Methods for screening odor inhibitors based on olfactory receptors and intermolecular interactions

Abstract

This invention discloses a method for screening odor inhibitors based on olfactory receptors and intermolecular interactions in the field of deodorization technology. The method includes: using an olfactory receptor database, multi-level screening or machine learning to predict olfactory receptors that a target odor pollutant can bind to or activate; multi-level screening or machine learning to predict potential inhibitors of the olfactory receptors; calculating the binding energy between the olfactory receptors and the target odor pollutant, and the screened or predicted inhibitors; preferentially selecting potential inhibitors with binding energies lower than those of the target odor pollutant; and using an electronic nose to test the odor-reducing effect of the potential inhibitors on the target odor pollutant. This invention proposes a screening method for odor inhibitors from the perspective of olfactory mechanisms. By utilizing the competitive and inhibitory effects of substances on olfactory receptors, it weakens the olfactory response of the target odor pollutant, thereby achieving the effect of odor suppression. Compared with manual olfactory screening methods, this method can effectively improve the development efficiency of deodorants and reduce workload, cost, and health risks.

Description

Technical Field

[0001] This invention relates to the field of deodorization technology, specifically a method for screening odor inhibitors based on the interaction between olfactory receptors and odor molecules. Background Technology

[0002] Odors are easily perceived and severely impact people's lives and work. Sources of odors are widespread, including garbage, sewage, feces, industrial production and livestock farming, as well as bacterial metabolism on the human body. Odors generated and released by garbage disposal facilities pollute the surrounding environment, leading to environmental complaints and the "not in my backyard" (NIMBY) effect, affecting the site selection, implementation, and operation of these facilities, and even social stability. Toilet odors not only affect the user experience but also the surrounding environment. Due to limitations in equipment space, energy consumption, and operating costs, centralized odor collection and treatment is unsuitable for facilities such as garbage rooms, small garbage transfer stations, and toilets. Furthermore, body odor affects a person's psychology and social interactions, causing anxiety, low self-esteem, social phobia, and even impacting work. Therefore, the development of odor inhibitors is of great significance.

[0003] The mechanisms of action of existing odor inhibitors include: physical adsorption / absorption, chemical reaction, biological action, odor masking, and their combined effects. Based on these mechanisms, some odor inhibitors have been developed. However, due to limitations in the rates of phase transfer, chemical reaction, and biological metabolism, physical adsorption / absorption, chemical reaction, and biological action usually do not produce immediate results; their onset time is measured in hours or even days.

[0004] Patent document CN116637049B discloses a pet disinfectant and deodorizer that uses plant extracts and antibacterial agents for pet disinfection and deodorization. After 24 hours of use, the removal rate of hydrogen sulfide, ammonia, and trimethylamine reaches approximately 90%. Patent document CN118634517B discloses a preparation method for a plant-based disinfectant and deodorizer based on optimized preparation parameters. After 2 hours of use, the removal rates of ammonia and hydrogen sulfide are less than 40% and 20%, respectively. Patent document CN119771158B provides a microbial compound deodorizer and its preparation method. After one day of use, the maximum removal rate of ammonia and hydrogen sulfide is less than 60%. Therefore, there is an urgent need to develop odor inhibitors that can rapidly improve environmental and human odor.

[0005] Odor inhibitors can rapidly improve olfactory perception by competing with odor pollutants for olfactory receptors and inhibiting their response. Existing odor inhibitors mostly use plant essential oils or fragrances, selecting their components solely based on the freshness and persistence of the scent. However, there may be as many as 40 billion odor substances. Verifying their odor-inhibiting effects on odor pollutants one by one through manual olfactory identification is labor-intensive, inefficient, and poses health risks. Therefore, there is currently a lack of efficient methods for screening odor inhibitors based on olfactory mechanisms (interactions between olfactory receptors and odor molecules). Summary of the Invention

[0006] The purpose of this invention is to provide a method for screening odor inhibitors based on the interaction between olfactory receptors and odor molecules, thereby improving the screening efficiency of odor inhibitors.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for screening odor inhibitors based on olfactory receptors and intermolecular interactions includes:

[0009] Step 1: Download the target file of olfactory receptor-ligand pairing data from the olfactory receptor database, and use the first multi-level screening method or the first machine learning prediction method to obtain the olfactory receptors that the target malodorous pollutants can bind to or activate.

