Compound safety risk screening method and system

By building a compound safety risk screening platform, using mass spectrometry methods and big data technology, standard spectral databases and screening models are established, the problem of risk substance detection in cosmetics is solved, efficient screening and early warning of compounds is achieved, and regulatory capabilities are improved .

CN119069033APending Publication Date: 2024-12-03CHINA INST FOR FOOD & DRUG CONTROL (MEDICAL DEVICE STANDARDS MANAGEMENT CENT OF THE STATE FOOD & DRUG ADMINISTRATION CHINA GENERAL INST FOR MEDICAL PROD INSPECTION)
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411119884.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The existing cosmetic testing methods cannot promptly follow up on banned substances illegally added by enterprises to products, resulting in increased supervision difficulties and ineffective screening of risky substances in compounds, affecting residents' health.

Method used

Based on mass spectrometry methods, the Internet, and big data technology, a compound safety risk screening platform is built, a high-resolution and low-resolution standard spectral library is established, and a data screening model is established through high-resolution and low-resolution algorithms are used to realize the risk of compound risk substances. Preliminary screening and early warning.

Benefits of technology

The preliminary screening of known or unknown substances in the compounds has been achieved, the scope of detection of risk substances has been expanded, and a high-throughput risk substance screening platform has been provided, which has improved the detection efficiency and accuracy, and has played an early warning role.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119069033A_ABST
    Figure CN119069033A_ABST
Patent Text Reader

Abstract

The invention discloses a compound safety risk screening method and system, and relates to the technical field of compound screening, and the method comprises the following steps: constructing a high-resolution standard spectrum library and a low-resolution standard spectrum library of risk substances; the spectrum library comprises retention time, parent ions, product ions and signal intensity information of the compounds; establishing a data screening model through a high-resolution algorithm and a low-resolution algorithm; uploading data of a to-be-detected compound, and comparing the data of the to-be-detected compound with the spectrum library by using the data screening model to determine whether the to-be-detected compound contains a corresponding risk substance; and returning a screening result to the user. On the basis of a mass spectrum method, the internet, big data and other technologies, a high-resolution standard spectrum library and a low-resolution standard spectrum library of the risky substances are constructed, and a compound screening model is established through a combinatorial algorithm, so that preliminary screening of the risky substances of the compounds is achieved, preliminary screening can be conducted on known or unknown substances in the compounds, and then the early warning effect is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of compound screening, and more particularly to a compound safety risk screening method and system. Background Art

[0002] With the rapid development of the economy, residents' demand for cosmetics is growing. Companies illegally add banned substances to their products, making supervision more difficult and preventing cosmetics-related testing methods from being updated in a timely manner.

[0004] The "Technical Specifications for Safety of Cosmetics" (2015 edition) clearly stipulates that the substances prohibited from being added include glucocorticoids (corticosteroids), sex hormones, antibiotics, anesthetics, psychotropic drugs, anti-infective drugs, antihistamines, etc. These drugs directly act on the surface of human skin, which will cause a series of adverse reactions and have a significant impact on people's physical health. Hormone drugs are clinically used to treat inflammation, allergic reactions, etc., but their direct action on the skin surface will cause the skin to become thinner and darker, and will also cause adverse reactions such as hormone-dependent dermatitis; anti-infective and antibiotic drugs are mainly used to treat infections caused by bacteria. Their long-term use will cause symptoms such as contact dermatitis and allergies, and are prone to drug resistance; anesthetics are often used for local or general anesthesia in surgery. The abuse of anesthetics will damage the body's higher nervous tissue and cause mental and psychological abnormalities. Alkaloid compounds are a class of natural products found in plants. Some alkaloids are toxic. After human contact, they will cause a series of adverse effects on physical and mental health, such as allergies, anemia, local anesthesia, etc. Pesticide residues entering the human body can cause various chronic or acute diseases, seriously affecting human health. Allergenic flavors and fragrances can give cosmetics a pleasant fragrance and can also mask the taste of the product itself. However, some flavors and fragrances are allergenic and can cause allergic reactions after adding them. Perfluorinated compounds are a class of substances with hydrophobic structures. Long-term exposure to perfluorinated substances increases the risk of human disease. Prostaglandin drugs are clinically used to treat diseases such as glaucoma and have the side effect of promoting eyelash growth, but long-term use can cause a series of adverse reactions in and around the eyes.

