Drug information storage method, system and medium for drug development process

By analyzing the local and global characteristics of the drug information matrix and adjusting the confusion parameters in the t-SNE algorithm, the problem of redundant storage of traditional Chinese medicine information in drug research and development is solved, and more efficient storage of drug information is achieved.

CN118737318BActive Publication Date: 2024-11-12YOUAN TECHNOLOGY (NANTONG) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411215882.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-11-12
Estimated Expiration
2044-09-02

AI Technical Summary

Technical Problem

In the process of drug research and development, drug information is massive and heterogeneous, resulting in redundant information occupying a large amount of storage space. The confusion parameters in existing data dimensionality reduction algorithms such as the t-distributed random neighborhood embedding algorithm (t-SNE) are difficult to determine. Influences the dimensionality reduction effect.

Method used

By analyzing the local and global characteristics of the drug information matrix, compute the component isolation group and information importance, combine the confusion default value in the t-SNE algorithm, adjust the confusion parameters to optimize the dimensionality reduction processing, and eliminate redundancy in drug information. data.

Benefits of technology

It effectively improves the efficiency of drug information storage, accurately eliminates redundant information, avoids the negative impact of too large or too small confusion parameter values ​​on the dimensionality reduction effect, and improves the efficiency of drug information storage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118737318B_ABST
    Figure CN118737318B_ABST
Patent Text Reader

Abstract

The present application relates to the field of digital data processing technology, and specifically to a method, system and medium for storing drug information in the process of drug development. The method includes: obtaining drug information of drug ingredients contained in each synthetic drug in the process of drug development, calculating the component separation degree of drug ingredients contained in each synthetic drug, and combining the types of drug ingredients contained in each synthetic drug to obtain information importance, and then obtain local importance; based on information importance combined with the average level of similarity between information vectors of each synthetic drug and other synthetic drugs, determine global importance; according to local importance and global importance, combined with the default value of perplexity in the t-SNE algorithm, obtain perplexity adjustment value, reduce the dimension of drug information data, and store the drug information data after dimension reduction. The present application can eliminate redundant information in drug information and improve the efficiency of drug information storage.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of digital data processing technology, and specifically to a drug information storage method, system and medium for drug research and development. Background Art

[0002] The drug development process usually goes through multiple stages and has a long development cycle. The drug information involved in the development process is often massive and heterogeneous, which makes drug information management in the drug development process highly complex and will greatly affect the progress of drug development.

[0003] By storing drug information, drug information in the research and development process can be effectively managed, so storing drug information in the drug research and development process has important practical significance. However, due to the massive and heterogeneous characteristics of drug information in the drug research and development process, there will be a large amount of redundant information in the drug information data, which will greatly occupy the information storage space, thereby affecting the efficiency of data storage.

[0004] In order to improve the efficiency of storing drug information, the existing technology often uses data dimensionality reduction algorithms to reduce the dimensionality of drug information and remove redundant information in drug information data, so as to improve the efficiency of drug information storage. For example, the t-distributed random neighbor embedding algorithm (t-SNE) can quickly remove redundant information in drug information data. However, the perplexity parameter in the t-distributed random neighbor embedding algorithm is difficult to determine. If the perplexity parameter is too large or too small, it will affect the effect of reducing the dimensionality of drug information, resulting in poor results in removing redundant information in drug information, and unable to effectively improve the efficiency of drug information storage. Summary of the invention

[0005] In order to solve the above technical problems, the purpose of this application is to provide a drug information storage method, system and medium for the drug development process. The technical solutions adopted are as follows:

[0006] In a first aspect, the present application embodiment provides a method for storing drug information in a drug development process, comprising the following steps:

[0007] Obtain drug information of drug ingredients contained in each synthetic drug during drug development to form information vectors of each synthetic drug;

[0008] Determine the component separation degree of each element in the information vector of each synthetic drug according to the difference between each element in the information vector of each synthetic drug and the core drug component of each synthetic drug;

[0009] According to the distribution disorder degree of the component separation degree of all elements in the information vector of each synthetic drug, combined with the types of drug ingredients contained in each synthetic drug, the information importance of each synthetic drug is obtained, and the local importance of the drug development process is determined through the information importance of all synthetic drugs in the drug development process;

[0010] Based on the difference between the information importance of each synthetic drug and other synthetic drugs, combined with the average similarity between the information vector of each synthetic drug and the information vector of other synthetic drugs, the global importance of the drug development process is determined;

[0011] According to the local importance and the global importance, combined with the default value of the perplexity in the t-SNE algorithm, the perplexity adjustment value is obtained, the drug information data is reduced in dimension, and the drug information data after the dimension reduction is stored.