[0010] Step 2: Use a second multi-level screening method or a second machine learning prediction method to obtain potential inhibitors of olfactory receptors, calculate the binding energy between olfactory receptors and target malodorous pollutants and the potential inhibitors, and screen for preferred potential inhibitors based on the binding energy.

[0011] Step 3: Evaluate the effectiveness of the potential inhibitors in reducing the odor of the target malodorous pollutants using the electronic nose test.

[0012] As a further aspect of the present invention: the olfactory receptor database includes the M2OR olfactory receptor database.

[0013] As a further aspect of the present invention: the interaction between the olfactory receptor and the odor molecule includes the binding, activation, and inhibition effects of the odor molecule on the olfactory receptor, as well as the competitive effect of the odor molecule when binding to the olfactory receptor.

[0014] As a further aspect of the present invention, the identifier of the target malodorous pollutant includes one or more of SMILES (Simplified molecular input line entry system), Molecule Name, CID (Chemical Identifier), CAS (Chemical Identifier), and InChIKey (Chemical Identifier).

[0015] As a further aspect of the present invention, the identifier of the olfactory receptor includes one or more of UniProt ID (protein identifier), Gene Name, Mutation, and Sequence.

[0016] As a further aspect of the present invention: the first multi-level screening method includes: first-level screening, second-level screening, and third-level screening;

[0017] The first-level screening includes: searching for the identifier of the target malodorous pollutant in the target file of the olfactory receptor-ligand pairing data, and screening for olfactory receptors that can bind to the target malodorous pollutant using Species = homo sapiens as the condition.

[0018] The second-level screening includes: based on the first-level screening, using Responsive = 1 as a condition, screening to obtain the olfactory receptors that the target malodorous pollutant may activate;

[0019] The third-level screening includes: based on the second-level screening, using the condition that Mixture (whether it is a mixture) = mono (pure substance) as a condition, screening to obtain the olfactory receptors that can be activated by the pure substance of the target malodorous pollutant.

[0020] As a further aspect of the present invention: the first machine learning prediction method includes:

[0021] Based on the olfactory receptor-ligand pairing data in the olfactory receptor database, a machine learning algorithm is used to predict olfactory receptors with Responsive=1 by using the property parameters of the target malodorous pollutant and the identifiers of olfactory receptors in the olfactory receptor database as input features, thereby screening out potential olfactory receptors for the target malodorous pollutant.

[0022] As a further aspect of the present invention, the property parameters include: MACCS (Molecular Access System) fingerprint, Henry coefficient, and octanol-water partition coefficient.

[0023] As a further aspect of the present invention: the second multi-level screening method includes: fourth-level screening, fifth-level screening, sixth-level screening, seventh-level screening and eighth-level screening;

[0024] The fourth level of screening includes: searching for the olfactory receptors obtained in the first step in the target file of the olfactory receptor-ligand pairing data, and screening for substances that can bind to the olfactory receptors;

[0025] The fifth level of screening includes: based on the fourth level of screening, screening for the first potential inhibitor with Responsive=0 as the condition;

[0026] The sixth level of screening includes: based on the fifth level of screening, screening for second potential inhibitors using Mixture=mono as the condition;

[0027] The seventh level of screening includes: based on the sixth level of screening, screening from the second potential inhibitors to obtain the third potential inhibitors common to the olfactory receptors obtained in the first step;

[0028] The eighth level of screening includes: calculating the binding energy of the olfactory receptor obtained in the first step with the target malodorous pollutant and the potential inhibitor, and screening for a fourth potential inhibitor whose binding energy with the olfactory receptor is lower than that with the target malodorous pollutant.

[0029] As a further aspect of the present invention: the second machine learning prediction method includes:

[0030] Based on the olfactory receptor-ligand pairing data in the olfactory receptor database, a machine learning algorithm is used to predict substances with Response=0, using the property parameters of odor molecules and the identifiers of olfactory receptors obtained in the first step as input features, and to screen out the fifth potential inhibitor. The binding energy between the olfactory receptors obtained in the first step and the target malodorous pollutant and the fifth potential inhibitor is calculated, and a sixth potential inhibitor with a binding energy with the olfactory receptor lower than that with the target malodorous pollutant is screened out.

[0031] As a further aspect of the present invention, the machine learning algorithm includes convolutional neural networks and recurrent neural networks.

