted into an expression form that can be recognized by a computer.

[0032] The method for converting the monomer structure obtained in the above step 1 is not particularly limited as long as it is a method that can convert the monomer structure, and examples thereof include: (1) a method of opening the carbon-carbon double bonds of the monomer structure (monomer structure itself (having carbon-carbon double bonds)) obtained in the above step 1 to convert it into a compound having saturated hydrocarbon groups such as hydrogen or methyl introduced at both ends (conversion into a single monomer (no carbon-carbon double bonds)), (2) a method of opening the carbon-carbon double bonds of the monomer structure (monomer structure itself) obtained in the above step 1 to convert it into a multimer bonded with carbon-carbon single bonds, and converting it into a compound having saturated hydrocarbon groups such as hydrogen or methyl introduced at both ends (conversion into a multimer having no carbon-carbon double bonds such as a dimer (no carbon-carbon double bonds) or a trimer (no carbon-carbon double bonds)). In the case of converting into the above multimer, it is preferred to open the carbon-carbon double bonds of the monomer structure (monomer structure itself) obtained in the above step 1 to convert it into a multimer bonded with carbon-carbon single bonds, and convert it into a compound having methyl introduced at both ends.

[0033] In the above step 2, the compound of the above monomer structure itself obtained in the above step 1 (the obtained monomer structure itself), or the above polymer without carbon-carbon double bonds, etc., is converted into an expression form recognizable by a computer, and as the above computer, ordinary personal computers, servers, workstations, compute nodes, microcomputers, etc. can be listed. The hardware structure of the computer can use known substances.

[0034] Examples of the computer-recognizable expression format include a known expression format such as the Simplified Molecular Input Line Entry System (SMILES) expression format and the MOL file expression format.

[0035] In the above-mentioned method for predicting physical properties of a polymer, in the above-mentioned step 2, for example, when the structure of the polymer is poly(2-methoxyethyl acrylate), first, the 2-methoxyethyl acrylate (monomer structure itself (having a carbon-carbon double bond)) obtained in the above-mentioned step 1 is converted into the following compounds (A) to (C), etc. (A) Single monomer (no carbon-carbon double bond): A compound in which the carbon-carbon double bond of 2-methoxyethyl acrylate is opened and methyl groups are introduced at both ends. (B) Dimer (no carbon-carbon double bond): A compound obtained by opening the carbon-carbon double bond of 2-methoxyethyl acrylate to form a dimer bonded by a carbon-carbon single bond and introducing methyl groups at both ends. (C) Trimer (no carbon-carbon double bond): A compound obtained by opening the carbon-carbon double bond of 2-methoxyethyl acrylate, making it a trimer bonded with a carbon-carbon single bond, and introducing methyl groups at both ends.

[0036] In the above step 2, 2-methoxyethyl acrylate and the above compounds (A) to (C) are further converted into, for example, the following SMILES expression (SMILES notation) which is an expression format recognizable by a computer. 2-Methoxyethyl acrylate: C=C(C(=O)OCCOC) (Hydrogens are not shown. Branches are recorded in ()) (A) Single monomer (no carbon-carbon double bond): CCC(C(=O)OCCOC)C (B) Dimer (no carbon-carbon double bond): CCC(C(=O)OCCOC)CC(C(=O)OCCOC)C (C) Trimer (no carbon-carbon double bond): CCC(C(=O)OCCCOC)CC(C(=O)OCCCOC)CC(C(=O)OCCCOC)C

[0037] In the above-mentioned method for predicting the physical properties of a polymer, in addition to the above-mentioned step 2, step 3 is performed: calculating a descriptor from an expression form that can be recognized by a computer. That is, for a polymer whose physical properties and structure are well-known, the monomer structure with carbon-carbon double bonds constituting the polymer (the monomer structure itself (with carbon-carbon double bonds)) or the compound after converting the monomer structure (the above-mentioned single monomer (without carbon-carbon double bonds), the above-mentioned dimer (without carbon-carbon double bonds), the above-mentioned trimer (without carbon-carbon double bonds) and other polymers without carbon-carbon double bonds) are converted into an expression form that can be recognized by a computer, and then the descriptor is calculated from the expression form.