[0012] Preferably, the process of determining the component separation degree of each element in the information vector of each synthetic drug is as follows:

[0013] Clustering all elements in the information vector of each synthetic drug respectively to obtain each core point in the information vector of each synthetic drug, and all core points in the information vector of each synthetic drug constitute the core component sequence of each synthetic drug;

[0014] The mean of the differences between each element in the information vector of each synthetic drug and all elements in the core component sequence of each synthetic drug is used as the component separation degree of each element in the information vector of each synthetic drug.

[0015] Preferably, the calculation formula for the information importance of each synthetic drug is:

[0016] , where is the information importance of the i-th synthetic drug, is the normalization function, is the number of types of drug ingredients contained in the i-th synthetic drug, is the information entropy of the component separation degree of all elements in the information vector of the i-th synthetic drug.

[0017] Preferably, the local importance in the drug development process is the average value of the information importance of all synthetic drugs in the drug development process.

[0018] Preferably, the global importance calculation formula in the drug development process is:

[0019] , where The global importance of the drug development process, is the information importance of the i-1th synthetic drug, is the mean similarity of drug ingredients of all recent synthetic drugs, To avoid values ​​with a denominator of zero.

[0020] Preferably, the similarity of drug ingredients of each recently synthesized drug is the average of the similarities between the information vectors of each recently synthesized drug and other synthesized drugs.

[0021] Preferably, the method of obtaining the perplexity adjustment value according to the local importance and the global importance in combination with the perplexity default value in the t-SNE algorithm further includes:

[0022] The sum of the adjustment coefficient and the number 1 for dimensionality reduction of drug information data in the drug development process is analyzed, and the product of the sum and the default value of perplexity in the t-SNE algorithm is used as the perplexity adjustment value; wherein, the expression of the adjustment coefficient K is, , is an exponential function with a natural constant as base, for the local importance in the drug development process.

[0023] Preferably, the dimensionality reduction processing of the drug information data further includes: using the perplexity adjustment value as the perplexity parameter of the t-SNE algorithm to perform dimensionality reduction processing on the drug information data of drug ingredients contained in all synthetic drugs in the recent drug development process.

[0024] In a second aspect, the embodiment of the present application further provides a drug information storage system for the drug development process, the system comprising:

[0025] The data acquisition module is used to obtain drug information of drug ingredients contained in each synthetic drug during the drug development process to form an information vector of each synthetic drug;

[0026] The data local importance analysis module determines the component separation degree of each element in the information vector of each synthetic drug according to the difference between each element in the information vector of each synthetic drug and the core drug component of each synthetic drug;

[0027] According to the distribution disorder degree of the component separation degree of all elements in the information vector of each synthetic drug, combined with the types of drug ingredients contained in each synthetic drug, the information importance of each synthetic drug is obtained, and the local importance of the drug development process is determined by the information importance of all synthetic drugs in the drug development process;

[0028] The data global importance analysis module determines the global importance of the drug development process based on the difference between the information importance of each synthetic drug and other synthetic drugs, combined with the average level of similarity between the information vectors of each synthetic drug and other synthetic drugs;

[0029] The data dimension reduction storage module obtains the perplexity adjustment value according to the local importance and the global importance and the default perplexity value in the t-SNE algorithm, performs dimension reduction processing on the drug information data, and stores the drug information data after dimension reduction.