[0032] As a further aspect of the present invention: the electronic nose inspection method includes:

[0033] The potential inhibitor obtained in the second step is added to the target malodorous pollutant gas, and the odor concentration and / or intensity before and after the addition are detected using an electronic nose to evaluate the odor reduction effect of the potential inhibitor on the target malodorous pollutant.

[0034] As a further aspect of the present invention: the target file includes a CSV format file or an xlsx format file. When the target file is a CSV format file, it needs to be converted into an xlsx format file.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] This invention, from the perspective of odor perception mechanisms, is based on the binding, activation, and inhibition of substances on olfactory receptors, as well as the competitive interaction between substances on olfactory receptors. Utilizing machine learning methods, it develops a screening method for odor inhibitors. Compared to manual olfactory screening, this method improves the targeting and accuracy of odor inhibitor screening, expands the screening range of olfactory receptors and inhibitors, thereby effectively improving the efficiency of deodorant development and reducing workload, cost, and health risks. Attached Figure Description

[0037] Figure 1 This is a flowchart of the method framework of the present invention;

[0038] Figure 2 This is a flowchart illustrating the method steps of the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] Please see Figures 1-2 In this embodiment of the invention, a method for screening odor inhibitors based on olfactory receptors and interactions between odor molecules includes the following steps:

[0041] S1: Download the target file of olfactory receptor-ligand pairing data from the olfactory receptor database, and use the first multi-level screening method or the first machine learning prediction method to obtain the olfactory receptors that the target malodorous pollutants can bind to or activate.

[0042] S2: Use a second multi-level screening method or a second machine learning prediction method to obtain potential inhibitors of olfactory receptors, calculate the binding energy between olfactory receptors and target malodorous pollutants and the potential inhibitors, and screen for preferred potential inhibitors based on the binding energy.

[0043] S3: The electronic nose test method was used to evaluate the effect of the potential inhibitor on reducing the odor of the target malodorous pollutant.

[0044] Example 1:

[0045] This embodiment targets methanethiol (CS), a major odor-causing substance in processes such as waste collection and treatment, as the odor pollutant.

[0046] Download the olfactory receptor-ligand pairing data from the M2OR olfactory receptor database in CSV format, then convert it to XLSX format. Using the data filtering tool in the table, filter for human olfactory receptors that can be activated by methanethiol itself in the following columns: "SMILES" column, "SMILES=CS"; "Species" column, "Species=homo sapiens"; "Responsive" column, "Responsive=1"; and "Mixture" column, "Mixture=mono".

[0047] Next, in the "UniProt ID" column of file F1, select the UniProtID of the olfactory receptor obtained in the previous step. In the "Responsive" column, use "Responsive=0" as the condition, and in the "Mixture" column, use "Mixture=mono" as the condition to filter out potential odor inhibitors of methanethiol. The screening results for each olfactory receptor are generated into a table; find all compounds common to the table and identify the potential odor inhibitors common to these olfactory receptors.

[0048] Finally, methanethiol gas (concentration of 0.1 ppm) and a mixture of methanethiol and each common potential odor inhibitor (concentration of 0.1 ppm) were prepared. The odor concentration, intensity and pleasantness of these two types of gases were detected using an electronic nose or manual olfaction. The effect of the screened potential odor inhibitors on improving the odor of methanethiol was evaluated, thereby discovering compounds that can effectively suppress the odor of methanethiol.

[0049] Example 2:

[0050] This example uses methanethiol (CS) as an example.

[0051] Download the olfactory receptor-ligand pairing data from the M2OR olfactory receptor database in CSV format, then convert it to XLSX format. Using the data filtering tool in the table, filter for human olfactory receptors that may be activated by methanethiol in the "SMILES" column with the condition "SMILES=CS", the "Species" column with the condition "Species=homo sapiens", and the "Responsive" column with the condition "Responsive=1".

[0052] Then, in table F2 of the file, in the UniProt ID column, select the UniProt ID of the olfactory receptor obtained in the previous step. In the "Responsive" column, with "Responsive=0" as the condition, filter out potential inhibitors, and generate a table for the screening results of each olfactory receptor; use Autodock software to calculate the binding energy of the olfactory receptor and methanethiol and potential inhibitors in each table; further, filter out potential odor inhibitors in each table whose binding energy with the olfactory receptor is lower than that with methanethiol.