[0005] Therefore, it is urgent to establish a high-throughput risk substance screening platform, expand the scope of risk substance detection, and break through the limitations of existing detection methods. Summary of the invention

[0006] In view of this, the present invention provides a compound safety risk screening method and system. Based on mass spectrometry, the Internet, big data and other technologies, a compound safety risk substance screening platform is constructed. The platform covers a variety of high-resolution and low-resolution standard mass spectral libraries of risk substances. The spectral library contains information such as the retention time, parent ion, product ion and its signal intensity of the compound. On this basis, a compound screening model is established through a combined algorithm to achieve preliminary screening of compound risk substances. Preliminary screening of known or unknown substances in the compound can be performed, thereby playing an early warning role.

[0007] In order to achieve the above object, the present invention adopts the following technical scheme: a compound safety risk screening method, comprising:

[0008] Construct high-resolution and low-resolution standard spectral libraries for risk substances;

[0009] Establish a data screening model through high-resolution algorithms and low-resolution algorithms;

[0010] Upload the data of the compounds to be tested, and use the data screening model to compare the data of the compounds to be tested with the spectral library to determine whether they contain the corresponding risk substances;

[0011] The screening results are returned to the user.

[0012] Preferably, the high-resolution algorithm comprises:

[0013] Check whether the parent ion of the extracted secondary spectrum is the same as the parent ion compared with the library; if the parent ion error is within m, they are considered to be the same, and then compare their product ions. If the product ion error is within f, it is considered a successful match;

[0014] If the standard library product ion is greater than 1, the parent ion is the same and at least two secondary product ions are the same, then the two compounds are judged to be the same compound; this can screen out some false matching results.

[0015] The similarity is calculated by the improved SS combination algorithm, and then the optimal matching result is selected by combining the relative retention time difference.

[0016] Preferably, the similarity is calculated by an improved SS combination algorithm, and the calculation formula is as follows:

[0017]

[0018] Among them, S D (U ω ,V ω ) represents the peak ratio algorithm, S C (U ω ,V ω) represents the cosine similarity algorithm, V is the compound in the library, U is the compound to be detected, ω is the mass-to-charge ratio and peak intensity information; Q represents the reference spectrum, R represents the unknown spectrum, N R is the number of mass-to-charge ratios in the unknown spectrum whose peak intensity is not zero, N Q∩R is the number of mass-to-charge ratios with non-zero peak intensities contained in both the unknown and reference spectra.

[0019] Preferably, the cosine similarity algorithm calculation formula is as follows:

[0020]

[0021] Where V is the compound in the library, U is the compound to be detected, ω is the mass-to-charge ratio and peak intensity information, U and V are matrices composed of ω, ω is obtained by multiplying the mass-to-charge ratio and relative peak intensity of the compound by the exponent of a weight factor, and the calculation formula of ω is as follows:

[0022]

[0023] Among them, x and y are weight factors, α n is the data of the nth mass-to-charge ratio in the mass spectrum, β n is the peak intensity value corresponding to the nth mass-to-charge ratio;

[0024] The calculation formula of the peak ratio algorithm is as follows:

[0025]

[0026] In the formula, u i 、v i are non-zero peaks of the same mass-to-charge ratio;

[0027] u i The peak value is less than v i When , n = 1; otherwise n = -1.

[0028] Preferably, the low-resolution algorithm comprises:

[0029] According to the importance of compound precursor ions, product ions and retention times, the data of the compound to be detected and the data in the spectral library are compared in turn, and the matching results are comprehensively considered to obtain the final matching results.

[0030] Preferably, the order of importance from most important to least important is precursor ion>product ion>retention time.

[0031] Preferably, when the mass difference between the parent ions of the compound to be detected and the matching compound is less than a first threshold, the match is successful;

[0032] Compare the product ions of the compound to be detected and the matching compound. When the product ions of the compound to be detected are all in the product ions of the matching compound or the product ions of the matching compound are all in the product ions of the compound to be detected, the match is successful; if the above conditions are not met, the product ions of the compound to be detected are matched one by one with the product ions of the matching compound, and the size of each product ion is compared to confirm whether the match is successful based on the similarity;

[0033] Compare the retention times of the detected compound and the matching compound. The smaller the retention time difference, the better the matching result.