[0030] In a third aspect, an embodiment of the present application also provides a drug information storage medium for the drug development process, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0031] From the above, it can be seen that the drug information storage method, system and medium provided by the present application for the drug development process have at least the following beneficial effects:

[0032] The present application takes into account that the perplexity parameter in the traditional t-SNE algorithm is difficult to determine, and that a perplexity parameter that is too large or too small will affect the effect of dimensionality reduction on drug information, resulting in poor effect of eliminating redundant information in drug information, and unable to effectively improve the efficiency of drug information storage. The present application analyzes the local and global characteristics of the drug information matrix in the drug development process, and uses the local feature complexity and global feature complexity of the drug information matrix to improve the perplexity parameter in the t-SNE algorithm, thereby avoiding the influence of too large or too small perplexity parameters on the dimensionality reduction effect, more accurately eliminating redundant information in drug information data, improving the effect of eliminating redundant information in drug information, and thereby effectively improving the efficiency of drug information storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 A flowchart of the steps of the drug information storage method for the drug development process provided in this application;

[0035] Figure 2 A schematic diagram of obtaining each core point in the information vector of each synthetic drug provided in this application;

[0036] Figure 3 A schematic diagram of the process of obtaining the component separation degree of each element in the information vector of each synthetic drug in the drug development process provided in this application;

[0037] Figure 4A block diagram of the drug information storage system provided for the drug development process in this application. DETAILED DESCRIPTION

[0038] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following, in combination with the accompanying drawings and preferred embodiments, describes in detail the drug information storage method, system and medium for the drug development process proposed in the present application, its specific implementation method, structure, characteristics and effects. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0039] Unless otherwise specified and limited, terms such as "comprises", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such articles or devices. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the article or device including the element. In addition, the term "and\or" used herein includes any and all combinations of one or more related listed items. All technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of this application.

[0040] The following is a detailed description of the specific scheme of the drug information storage method, system and medium for the drug development process provided by the present application in conjunction with the accompanying drawings.

[0041] See also Figure 1 , which shows a flowchart of a method for storing drug information in a drug development process provided by an embodiment of the present application, including the following steps:

[0042] Step 1: Obtain drug information of drug ingredients contained in each synthetic drug during the drug development process to form an information vector of each synthetic drug.

[0043] The purpose of this embodiment is to improve the effect of eliminating redundant information in drug information by improving the method of determining the value of the perplexity parameter in the t-SNE algorithm, and to more accurately eliminate the redundant information existing in the drug information data, thereby effectively improving the efficiency of drug information storage.

[0044] In the process of drug development, synthetic drugs are usually synthesized by multiple attempts at different drug component ratios. This embodiment obtains the latest drug information of various synthetic drugs in the drug development process in the past month from the drug development platform, that is, the drug information of the last synthetic attempt of the synthetic drug, and the drug information is the pH, content and proportion data of various drug components contained in each synthetic drug;

[0045] The pH value, content and proportion data of various pharmaceutical ingredients contained in each synthetic drug are used to form an information vector of each synthetic drug. Each element in the information vector represents the three-dimensional coordinates of each pharmaceutical ingredient, that is, each element in this embodiment contains three-dimensional values, where the first dimension is the pH value of the pharmaceutical ingredient, the second dimension is the content of the pharmaceutical ingredient, and the third dimension is the proportion of the pharmaceutical ingredient.

[0046] In this embodiment, a matrix consisting of information vectors of all synthetic drugs in the drug development process within the most recent month is recorded as a drug information high-dimensional matrix in the drug development process.

[0047] The high-dimensional matrix of drug information in the drug development process is preprocessed. The present invention adopts a data cleaning method to process missing values ​​and unify the data format of the high-dimensional matrix of drug information to obtain the drug information matrix in the drug development process, wherein data cleaning is a well-known technology and the specific process will not be repeated.

[0048] Step 2: Determine the component separation degree of each element in the information vector of each synthetic drug based on the differences between each element in the information vector of each synthetic drug and the core drug ingredients of each synthetic drug.

[0049] The t-SNE algorithm is a high-dimensional visualization and dimensionality reduction technology. This dimensionality reduction algorithm can cluster drug information data with similar structures together to eliminate redundant information in drug information data. However, the dimensionality reduction effect of the t-SNE algorithm is very sensitive to the value of the perplexity parameter. The larger the perplexity parameter value, the dimensionality reduction algorithm tends to maintain the global structure of the drug information data; and the smaller the perplexity parameter, the dimensionality reduction algorithm tends to focus on the local structure of the drug information data. In other words, if the perplexity parameter value in the algorithm is too large or too small, it will affect the effect of dimensionality reduction on drug information data, resulting in the inability to accurately eliminate redundant information in drug information data. Therefore, it is necessary to improve the perplexity value method in the traditional t-SNE algorithm.