[0053] Specifically, the steps for calculating the binding energy using Autodock software include: downloading the olfactory receptor's PDB file from the AlphaFold ProteinStructure Database; downloading the methanethiol and potential inhibitor's SDF files from the PubChem database; converting the SDF files of the methanethiol and potential inhibitor to PDB format using Open Babel software; importing the olfactory receptor's PDB file into Autodock software, performing dehydration and hydrogenation treatment to generate a PDBQT file; importing the methanethiol and potential inhibitor's PDB files into Autodock software, performing hydrogenation treatment to generate a PDBQT file; then, setting up the docking box, performing molecular docking according to the software instructions, and thus generating the binding energy between the olfactory receptor and the methanethiol and potential inhibitor.

[0054] Finally, methanethiol gas (concentration of 0.05 ppm) and a mixture of methanethiol and each screened potential odor inhibitor (concentration of 0.05 ppm) were prepared. The odor intensity of these two types of gases was detected using an electronic nose to evaluate the effect of the screened potential odor inhibitors on improving the odor of methanethiol, thereby discovering compounds that can effectively suppress the odor of methanethiol.

[0055] Compared to Example 1, this example removes the restriction of Mixture=mono when screening olfactory receptors for target malodorous pollutants, thus expanding the screening range of olfactory receptors; when screening potential inhibitors, it removes the restriction of Mixture=mono and the restriction of identifying potential inhibitors common to olfactory receptors, thus expanding the screening range of potential inhibitors; the concentrations of methanethiol and potential inhibitors in the mixed gas are adjusted according to the actual concentration of methanethiol in the environment, and only electronic nose is used to evaluate odor intensity.

[0056] Example 3:

[0057] This example uses methanethiol (SMILES is "CS") as an example.

[0058] Download the olfactory receptor-ligand pairing data in CSV format from the M2OR olfactory receptor database, and then convert it to XLSX format file F3. Based on file F3, convert the smiles of odor substances into MACCS fingerprints using RDKit, and construct a tabular file F4 containing the MACCS fingerprints of odor substances, Henry's Law coefficient, octanol-water partition coefficient, UniProt ID of olfactory receptors, and responsiveness. This serves as the dataset for the machine learning model, with a training and testing dataset ratio of 8:2. Using the compound's MACCS fingerprint, Henry's coefficient, octanol-water partition coefficient, and olfactory receptor UniProt ID as input features, a convolutional neural network algorithm is used to build model M1, which predicts the response between the olfactory receptor and the ligand, i.e., Responsive. Using the trained model M1, the MACCS fingerprint, Henry's coefficient, octanol-water partition coefficient of methanethiol, and the known UniProt ID of human olfactory receptors are input to predict the response between methanethiol and the olfactory receptor, generating a prediction result table file F5. Then, in the "Responsive" column of file F5, the olfactory receptor OR1 that methanethiol may activate is screened with "Responsive=1" as the condition.

[0059] Then, in the "UniProt ID" column of file F3, select the UniProt ID of olfactory receptor OR1. In the "Responsive" column, use "Responsive=0" as the screening condition to screen for potential odor inhibitors of methanethiol.

[0060] Finally, methanethiol gas (concentration of 0.5 ppm) and a mixture of methanethiol and each screened potential odor inhibitor (concentration of 0.5 ppm) were prepared. The odor concentrations of these two types of gases were detected using an electronic nose to evaluate the effect of the screened potential odor inhibitors on improving the odor of methanethiol, thereby identifying compounds that can effectively suppress the odor of methanethiol.

[0061] Compared to Example 2, this example uses a convolutional neural network algorithm to predict olfactory receptors based on known olfactory receptors, thus expanding the range of olfactory receptors that methanethiol may activate. The concentrations of methanethiol and potential inhibitors in the mixed gas are adjusted according to the actual methanethiol concentration in the environment, and odor concentration is used as an evaluation index.

[0062] Example 4:

[0063] This example uses methanethiol (CS) as an example.