[0034] Preferably, confirming whether the match is successful based on the similarity of the product ions includes:

[0035] Calculating the mass difference between the product ion of the compound to be detected and the product ion of the matching compound, when the mass difference is less than or equal to a second threshold, the product ion of the compound to be detected and the product ion of the matching compound are corresponding;

[0036] Each product ion has a weight, which is 1 / the number of product ions. The final similarity is the sum of the similarity of each product ion multiplied by the weight.

[0037] The similarity of the product ions is that the smaller the mass difference, the greater the similarity. When the mass difference is greater than the second threshold, the similarity is 0; when the mass difference is between 0 and the second threshold, the similarity and the mass difference satisfy a linear function relationship.

[0038] A compound safety risk screening system, comprising:

[0039] Spectral library construction module, used to construct high-resolution and low-resolution standard spectral libraries of risk substances;

[0040] A model building module, used to build a data screening model through a high-resolution algorithm and a low-resolution algorithm;

[0041] The compound comparison module is used to upload the data of the compound to be tested and compare the data of the compound to be tested with the spectral library using the data screening model to determine whether it contains the corresponding risk substances;

[0042] The result output module is used to return the screening results to the user.

[0043] Through the above technical solutions, it can be known that compared with the prior art, the present invention discloses a method and system for screening compound safety risks, including: constructing high-resolution and low-resolution standard spectral libraries of risk substances; the spectral libraries include the retention time, parent ions, product ions and signal intensity information of the compounds; establishing a data screening model through high-resolution algorithms and low-resolution algorithms; uploading the data of the compounds to be detected, and using the data screening model to compare the data of the compounds to be detected with the spectral library to determine whether they contain corresponding risk substances; and returning the screening results to the user. Based on mass spectrometry, the Internet, big data and other technologies, the present invention proposes a method and system for screening compound safety risks, and constructs a compound safety risk substance screening platform, which covers more than 2,100 high-resolution and low-resolution standard mass spectral libraries of risk substances. The spectral library contains information such as the retention time, parent ions, product ions and signal intensity of the compounds. On this basis, a compound screening model is established through a combined algorithm, thereby realizing the preliminary screening of compound risk substances, and preliminary screening of known or unknown substances in the compounds can be performed, thereby playing an early warning role. The present invention realizes basic query and management of risk substances, including basic chemical information, detection standards, compound residue limit standards, and spectral information of risk substances; realizes screening function, and users can select the instrument manufacturer for uploading data, the instrument resolution and analysis method used, and screen the detection data uploaded by users by calling the script of the matching model built at the bottom layer, and return the verified screening results to users for convenience of user analysis, and users can download and save the matching results. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0045] Figure 1 A schematic diagram of a process for screening compound safety risks provided by the present invention.

[0046] Figure 2 A schematic diagram of the structure of a compound safety risk screening system provided by the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] The embodiment of the present invention discloses a compound safety risk screening method and system. The embodiment of the present invention constructs a compound safety risk substance screening platform based on mass spectrometry, the Internet, big data and other technologies. The platform covers a variety of high-resolution and low-resolution standard mass spectral libraries of risk substances. The spectral library contains information such as the retention time, parent ion, product ion and its signal intensity of the compound. On this basis, a compound screening model is established through a combined algorithm to achieve preliminary screening of compound risk substances. Preliminary screening of known or unknown substances in the compound can be performed, thereby playing an early warning role.

[0049] The present invention discloses a method for screening compound safety risks. Figure 1 As shown, including:

[0050] Based on mass spectrometry, the Internet, big data and other technologies, high-resolution and low-resolution standard spectral libraries of risk substances are constructed;

[0051] The spectral library includes the retention time, precursor ion, product ion and signal intensity information of the compound;

[0052] Establish a data screening model through high-resolution algorithms and low-resolution algorithms;

[0053] Upload the data of the compounds to be tested, and use the data screening model to compare the data of the compounds to be tested with the spectral library to determine whether they contain the corresponding risk substances;

[0054] The screening results are returned to the user.