[0038] The physical properties of the polymer are not particularly limited, and examples thereof include glass transition temperature (Tg), melting point, refractive index, contact angle, surface free energy, density, water absorption, amount of intermediate water, amount of bound water, selective adhesion ability to specific cells such as cancer cells, biocompatibility, etc. Among them, the amount of intermediate water, selective adhesion ability to specific cells, and biocompatibility can be appropriately predicted.

[0039] In the above step 3, as the expression form when the monomer structure itself constituting the polymer or the compound after the monomer structure is converted into an expression form that can be recognized by the computer, the aforementioned SMILES expression form, MOLfile expression form, etc. can be listed. Then, the descriptors (numerical data) that digitize the chemical structure are calculated by the formulas of these expressions. For example, 208 or 400 descriptors are calculated from SMILES to RDkit using Python. In addition, descriptors such as mordred, PaDEL, alvaDesc can also be used to calculate descriptors. Further, as other methods for digitizing the chemical structure, fingerprints can also be listed.

[0040] In the above-mentioned method for predicting the physical properties of the polymer, in the above-mentioned step 3, for example, SMILES expressions of the monomer structures (the monomer structure itself (with carbon-carbon double bonds)) constituting the polymers or the compounds after conversion of the monomer structures (the above-mentioned single monomer (without carbon-carbon double bonds), the above-mentioned dimer (without carbon-carbon double bonds), the above-mentioned trimer (without carbon-carbon double bonds)) can be prepared respectively from 11 kinds of polymers with physical properties of known amounts of intermediate water, and descriptors of 208 numerical data reflecting the chemical structures can be calculated respectively from all SMILES expressions.

[0041] In addition, the above step 3 can be performed separately from steps 1 and 2, regardless of the order of steps 1 and 2. For example, step 3 can be performed after steps 1 and 2, or step 3 can be performed separately at the same time as steps 1 and 2, or step 1 and 2 can be performed after step 3 is performed.

[0042] In the above-mentioned method for predicting the physical properties of a polymer, after the above-mentioned steps 1, 2, and 3, step 4 is performed: the physical properties of the polymer are obtained and used as target variables, and the descriptors are used as explanatory variables, and regression calculations are performed to prepare a regression model. That is, in the above-mentioned step 4, the physical properties of the polymer are set as target variables, and the descriptors (for example, the above-mentioned 208 descriptors) are set as explanatory variables, and regression calculations are performed to prepare a regression equation (regression model).

[0043] The algorithm used in model construction can utilize known algorithms such as algorithms using machine learning. When performing regression calculation to make a regression equation (regression model), as regression calculation, multiple regression calculation, support vector regression calculation, decision tree calculation, random forest regression calculation, XGBoost regression calculation, LightGBM regression calculation, CatBoost regression calculation, Gaussian process regression calculation, neural network, deep learning and other known methods can be used. Regression calculation can use one regression calculation or a combination of two or more regression calculations.

[0044] In the above-mentioned method for predicting the physical properties of the polymer, in the above-mentioned step 4, for example, the four types of monomer structures constituting the polymer (the monomer structure itself (with carbon-carbon double bonds)), the compounds after the conversion of the monomer structure (the above-mentioned single monomer (without carbon-carbon double bonds), the above-mentioned dimer (without carbon-carbon double bonds), and the above-mentioned trimer (without carbon-carbon double bonds)) can be used, and the amount of intermediate water can be set as the target variable and the 208 descriptors can be set as explanatory variables, and four models for predicting the amount of intermediate water can be prepared by an ensemble learning method of random forest, XGBoost regression calculation, LightGBM regression calculation or CatBoost regression calculation. In addition, in the case of Boost regression calculation, neural network, and deep learning, in order to reduce overfitting, the entire data set can be used for training and the hyper parameters can be adjusted. In order to adjust the hyper parameters, there are methods such as grid search, random search, and Bayesian optimization. Hyper parameters refer to parameters set by people before starting learning. Hyper parameters are values ​​necessary to control the behavior of various algorithms. Representative examples include the maximum depth of a decision tree, the maximum number of leaves of a decision tree, the minimum number of samples contained in a leaf, the proportion of randomly extracted samples in each decision tree, the proportion of randomly extracted columns in each decision tree, the L1 regularization term for the leaf weights of a decision tree, the L2 regularization term for the leaf weights of a decision tree, the lower limit of the loss reduction caused by adding leaves of a decision tree, the optimization method of the loss function, the number of iterations of the Boosting algorithm, the number of epochs, the learning rate, the threshold, the minibatch size, the number of layers, the number of neurons in a layer, etc.