[0064] Download the olfactory receptor-ligand pairing data from the M2OR olfactory receptor database in CSV format, and then convert it to XLSX format file F6. Based on file F6, convert the smiles of odor substances into MACCS fingerprints using RDKit, and create a tabular file F7 containing the MACCS fingerprints of odor substances, the UniProt IDs of olfactory receptors, and the responsiveness. This tabular file serves as the dataset for the machine learning model, with a training and testing dataset ratio of 7:3. Using the MACCS fingerprint of the compound and the UniProt ID of the olfactory receptor as input features, a recurrent neural network algorithm is used to build model M2, which predicts the response between the olfactory receptor and the ligand, i.e., Responsive. A first machine learning method is used to predict the olfactory receptor. Using the trained model M2, the MACCS fingerprint of methanethiol and the known UniProt ID of the human olfactory receptor are input to predict the response between methanethiol and the olfactory receptor, generating a prediction result table file F8. In the "Responsive" column of file F8, olfactory receptors OR2 that may be activated by methanethiol are screened with "Responsive=1". A second machine learning method is used to predict potential inhibitors. Using the trained model M2, the MACCS fingerprint of the compound in file F6 and the UniProt ID of the olfactory receptor OR2 are input. ID, predict the response between olfactory receptor OR2 and the compound, and obtain the prediction results table file F9. In the "Responsive" column of file F9, with "Responsive=0" as the condition, screen for potential inhibitors of olfactory receptor OR2, thereby obtaining potential odor inhibitors of methanethiol.

[0065] Finally, methanethiol gas (concentration of 0.5 ppm) and a mixture of methanethiol and each screened potential odor inhibitor (concentration of 0.5 ppm) were prepared. The odor intensity of these two types of gases was detected by electronic nose and manual olfaction to evaluate the effect of the screened potential odor inhibitors on improving the odor of methanethiol, thereby discovering compounds that can effectively suppress the odor of methanethiol.

[0066] Compared to Example 3, this example changes the prediction model algorithm for olfactory receptors, input features, and the ratio of training and testing datasets. Furthermore, it increases the prediction of olfactory receptor inhibitors, expands the screening range of potential olfactory receptor inhibitors, and uses an electronic nose and manual olfaction to evaluate the odor suppression effect.

[0067] Example 5:

[0068] This example uses a mixture of methanethiol (SMILES is CS) and acetic acid (SMILES is CC(=O)O).

[0069] Download the olfactory receptor-ligand pairing data from the M2OR olfactory receptor database in CSV format, then convert it to an XLSX file (F10). Using the data filtering tools in the table, filter the human olfactory receptors OR3 and OR4 that can be activated by methanethiol and acetic acid, respectively, in the SMILES column using "SMILES=CS" and "SMILES=CC(=O)O", the "Species" column using "Species=homo sapiens", the "Responsive" column using "Responsive=1", and the "Mixture" column using "Mixture=mono". Then, in the "UniProt ID" column of file F10, select the UniProt IDs for olfactory receptors OR3 and OR4. The ID is used to filter potential odor inhibitors of methanethiol and acetic acid in the "Responsive" column with "Responsive=0" and in the "Mixture" column with "Mixture=mono". The screening results for each olfactory receptor are generated into a table. Then, compounds common to all the tables are searched to identify potential odor inhibitors common to these olfactory receptors.

[0070] Finally, a mixture of methanethiol and acetic acid (each at a concentration of 0.1 ppm) and a mixture of methanethiol, acetic acid, and each potential odor inhibitor (each at a concentration of 0.1 ppm) were prepared. An electronic nose was used to detect the odor intensity of these two types of gases, and the effect of the screened potential odor inhibitors on improving the mixed odor of methanethiol and acetic acid was evaluated, thereby identifying compounds that can effectively suppress the mixed odor of methanethiol and acetic acid.

[0071] Compared to the previous four embodiments, this embodiment expands from a single methanethiol to a mixture of methanethiol and acetic acid, enhancing the feasibility of practical applications.

[0072] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for screening odor inhibitors based on the interaction between olfactory receptors and odor molecules, characterized by, The method comprises the following steps: The first step is to download the target file of olfactory receptor-ligand pairing data from the olfactory receptor database, and use the first multi-level screening method or the first machine learning prediction method to obtain the olfactory receptor that can bind or activate the target malodorous pollutant; The second step is to obtain the potential inhibitor of the olfactory receptor by using the second multi-level screening method or the second machine learning prediction method, to calculate the binding energy of the olfactory receptor, the target malodorous pollutant and the potential inhibitor, and to screen the preferred potential inhibitor based on the binding energy; The third step is to evaluate the effect of the potential inhibitor on the reduction of the odor of the target malodorous pollutant by using the electronic nose test method.

2. The method of screening odor inhibitors based on the intermolecular interactions between olfactory receptors and odor molecules according to claim 1, characterized in that: The olfactory receptor database comprises an M2OR olfactory receptor database.

3. The method of screening odor inhibitors based on the interaction between olfactory receptors and odor molecules according to claim 1, characterized in that: The interaction between the olfactory receptor and the odor molecule includes the combination, activation and inhibition of the odor molecule to the olfactory receptor, and the competition of the odor molecule when combined with the olfactory receptor.