[0055] Specifically, the high-resolution algorithm includes:

[0056] Whether the parent ion of the extracted secondary spectrum is the same as the parent ion compared with the library; if the parent ion error is within m (m = 0.008), it is considered to be the same, and then the product ions are compared. If the product ion error is within f (f = 0.01), it is considered a successful match;

[0057] If the standard library product ion is greater than 1, the parent ion is the same and at least two secondary product ions are the same, then the two compounds are judged to be the same compound; this can screen out some false matching results.

[0058] The similarity is calculated by the improved SS combination algorithm, and then combined with the relative retention time difference, the result with higher similarity and lower relative retention time difference is selected as the optimal matching result.

[0059] Specifically, the similarity is calculated by the improved SS combination algorithm, and the calculation formula is as follows:

[0060]

[0061] Among them, S D (U ω ,V ω ) represents the peak ratio algorithm, S C (U ω ,V ω ) represents the cosine similarity algorithm, V is the compound in the spectral library, U is the compound to be detected, ω is the mass-to-charge ratio and peak intensity information; N R is the number of mass-to-charge ratios in the unknown spectrum whose peak intensity is not zero, N Q∩R It is the number of mass-to-charge ratios with non-zero peak intensities in both the unknown spectrum and the reference spectrum. Q represents the reference spectrum and R represents the unknown spectrum.

[0062] Specifically, the calculation formula of the cosine similarity algorithm is as follows:

[0063]

[0064] Where V is the compound in the library, U is the compound to be detected, ω is the mass-to-charge ratio and peak intensity information, U and V are matrices composed of ω, ω is obtained by multiplying the mass-to-charge ratio and relative peak intensity of the compound by the exponent of a weight factor, and the calculation formula of ω is as follows:

[0065]

[0066] Among them, x and y are weight factors, x = 1.3, y = 0.53, α and β represent the mass-to-charge ratio and relative peak intensity respectively; α n is the data of the nth mass-to-charge ratio in the mass spectrum, β n It is the peak intensity value corresponding to the nth mass-to-charge ratio; when n=1, it is the first mass-to-charge ratio, n=2, it is the second mass-to-nuclear ratio, and n=q is the qth mass-to-charge ratio.

[0067] The calculation formula of the peak ratio algorithm is as follows:

[0068]

[0069] In the formula, u i 、v i are non-zero peaks of the same mass-to-charge ratio;

[0070] u i The peak value is less than v i When , n = 1; otherwise n = -1.

[0071] Specifically, the low-resolution algorithm includes:

[0072] According to the importance of compound precursor ions, product ions and retention times, the data of the compound to be detected and the data in the spectral library are compared in turn, and the matching results are comprehensively considered to obtain the final matching results.

[0073] Specifically, the order of importance from most important to least important is parent ion > product ion > retention time.

[0074] Specifically, the parent ions of the compound to be detected and the matching compound are compared. Taking into account the certain errors between different instruments, the match is successful when the mass difference is less than the first threshold; the first threshold is 0.2Da. When the mass difference is within 0.2Da (that is, the difference between the parent ions of the unknown compound and the matching compound is less than 0.2Da), the algorithm considers them to be a successful match.

[0075] Compare the product ions of the compound to be detected and the matching compound. When the product ions of the compound to be detected are all in the product ions of the matching compound or the product ions of the matching compound are all in the product ions of the compound to be detected (the product ions of the unknown compound and the matching compound can be regarded as two sets A and B, A is a subset of B or B is a subset of A), the match is successful; they are considered to be the same compound, that is, the similarity is 1. If the above situation is not met, the product ions of the compound to be detected are matched one by one with the product ions of the matching compound, and the size of each product ion is compared to confirm whether the match is successful based on the similarity;

[0076] Compare the retention times of the detected compound and the matching compound. The smaller the retention time difference, the better the matching result.

[0077] Specifically, whether the match is successful is confirmed based on the similarity of the product ions, including:

[0078] The mass difference between the product ions of the detected compound and the product ions of the matching compound is calculated. When the mass difference is less than or equal to the second threshold, the second threshold is 0.3 Da. If the mass difference is within 0.3 Da, it is considered that the product ions of the detected compound and the product ions of the matching compound correspond to each other.