[0050] Furthermore, considering that there are certain differences between the dimensional features of different drug ingredients contained in the same drug, the greater the outlier feature of the dimensional parameters of the drug ingredients contained in a certain synthetic drug, the more difficult it is to design the synthetic drug, and to a certain extent, the more important the drug ingredient information contained in the synthetic drug is. In order to improve the perplexity parameter in the dimensionality reduction algorithm based on the local and global structural features of drug information data, it is first necessary to analyze the outlier features of the drug ingredient information in the synthetic drug.

[0051] All elements in the information vector of each synthetic drug are used as the input of the DBSCAN clustering algorithm (Density-Based Spatial Clustering of Applications with Noise). In this embodiment, the density threshold minPts in the algorithm is set to 4, and the neighborhood radius is The value is 0.3, and the DBSCAN clustering algorithm outputs all the core points in the information vector of each synthetic drug. The DBSCAN algorithm is a well-known technology, and the specific process is not repeated here. The specific schematic diagram of core point acquisition is as follows: Figure 2 As shown, Figure 2 The middle triangle position represents the core point in the information vector of the synthetic drug, and the cross position represents other data points in the information vector of the synthetic drug.

[0052] It should be noted that each element in the information vector of each synthetic drug is a three-dimensional information, which is used to characterize the three-dimensional coordinates of each drug component; all the core points in the information vector of each synthetic drug are obtained by the DBSCAN clustering algorithm, and the core points represent a certain main drug component of the synthetic drug. The sequence composed of all the core points in the information vector of each synthetic drug is recorded as the core component sequence of each synthetic drug. The elements in the core component sequence of the synthetic drug characterize the distribution of the main drug components in the synthetic drug. The greater the difference in distance between the element in the information vector and the core component sequence, to a certain extent, it means that the outlier characteristics of the drug component corresponding to the element are greater, that is, the difference between the content and proportion of the drug component corresponding to the element and the content and proportion of other synthetic components is greater.

[0053] The mean of the difference between each element in the information vector of each synthetic drug and all elements in the core component sequence of each synthetic drug is used as the component separation degree of each element in the information vector of each synthetic drug. For ease of understanding, in this embodiment, the calculation formula is: , where is the component separation degree of the jth element in the information vector of the i-th synthetic drug, is the number of elements in the core component sequence of the i-th synthetic drug, is the Euclidean distance, is the jth element in the information vector of the i-th synthetic drug, is the hth element in the core component sequence of the i-th synthetic drug, For elements With elements The Euclidean distance between .

[0054] Understandably, The larger the element, the The greater the difference between the drug information of the corresponding drug component and the drug information of other drug components, that is, the higher the outlier degree of the drug component corresponding to the element, the greater the corresponding component separation degree.

[0055] For details, please refer to Figure 3 , Figure 3 The figure shows a schematic diagram of the process of obtaining the component separation degree of each element in the information vector of each synthetic drug during the drug development process.

[0056] Step 3: According to the distribution disorder of the component separation degree of all elements in the information vector of each synthetic drug, combined with the types of drug ingredients contained in each synthetic drug, the information importance of each synthetic drug is obtained, and the local importance of the drug development process is determined by the information importance of all synthetic drugs in the drug development process.

[0057] Furthermore, considering that the more drug ingredients a synthetic drug contains and the more chaotic the outlier characteristics between different drug ingredients are, it indicates that the design difficulty of the synthetic drug is higher and the drug ingredient information contained in the synthetic drug is more important.

[0058] Based on the above analysis, the information importance of each synthetic drug is calculated:

[0059] , where is the information importance of the i-th synthetic drug, is the normalization function, is the number of types of drug ingredients contained in the i-th synthetic drug, is the information entropy of the component separation degree of all elements in the information vector of the i-th synthetic drug, wherein the calculation of information entropy is a well-known technology and the specific process is not repeated here. The information importance reflects the complexity of the composition of the drug components in the synthetic drug, that is, the greater the information importance, the higher the complexity of the drug component composition of the synthetic drug, and the more important the drug component information contained in the synthetic drug.