[0045] In the above-mentioned method for predicting the physical properties of a polymer, after the above-mentioned steps 1, 2, 3, and 4, step 5 is performed: the expression form of the new monomer is calculated from the expression form that can be recognized by a computer. That is, in the above-mentioned step 5, the above-mentioned expression form of the new monomer is calculated (generated) from the above-mentioned expression form (SMILES expression form, MOL file expression form, etc.) such as the monomer structure (having carbon-carbon double bonds) constituting the polymer.

[0046] The calculation (generation) of the expression form that the computer can recognize of the new monomer can, for example, be performed by decomposing the chemical structure of the compound in the expression form into fragments and then mechanically bonding them through the BRICS (Breaking of Retrosynthetically Interesting Chemical Substructures) algorithm, RECAP (Retrosynthetic Combinatiorial Analysis Procedure) algorithm, DOCS (Design Of Genuine Structures) algorithm, etc., to generate (calculate) the expression form of the new compound. In addition, as the expression form that the computer can recognize, the aforementioned SMILES expression form, MOLfile expression form, etc. can be listed.

[0047] For example, when using RECAP, for organic molecules, attention is paid to the chemical bonds of amide, ester, amine, NC in urea, ether, C=C, ammonium, NS in sulfonamide, aromatic ring-aromatic ring, N (in aromatic ring)-C (sp3) and N (in lactam ring)-C (sp3), and the chemical bonds are decomposed into fragments at the places where the chemical bonds in the molecule are easily broken. When using BRICS, attention is paid to 16 kinds of chemical bonds in the same way as RECAP, and they are decomposed into fragments.

[0048] In addition, when calculating (generating) a computer-recognizable expression of a new monomer, compounds without carbon-carbon double bonds may be generated. In this case, in order to predict the physical properties of the polymer, compounds without carbon-carbon double bonds are removed from the generated monomers.

[0049] In the above-mentioned method for predicting the physical properties of the polymer, in the above-mentioned step 5, for example, using BRICS, 74 compounds are generated from the SMILES expressions of 11 monomer structures (monomer structures themselves (having carbon-carbon double bonds)) constituting 11 polymers whose physical properties of the amount of the above-mentioned intermediate water are known, using BRICS, etc., among which 33 compounds that can become monomers are selected (excluding 41), and SMILES expressions of 33 new monomers are calculated (generated).

[0050] In the above-mentioned method for predicting the physical properties of a polymer, step 6 is performed: the expression form (the expression form that can be recognized by a computer) of the new monomer calculated (generated) in the above-mentioned step 5 is itself or after conversion, and a descriptor is calculated. That is, in the above-mentioned step 6, first, the expression form (the expression form that can be recognized by a computer) of the new monomer calculated (generated) in the above-mentioned step 5 is converted into the expression form of a compound of the structure of the monomer itself (the monomer structure itself) or the compound after the structure of the monomer is converted, using the same method as the above-mentioned step 2, and then, the descriptor (numerical data) is calculated from the expression form using the same method as the above-mentioned step 3.