[0079] Each product ion has a weight, which is 1 / the number of product ions. The final similarity is the sum of the similarity of each product ion multiplied by the weight.

[0080] The similarity of the product ions is that the smaller the mass difference, the greater the similarity. When the mass difference is greater than the second threshold, the similarity is 0. When the mass difference is between 0 and the second threshold, the similarity and the mass difference satisfy a linear function relationship, which roughly satisfies y=1-0.25x (y represents the similarity, and x represents the size of the difference).

[0081] If there is only one matching result, the similarity is not considered (because if the parent ion matches successfully, it is likely to be the same compound, but since the product ions selected for the detected compound and the matching compound may be different, the product ions produced by different instruments have different sizes, so when there is only one matching result and the similarity is low, it is also retained). When there are multiple matching results, the matching results with lower similarity are eliminated. If the similarity of all matching results is low, the matching result with the greatest similarity is retained.

[0082] In a specific embodiment of the present invention, the embodiment of the present invention is applied to the field of cosmetic technology to screen for risky compound substances in cosmetics.

[0083] Specifically, the high-resolution algorithm: first extract the parent ion of the secondary spectrum and the parent ion of the library comparison to see if they are the same. If the parent ion error is within 0.008, they are considered the same, and further compare their fragment ions. If the fragment ion error is within 0.01, it is considered a successful match. If the standard library fragment ion is greater than 1, the two compounds can be determined to be the same compound only when the parent ions are the same and at least two secondary fragment ions are the same. This can filter out some wrong matching results. The similarity is calculated through the improved SS combination algorithm, and then combined with the relative retention time difference, the result with higher similarity and lower relative retention time difference is selected as the optimal matching result.

[0084] Calculate the similarity score: Calculate based on the screening comparison algorithm improved by the SS combination algorithm. The SS combination algorithm is the SS combination algorithm proposed by Stein and Scott, including the cosine similarity algorithm (also called the weighted dot product algorithm) S C (U ω ,V ω ) and peak ratio algorithm S D (U ω ,V ω ). The calculation formula of the cosine similarity algorithm is as follows:

[0085]

[0086] Where V is the compound in the library, U is the unknown compound, ω is the mass-to-charge ratio and peak intensity information, U and V are matrices composed of ω, ω is obtained by multiplying the mass-to-charge ratio and relative peak intensity of the compound by the exponent of a weight factor, and the calculation formula of ω is as follows:

[0087]

[0088] Among them, x and y are weight factors, x = 1.3, y = 0.53, α and β represent the mass-to-charge ratio and relative peak intensity respectively; α n is the data of the nth mass-to-charge ratio in the mass spectrum, β nIt is the peak intensity value corresponding to the nth mass-to-charge ratio; when n=1, it is the first mass-to-charge ratio, n=2, it is the second mass-to-nuclear ratio, and n=q is the qth mass-to-charge ratio.

[0089] The calculation formula of the peak ratio algorithm is as follows:

[0090]

[0091] Where u i , v i are non-zero peaks with the same mass-to-charge ratio. When the former peak ratio is smaller than the latter, n = 1, otherwise n = -1. R is the number of mass-to-charge ratios in the unknown spectrum whose peak intensity is not zero, N Q∩R It is the number of mass-to-charge ratios with non-zero peak intensities in both the unknown spectrum and the reference spectrum. Q represents the reference spectrum and R represents the unknown spectrum.

[0092] Finally, S C and S D Multiply them by the corresponding weights and combine them together to calculate the final similarity. The calculation formula is as follows:

[0093]

[0094] Compared with the SS combined algorithm proposed by Stein and Scott, the improved combined algorithm tends to use peak ratio calculation when the similarity of mass spectra is low and the intensities of the same mass-to-charge ratio in different spectra will differ greatly. When the similarity is high, the number of the same mass-to-charge ratio increases and the difference between the corresponding intensities of the same mass-to-charge ratio decreases. At this time, it tends to use cosine similarity calculation, which can further improve the similarity between mass spectra.