[0060] Considering that the higher the average level of information importance of all synthetic drugs in the drug information matrix during drug development, it indicates to a certain extent that the local structural characteristics of different synthetic drugs in the drug information matrix during drug development are more complex, and at this time, the dimensionality reduction algorithm should be more inclined to focus on the local structure of drug information data. Therefore, combined with the average level of information importance of all synthetic drugs during drug development, the local importance during drug development is obtained. In this embodiment, the local importance calculation formula is: , where for the local importance in the drug development process, It is the number of types of synthetic drugs contained in the drug information matrix during the drug development process, that is, the number of types of synthetic drugs in the recent drug development process.

[0061] It can be understood that local importance reflects the complexity of the local structural characteristics of different synthetic drugs in the drug information matrix. That is, the greater the local importance, the more complex the local structural characteristics of different synthetic drugs. At this time, the dimensionality reduction algorithm should tend to focus on the local structure of drug information data.

[0062] Step 4: Based on the difference between the information importance of each synthetic drug and other synthetic drugs, combined with the average value of the similarity between the information vector of each synthetic drug and the information vector of other synthetic drugs, determine the global importance in the drug development process.

[0063] Generally, there is a certain correlation between the pH value, content and proportion of drug ingredients of different synthetic drugs. For example, the pH value and content of drug ingredients of synthetic drugs with the same effect tend to be consistent. At this time, there is a high similarity between the pH value, content and proportion of drug ingredients of different synthetic drugs. Based on the local importance of the drug information matrix and combined with the global characteristics of the drug information matrix, the perplexity parameter in the t-SNE algorithm is improved, and the global characteristics of the drug information matrix need to be further analyzed.

[0064] Based on the above analysis, combined with the similarity between the information vectors of each synthetic drug and other synthetic drugs, the similarity of the drug ingredients of each synthetic drug is calculated. Preferably, the calculation formula in this embodiment is:

[0065] , where is the similarity of the drug ingredients of the i-th synthetic drug, is the cosine similarity function, and are the information vectors of the i-th and s-th synthetic drugs respectively, is the information vector and It should be noted that the implementer may also use other similarity measurement methods to analyze the similarity between two information vectors, and this embodiment does not impose any special limitation on this.

[0066] It can be understood that the similarity of drug ingredients reflects the similarity characteristics between the content and proportion of drug ingredients of different synthetic drugs, that is, the greater the similarity of drug ingredients, the greater the similarity characteristics between the pH, content and proportion of drug ingredients of different synthetic drugs. The more synthetic drugs with the same function are developed in the drug research and development in the past month, the less complex the global characteristics of the drug information matrix are.

[0067] If there are high similarity characteristics among the drug ingredients of all synthetic drugs, and the difference in the change of information importance between different synthetic drugs is smaller, it can be explained that the drug effects of different recently developed synthetic drugs are more correlated, and the complexity of the global structural characteristics of the drug information matrix in the drug development process is smaller.

[0068] Based on the above analysis, calculate the global importance in the drug development process:

[0069] , where The global importance of the drug development process, is the information importance of the i-1th synthetic drug, is the average of the similarity of the drug ingredients of all recent synthetic drugs. In this embodiment, recent refers to the most recent month. It refers to the average similarity of drug ingredients of all synthetic drugs in the drug development process within the last month. In order to avoid the denominator being zero, in this embodiment, the error parameter is set to 0.1. The global importance reflects the complexity of the global structural features between different synthetic drugs in the drug information matrix, that is, the greater the global importance, the higher the complexity of the global structural features between different synthetic drugs, and the dimensionality reduction algorithm should maintain the global structure of the drug information data.

[0070] Step 5: According to the local importance and global importance, combined with the default value of perplexity in the t-SNE algorithm, the perplexity adjustment value is obtained, the drug information data is reduced in dimension, and the drug information data after dimension reduction is stored.