[0051] In the above-mentioned method for predicting the physical properties of the polymer, in the above-mentioned step 6, for example, for the SMILES expressions of the 33 new monomers calculated (generated) above, SMILES expressions of the monomer structures of the 33 new monomers (the monomer structure itself (with carbon-carbon double bonds)) or the compounds after the conversion of the monomer structures (the above-mentioned single monomer (without carbon-carbon double bonds), the above-mentioned dimer (without carbon-carbon double bonds), the above-mentioned trimer (without carbon-carbon double bonds)) are made, and then, the descriptors (numerical data) are calculated respectively.

[0052] In the above-mentioned method for predicting the physical properties of a polymer, step 7 is performed: the descriptor of the above-mentioned new monomer is applied to the above-mentioned regression model to predict the physical properties of the polymer polymerized from the above-mentioned new monomer. That is, in the above-mentioned step 7, the physical properties of the polymer of the new monomer are predicted by applying the descriptor (numerical data) of the new monomer calculated in the above-mentioned step 6 to the regression model prepared in the above-mentioned step 4.

[0053] In the above step 7, for example, by applying the descriptors (numerical data) of the monomer structures of the above 33 new monomers calculated (generated) in the above step 6 (the monomer structure itself (with carbon-carbon double bonds)) or the compounds after the conversion of the monomer structures (the above single monomer (without carbon-carbon double bonds), the above dimer (without carbon-carbon double bonds), the above trimer (without carbon-carbon double bonds)) to the regression model prepared in the above step 4, the amount of intermediate water in the polymer obtained by polymerizing the above 33 new monomers is predicted.

[0054] Specifically, Figure 1 The relationship between the predicted value of the amount of intermediate water in a polymer polymerized from the monomer structure of 33 new monomers (monomer structure itself (with carbon-carbon double bond)) and the predicted value of the amount of intermediate water in a polymer polymerized from the compound converted from the monomer structure (the above single monomer (without carbon-carbon double bond), the above dimer (without carbon-carbon double bond), and the above trimer (without carbon-carbon double bond)) is shown using the regression model prepared in the above step 4. The results show that there is no difference in the predicted value of the amount of intermediate water with or without carbon-carbon double bonds, and there is no difference even if the number of polymerized monomers is increased.

[0055] Furthermore, when the upper limit of the compounds obtained from the monomer structures (monomer structures themselves (with carbon-carbon double bonds)) of the above 11 polymers using the BRICS algorithm was set to 2000 and compounds were generated, more than 1600 compounds were generated (SMILES expression). Among them, compounds without carbon-carbon double bonds and compounds with two or more carbon-carbon double bonds were removed, and 625 compounds with only one carbon-carbon double bond were selected. The descriptors of these 625 compounds were calculated, and the amount of intermediate water in these 625 compounds was predicted by the prediction model of compounds with only one carbon-carbon double bond.

[0056] Among them, the intermediate water volume is 0.01~0.15 as shown in Table 1-1, Table 1-2 (continued from Table 1-1), and Table 1-3 (continued from Table 1-1). In addition, the intermediate water volume of 0.22~0.5 is shown in Table 2-1, Table 2-2 (continued from Table 2-1), Table 2-3 (continued from Table 2-1), Table 2-4 (continued from Table 2-1), Table 2-5 (continued from Table 2-1), Table 2-6 (continued from Table 2-1), and Table 2-7 (continued from Table 2-1).

[0057]

Table 1-1

[0058]

Table 1-2

[0059]

Table 1-3

[0060]

Table 2-1

[0061]

Table 2-2

[0062]

Table 2-3

[0063]

Table 2-4

[0064]

Table 2-5

[0065]

Table 2-6

[0066]

Table 2-7

[0067] By using the above-described machine learning method, a regression equation (regression model) for predicting the physical properties of a polymer can be prepared. In addition, it is known that when the amount of intermediate water is 0.01 to 0.15 (Tables 1-1 to 1-3), it becomes a polymer with good adhesion to specific cells such as cancer cells. Therefore, the polymer whose amount of intermediate water is 0.01 to 0.15 calculated by the above-mentioned polymer property prediction method has excellent selective adhesion to specific cells, especially excellent selective adhesion to cancer cells. For example, the polymers polymerized by the monomers expressed in the SMILES form of Tables 1-1 to 1-3, whose amount of intermediate water is 0.01 to 0.15, have good adhesion to cancer cells.