[0095] Low-resolution algorithm: Based on the data parsed from the original file and the data in the currently built spectral library, the data that can be used to design the algorithm include parent ion, 1 to 3 product ions, and retention time (the retention time in the spectral library is temporarily missing). The order of importance from most important to least important is parent ion --> product ion --> retention time.

[0096] The first step is to compare the parent ions of the unknown compound and the matching compound. Taking into account the certain errors between different instruments, if the mass difference is within 0.2Da (that is, the difference between the parent ions of the unknown compound and the matching compound is less than 0.2Da), the algorithm considers them to be a successful match.

[0097] The second step is to compare their product ions. When the product ions of the unknown compound are all in the product ions of the matching compound or the product ions of the matching compound are all in the product ions of the unknown compound (the product ions of the unknown compound and the matching compound can be regarded as two sets A and B, A is a subset of B or B is a subset of A), they are considered to be the same compound, that is, the similarity is 1. If the above situation is not met, the product ions of the unknown compound and the matching compound are matched one by one, that is, the size of each product ion is compared. Considering that there are certain errors between different instruments, they are considered to be corresponding within a mass difference of 0.3Da. Each product ion has a weight and is 1 / the number of product ions. The final similarity is the sum of the similarity of each product ion multiplied by the weight. The similarity of the product ions is that the smaller the mass difference, the greater the similarity. When the mass difference is greater than 0.3Da, the similarity is 0; when the mass difference is between 0 and 0.3Da, the similarity roughly satisfies y=1-0.25x (y represents the similarity, and x represents the size of the difference).

[0098] The third step is to consider the retention time difference. The smaller the retention time difference, the better the matching result. In the case of close similarity, the smaller the retention time difference is, the better the matching result is.

[0099] Finally, the matching results are comprehensively considered. If there is only one matching result, the similarity is not considered (because if the parent ion matches successfully, it is largely the same compound, but because the product ions selected for the unknown compound and the matching compound may be different, and the product ions produced by different instruments have different sizes, so when there is only one matching result and the similarity is low, it is retained). When there are multiple matching results, the matching results with lower similarity are eliminated. If the similarity of all matching results is low, the one with the largest similarity is retained.

[0100] The embodiment of the present invention is based on mass spectrometry, the Internet, big data and other technologies to build a high-resolution and low-resolution standard mass spectral library of more than 2,100 cosmetic risk substances. The library contains information such as the retention time, parent ion, fragment ion and its signal intensity of the compound. On this basis, a data screening model is established through a combined algorithm. Researchers and monitoring personnel can upload the spectrum file generated by the instrument online, and use the model to compare to determine whether it contains the corresponding risk substance, thereby realizing the preliminary screening of cosmetic risk substances, which is used to determine whether the risk substance exists in the cosmetic matrix, and then play an early warning role. The embodiment of the present invention realizes the basic query and management of cosmetic risk substances, including the basic chemical information, detection standards, compound residue limit standards, and spectral information of risk substances; to realize the screening function, the user can select the instrument manufacturer of the uploaded data, the instrument resolution and analysis method used, and the test data uploaded by the user is screened by calling the script of the matching model built at the bottom layer, and the verified screening results are returned to the user, which is convenient for the user to analyze, and the user can download and save the matching results.

[0101] The embodiment of the present invention also discloses a compound safety risk screening system, such as Figure 2 As shown, including:

[0102] Spectral library construction module, used to construct high-resolution and low-resolution standard spectral libraries of risk substances;

[0103] A model building module, used to build a data screening model through a high-resolution algorithm and a low-resolution algorithm;

[0104] The compound comparison module is used to upload the data of the compound to be tested and compare the data of the compound to be tested with the spectral library using the data screening model to determine whether it contains the corresponding risk substances;

[0105] The result output module is used to return the screening results to the user.

[0106] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0107] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for screening compound safety risks, It is characterized in that include: Construct high-resolution and low-resolution standard spectral libraries for risk substances; Establish a data screening model through high-resolution algorithms and low-resolution algorithms; Upload the data of the compounds to be tested, and use the data screening model to compare the data of the compounds to be tested with the spectral library to determine whether they contain the corresponding risk substances; The screening results are returned to the user.