[0071] According to the local importance and global importance of the drug information matrix in the drug development process, the perplexity parameter in the t-SNE algorithm is improved. If the global importance of the drug information matrix is ​​large, while the local importance is small, it means that the global features of the drug information matrix contain richer synthetic drug information. At this time, a larger perplexity parameter should be taken, so that the dimensionality reduction algorithm tends to maintain the global structural characteristics of the drug information data; if the global importance of the drug information matrix is ​​small, while the local importance is large, it means that the local features of the drug information matrix contain richer synthetic drug information. At this time, a smaller perplexity parameter should be taken, so that the dimensionality reduction algorithm tends to focus on the local structural characteristics of the drug information data.

[0072] Furthermore, combining the local importance and global importance in the drug development process, the adjustment coefficient for reducing the dimension of the drug information matrix in the drug development process is calculated:

[0073] , where It is the adjustment coefficient for reducing the dimension of drug information data in the process of drug development. is an exponential function with a natural constant as the base. The adjustment coefficient ranges from 0 to 1. When the global importance is greater and the local importance is smaller, the adjustment coefficient approaches 1, making the perplexity adjustment value larger; when the global importance is smaller and the local importance is greater, the adjustment coefficient approaches 0, making the perplexity adjustment value smaller.

[0074] According to the adjustment coefficient and the default value of the perplexity in the t-SNE algorithm, the perplexity adjustment value when reducing the dimension of the drug information matrix in the drug development process is determined. The specific calculation formula in this embodiment is:

[0075] , where is the perplexity adjustment value for dimensionality reduction of drug information matrix in drug development process, is the default value of the perplexity in the t-SNE algorithm. In this embodiment, the default value The value is 30. The larger the perplexity adjustment value, the more the dimensionality reduction algorithm tends to maintain the global structural characteristics of the drug information data; conversely, the smaller the perplexity adjustment value, the more the dimensionality reduction algorithm tends to focus on the local structural characteristics of the drug information data.

[0076] Finally, the drug information in the research and development process is reduced in dimension, the redundant information in the drug information data is eliminated, the perplexity adjustment value is used as the perplexity parameter of the t-SNE algorithm, and the drug information data of the drug ingredients contained in all synthetic drugs in the recent drug research and development process are reduced in dimension.

[0077] Specifically, in this embodiment, the drug information matrix in the drug development process is used as the input of the t-SNE algorithm, the perplexity adjustment value is used as the perplexity parameter in the t-SNE algorithm, and the drug information matrix after dimensionality reduction is output, wherein the specific dimensionality reduction process of the t-SNE algorithm is a well-known technology and will not be repeated in this embodiment.

[0078] A distributed data storage algorithm is used to store the reduced-dimensional drug information matrix, and the reduced-dimensional drug information matrix is ​​used as the input of a consistent hashing algorithm. The reduced-dimensional drug information matrix is ​​stored in a storage hard disk of a computer server using the consistent hashing algorithm. The storage hard disk of the computer server is a drug information storage medium, and the consistent hashing algorithm is a well-known technology, and the specific process will not be repeated here.

[0079] At this point, the drug information storage method in the drug development process is completed.

[0080] On the other hand, an embodiment of the present application provides a drug information storage system for a drug development process, the system comprising:

[0081] The data acquisition module is used to obtain drug information of drug ingredients contained in each synthetic drug during the drug development process to form an information vector of each synthetic drug;

[0082] The data local importance analysis module determines the component separation degree of each element in the information vector of each synthetic drug according to the difference between each element in the information vector of each synthetic drug and the core drug component of each synthetic drug;

[0083] According to the distribution disorder degree of the component separation degree of all elements in the information vector of each synthetic drug, combined with the types of drug ingredients contained in each synthetic drug, the information importance of each synthetic drug is obtained, and the local importance of the drug development process is determined by the information importance of all synthetic drugs in the drug development process;

[0084] The data global importance analysis module determines the global importance of the drug development process based on the difference between the information importance of each synthetic drug and other synthetic drugs, combined with the average level of similarity between the information vectors of each synthetic drug and other synthetic drugs;

[0085] The data dimension reduction storage module obtains the perplexity adjustment value according to the local importance and the global importance and the default perplexity value in the t-SNE algorithm, performs dimension reduction processing on the drug information data, and stores the drug information data after dimension reduction.

[0086] Specifically, the block diagram of the drug information storage system in the drug development process in this embodiment is as follows: Figure 4 shown.