[0068] Furthermore, it is known that when the amount of intermediate water is 0.22 to 0.5 (Tables 2-1 to 2-7), the polymer has good biocompatibility. Therefore, the polymer whose intermediate water amount is 0.22 to 0.5 calculated by the above-mentioned polymer property prediction method has excellent biocompatibility. For example, the polymers polymerized by the monomers expressed in the SMILES form of Tables 2-1 to 2-7 have good biocompatibility when the amount of intermediate water is 0.22 to 0.5.

[0069] The above is the result of using random forest regression calculation with double bonds. In contrast, prediction models for single monomers without double bonds were made using XGBoost regression calculation and CatBoost regression calculation (hyperparameters were manually adjusted and RMSE was reduced respectively). In addition, the Optuna framework was used in XGBoost regression calculation, and hyperparameters were adjusted and prediction models were made using Bayesian optimization. The monomers expressed by SMILES in Table 3 were converted into single monomers without double bonds (double bonds were opened and methyl groups were introduced at both ends) and descriptors were calculated. The amount of intermediate water predicted by the above three prediction models is shown in Table 3. It can be seen that the predicted values ​​are within a certain range, and the predicted values ​​are relatively accurate to a certain extent.

[0070]

Table 3

[0071] The present invention (1) is: a method for predicting the physical properties of a polymer, comprising the following steps: Step 1: Obtain the monomer structure from the polymer structure, Step 2: The monomer structure itself or after conversion is further converted into an expression form that can be recognized by a computer, Step 3: Calculate the descriptor based on the expression form that the computer can recognize. Step 4: Obtain the physical properties of the polymer and use them as the target variable, use the descriptor as the explanatory variable, perform regression calculations, and create a regression model. Step 5: Based on the expressions that the computer can recognize, calculate the expression form of the new monomer. Step 6: Calculate the descriptor of the new monomer expression itself or after conversion, Step 7: Apply the descriptor of the new monomer to the regression model to predict the physical properties of the polymer polymerized from the new monomer.

[0072] The present invention (2) is: a method for predicting the physical properties of a polymer according to the present invention (1), wherein the expression form recognizable by the computer is a SMILES expression form or a MOL file expression form.

[0073] The present invention (3) is: a method for predicting the physical properties of a polymer according to the present invention (1) or (2), wherein the regression calculation is selected from at least one of multiple regression calculation, support vector regression calculation, decision tree calculation, random forest regression calculation, XGBoost regression calculation, LightGBM regression calculation, CatBoost regression calculation, Gaussian process regression calculation, neural network and deep learning.

[0074] The present invention (4) is: a method for predicting the physical properties of a polymer according to any combination of any one of the present inventions (1) to (3), wherein the algorithm for calculating the expression form of the new monomer is at least one selected from BRICS, RECAP and DOCS.

[0075] The present invention (5) is: a method for predicting the physical properties of a polymer according to any combination of any one of the present inventions (1) to (4), wherein the physical property of the polymer is at least one selected from glass transition temperature, melting point, refractive index, contact angle, surface free energy, density, water absorption rate, amount of intermediate water, amount of bound water, selective adhesion ability to specific cells and biocompatibility.

[0076] The present invention (6) is: a method for predicting the physical properties of a polymer according to any combination of any one of the present inventions (1) to (5), wherein the method for converting the monomer structure is to open the carbon-carbon double bond of the monomer structure, set it as a polymer bonded by a carbon-carbon single bond, and set it as a compound with methyl groups introduced at both ends.

[0077] The present invention (7) is: a method for predicting the physical properties of a polymer according to any combination of any one of the present inventions (1) to (6), wherein the step 2 is: converting the monomer structure itself or converting it into a polymer of the monomer structure into an expression form that can be recognized by a computer.

[0078] The present invention (8) is: a polymer having an intermediate water content of 0.01 to 0.15 as calculated by the polymer property prediction method according to any combination of the present inventions (1) to (7).