4. The method of screening odor inhibitors based on the interaction between olfactory receptors and odor molecules according to claim 1, characterized in that: The identification of the target malodorous pollutant includes one or more of SMILES, Molecule Name, CID, CAS and InChIKey.

5. The method of screening odor inhibitors based on the interaction between olfactory receptors and odor molecules according to claim 1, characterized in that: The identification of the olfactory receptor includes one or more of UniProt ID, Gene Name, Mutation and Sequence.

6. The method of screening odor inhibitors based on the intermolecular interaction of olfactory receptors and odor molecules according to claim 1 or 4, characterized in that: The first multi-level screening method comprises first-level screening, second-level screening and third-level screening. The first-level screening comprises searching for the identification of the target malodorous pollutant in the target file of the olfactory receptor-ligand pairing data, and screening the olfactory receptor that can bind the target malodorous pollutant with the condition of Species=homo sapiens; The second-level screening comprises screening the olfactory receptor that can be activated by the target malodorous pollutant based on the first-level screening with the condition of Responsive=1; The third-level screening comprises screening the olfactory receptor that can be activated by the pure substance of the target malodorous pollutant based on the second-level screening with the condition of Mixture=mono.

7. The method of screening odor inhibitors based on the interaction between olfactory receptors and odor molecules according to claim 1, characterized in that: The first machine learning prediction method comprises: Based on the olfactory receptor-ligand pairing data of the olfactory receptor database, a machine learning algorithm is used to predict the Responsive=1 olfactory receptor with the property parameters of the target malodorous pollutant and the identification of the olfactory receptor in the olfactory receptor database as input features, and to screen the potential olfactory receptor of the target malodorous pollutant.

8. The method of claim 7, wherein the method is based on the intermolecular interactions between olfactory receptors and odorant molecules. The property parameters include MACCS fingerprint, Henry's coefficient and octanol-water partition coefficient.

9. The method of claim 1, wherein the method is for screening odor inhibitors based on the intermolecular interactions between olfactory receptors and odor molecules. The second multi-level screening method comprises fourth-level screening, fifth-level screening, sixth-level screening, seventh-level screening and eighth-level screening; The fourth-level screening comprises searching for the olfactory receptor obtained in the first step in the target file of the olfactory receptor-ligand pairing data, and screening the substance that can bind the olfactory receptor; The fifth-level screening comprises screening the first potential inhibitor based on the fourth-level screening with the condition of Responsive=0; The sixth level screening includes: on the basis of the fifth level screening, screening the second potential inhibitor with the condition of Mixture = mono; The seventh level screening includes: on the basis of the sixth level screening, screening the third potential inhibitor shared by the olfactory receptors obtained in the first step from the second potential inhibitors; The eighth level screening includes: calculating the binding energy of the olfactory receptors obtained in the first step with the target malodorous pollutants and the potential inhibitors, and screening the fourth potential inhibitor with the binding energy of the olfactory receptors lower than the binding energy of the target malodorous pollutants and the olfactory receptors.

10. The method of screening odor inhibitors based on the interaction between olfactory receptors and odor molecules according to claim 1, wherein The second machine learning prediction method includes: Based on the olfactory receptor-ligand pairing data of the olfactory receptor database, a machine learning algorithm is used to take the property parameters of the odor molecules and the identification of the olfactory receptors obtained in the first step as input features to predict the substances with Response = 0, and to screen the fifth potential inhibitor; the binding energy of the olfactory receptors obtained in the first step with the target malodorous pollutants and the fifth potential inhibitor is calculated, and the sixth potential inhibitor with the binding energy of the olfactory receptors lower than the binding energy of the target malodorous pollutants and the olfactory receptors is screened.

11. The method for screening odor inhibitors based on the interaction between olfactory receptors and odor molecules according to claim 7 or 10, characterized in that: The machine learning algorithm includes: convolutional neural network and recurrent neural network.

12. The method of screening odor inhibitors based on the interaction between olfactory receptors and odor molecules according to claim 1, characterized in that: The electronic nose test method includes: The target malodorous pollutant gas is added with the potential inhibitor obtained in the second step, and the electronic nose is used to detect the odor concentration and / or intensity before and after the addition to evaluate the effect of the potential inhibitor on the odor reduction of the target malodorous pollutant.

Images