2. A compound safety risk screening method according to claim 1, It is characterized in that The high-resolution algorithm includes: Check whether the parent ion of the extracted secondary spectrum is the same as the parent ion compared with the library; if the parent ion error is within m, they are considered to be the same, and then compare their product ions. If the product ion error is within f, it is considered a successful match; If the standard library product ion is greater than 1, the parent ions are the same and at least two secondary product ions are the same, then the two compounds are judged to be the same compound; The similarity is calculated by the improved SS combination algorithm, and then the optimal matching result is selected by combining the relative retention time difference.

3. A compound safety risk screening method according to claim 2, It is characterized in that The similarity is calculated by the improved SS combination algorithm, and the calculation formula is as follows: Among them, S D (U ω ,V ω ) represents the peak ratio algorithm, S C (U ω ,V ω ) represents the cosine similarity algorithm, V is the compound in the library, U is the compound to be detected, ω is the mass-to-charge ratio and peak intensity information; Q represents the reference spectrum, R represents the unknown spectrum, N R is the number of mass-to-charge ratios with non-zero peak intensities in the unknown spectrum, N Q∩R is the number of mass-to-charge ratios with non-zero peak intensities contained in both the unknown and reference spectra.

4. A compound safety risk screening method according to claim 3, It is characterized in that The calculation formula of the cosine similarity algorithm is as follows: Where V is the compound in the library, U is the compound to be detected, ω is the mass-to-charge ratio and peak intensity information, U and V are matrices composed of ω, ω is obtained by multiplying the mass-to-charge ratio and relative peak intensity of the compound by the exponent of a weight factor, and the calculation formula of ω is as follows: Among them, x and y are weight factors, α n is the data of the nth mass-to-charge ratio in the mass spectrum, β n is the peak intensity value corresponding to the nth mass-to-charge ratio; The calculation formula of the peak ratio algorithm is as follows: In the formula, u i 、v i are non-zero peaks of the same mass-to-charge ratio; u i The peak value is less than v i When , n = 1; otherwise n = -1.

5. A compound safety risk screening method according to claim 1, It is characterized in that The low-resolution algorithm includes: According to the importance of compound precursor ions, product ions and retention times, the data of the compound to be detected and the data in the spectral library are compared in turn, and the matching results are comprehensively considered to obtain the final matching results.

6. A compound safety risk screening method according to claim 5, It is characterized in that The order of importance from most important to least important is precursor ion > product ion > retention time.

7. A compound safety risk screening method according to claim 5, It is characterized in that Comparing the parent ions of the compound to be detected and the matching compound, when the mass difference is less than the first threshold, the match is successful; Compare the product ions of the compound to be detected and the matching compound. When the product ions of the compound to be detected are all in the product ions of the matching compound or the product ions of the matching compound are all in the product ions of the compound to be detected, the match is successful; if the above conditions are not met, the product ions of the compound to be detected are matched one by one with the product ions of the matching compound, and the size of each product ion is compared to confirm whether the match is successful based on the similarity; Compare the retention times of the detected compound and the matching compound. The smaller the retention time difference, the better the matching result.

8. A compound safety risk screening method according to claim 7, It is characterized in that Confirm whether the match is successful based on the similarity of the product ions, including: Calculating the mass difference between the product ion of the compound to be detected and the product ion of the matching compound, when the mass difference is less than or equal to a second threshold, the product ion of the compound to be detected and the product ion of the matching compound are corresponding; Each product ion has a weight, which is 1 / the number of product ions. The final similarity is the sum of the similarity of each product ion multiplied by the weight. The similarity of the product ions is that the smaller the mass difference, the greater the similarity. When the mass difference is greater than the second threshold, the similarity is 0; when the mass difference is between 0 and the second threshold, the similarity and the mass difference satisfy a linear function relationship.

9. A compound safety risk screening system, It is characterized in that include: Spectral library construction module, used to construct high-resolution and low-resolution standard spectral libraries of risk substances; A model building module, used to build a data screening model through a high-resolution algorithm and a low-resolution algorithm; The compound comparison module is used to upload the data of the compounds to be tested and compare the data of the compounds to be tested with the spectral library using the data screening model to determine whether the corresponding risk substances are contained; The result output module is used to return the screening results to the user.