[0079] The present invention (9) is: the polymer according to the present invention (8), wherein the polymer is a polymer obtained by polymerizing monomers represented by the SMILES expression form of Tables 1-1 to 1-3.

[0080] The present invention (10) is: the polymer according to the present invention (8) or (9), wherein the polymer has selective adhesion ability to specific cells.

[0081] The present invention (11) is: the polymer according to the present invention (10), wherein the specific cells are cancer cells.

[0082] The present invention (12) is: a polymer having an intermediate water content of 0.22 to 0.5 as calculated by the polymer property prediction method according to any combination of any one of the present inventions (1) to (7).

[0083] The present invention (13) is: the polymer according to the present invention (12), wherein the polymer is a polymer obtained by polymerizing monomers represented by the SMILES expression of Tables 2-1 to 2-7.

[0084] The present invention (14) is: the polymer according to the present invention (12) or (13), wherein the polymer is biocompatible.

Claims

1. A method for predicting physical properties of a polymer, characterized in that: It includes the following steps: Step 1: Obtain the monomer structure from the polymer structure, Step 2: The monomer structure itself or after conversion is further converted into an expression form that can be recognized by a computer, Step 3: Calculate the descriptor based on the expression form that the computer can recognize. Step 4: Obtain the physical properties of the polymer and use them as the target variable, use the descriptor as the explanatory variable, perform regression calculations, and create a regression model. Step 5: Based on the expressions that the computer can recognize, calculate the expression form of the new monomer. Step 6: Calculate the descriptor of the new monomer expression itself or after conversion, Step 7: Apply the descriptor of the new monomer to the regression model to predict the physical properties of the polymer polymerized from the new monomer.

2. The method for predicting the physical properties of a polymer according to claim 1, wherein: The expression form recognizable by the computer is a SMILES expression form or a MOL file expression form.

3. The method for predicting the physical properties of a polymer according to claim 1, wherein: The regression calculation is selected from at least one of multiple regression calculation, support vector regression calculation, decision tree calculation, random forest regression calculation, XGBoost regression calculation, LightGBM regression calculation, CatBoost regression calculation, Gaussian process regression calculation, neural network and deep learning.

4. The method for predicting the physical properties of a polymer according to claim 1, wherein: The algorithm for calculating the expression form of the new monomer is at least one selected from BRICS, RECAP and DOCS.

5. The method for predicting the physical properties of a polymer according to claim 1, wherein: The physical property of the polymer is at least one selected from glass transition temperature, melting point, refractive index, contact angle, surface free energy, density, water absorption, amount of intermediate water, amount of bound water, selective adhesion ability to specific cells and biocompatibility.

6. The method for predicting the physical properties of a polymer according to claim 1, wherein: The method for converting the monomer structure is a method of opening the carbon-carbon double bond of the monomer structure to form a polymer bonded by a carbon-carbon single bond and introducing methyl groups at both ends thereof.

7. The method for predicting the physical properties of a polymer according to claim 1, wherein: The step 2 is: converting the monomer structure itself or converting it into a polymer of the monomer structure into an expression form that can be recognized by a computer.

8. A polymer, characterized in that The polymer has an intermediate water content of 0.01 to 0.15 as calculated by the polymer property prediction method according to claim 1.

9. The polymer according to claim 8, wherein The polymer is a polymer obtained by polymerizing monomers expressed in the SMILES form of Tables 1-1 to 1-3.

10. The polymer according to claim 8, wherein The polymer has selective adhesion capability to specific cells.

11. The polymer according to claim 10, wherein The specific cells are cancer cells.

12. A polymer, characterized in that This is a polymer having an intermediate water content of 0.22 to 0.5 as calculated by the polymer property prediction method according to claim 1.

13. The polymer according to claim 12, wherein The polymer is a polymer obtained by polymerizing monomers expressed in the SMILES form of Tables 2-1 to 2-7.

14. The polymer according to claim 12, wherein The polymer is biocompatible.

Images