[0087] Based on the same inventive concept as the above method, an embodiment of the present application also provides a drug information storage medium for the drug development process, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above-mentioned drug information storage methods for the drug development process are implemented.

[0088] It is to be understood that the sequence of the embodiments of the present application described above is for description only and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of the present specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0089] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0090] The above content is only an implementation method of the present application and is not intended to limit the scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly used in other related technical fields, are also included in the protection scope of the present application.

Claims

1. A drug information storage method for drug development, characterized in that: The following steps are involved: Obtain drug information of drug components contained in each synthetic drug during drug development to form information vectors of each synthetic drug. Drug information includes pH value, content and proportion of drug components. Determine the component separation degree of each element in the information vector of each synthetic drug according to the difference between each element in the information vector of each synthetic drug and the core drug component of each synthetic drug; According to the distribution disorder of the component separation degree of all elements in the information vector of each synthetic drug, combined with the types of drug ingredients contained in each synthetic drug, the information importance of each synthetic drug is obtained, and the local importance of the drug development process is determined through the information importance of all synthetic drugs in the drug development process; Based on the difference between the information importance of each synthetic drug and other synthetic drugs, combined with the average similarity between the information vector of each synthetic drug and the information vector of other synthetic drugs, the global importance of the drug development process is determined; According to the local importance and the global importance, combined with the default value of the perplexity in the t-SNE algorithm, the perplexity adjustment value is obtained, the drug information data is reduced in dimension, and the drug information data after the dimension reduction is stored; The process of determining the component separation degree of each element in the information vector of each synthetic drug is as follows: Clustering all elements in the information vector of each synthetic drug respectively to obtain each core point in the information vector of each synthetic drug, and all core points in the information vector of each synthetic drug constitute the core component sequence of each synthetic drug; The mean of the differences between each element in the information vector of each synthetic drug and all elements in the core component sequence of each synthetic drug is used as the component separation degree of each element in the information vector of each synthetic drug; The calculation formula for the information importance of each synthetic drug is: , where is the information importance of the i-th synthetic drug, is the normalization function, is the number of types of drug ingredients contained in the i-th synthetic drug, is the information entropy of the component separation degree of all elements in the information vector of the i-th synthetic drug.

2. The drug information storage method for drug development process according to claim 1, characterized in that: The local importance in the drug development process is the average value of the information importance of all synthetic drugs in the drug development process.

3. The drug information storage method for drug development process according to claim 1, characterized in that: The global importance calculation formula in the drug development process is: , where The global importance of the drug development process, is the information importance of the i-1th synthetic drug, is the mean similarity of drug ingredients of all recent synthetic drugs, To avoid values ​​with a denominator of zero.

4. The drug information storage method for drug development process according to claim 3, characterized in that: The drug composition similarity of each recently synthesized drug is the average of the similarities between the information vectors of each recently synthesized drug and other synthesized drugs.

5. The drug information storage method for drug development process according to claim 3, characterized in that: The method of obtaining a perplexity adjustment value based on the local importance and the global importance in combination with the perplexity default value in the t-SNE algorithm further includes: The sum of the adjustment coefficient and the number 1 for dimensionality reduction of drug information data in the drug development process is analyzed, and the product of the sum and the default value of perplexity in the t-SNE algorithm is used as the perplexity adjustment value; wherein, the expression of the adjustment coefficient K is, , is an exponential function with a natural constant as base, for the local importance in the drug development process.

6. The method for storing drug information in the drug development process according to claim 1, characterized in that: The dimensionality reduction processing of the drug information data further includes: using the perplexity adjustment value as the perplexity parameter of the t-SNE algorithm to perform dimensionality reduction processing on the drug information data of drug ingredients contained in all synthetic drugs in the recent drug development process.

7. A drug information storage system for drug development, implementing the method as claimed in claim 1, characterized in that: The system comprises: The data acquisition module is used to obtain drug information of drug ingredients contained in each synthetic drug during the drug development process to form an information vector of each synthetic drug; The data local importance analysis module determines the component separation degree of each element in the information vector of each synthetic drug according to the difference between each element in the information vector of each synthetic drug and the core drug component of each synthetic drug; According to the distribution disorder degree of the component separation degree of all elements in the information vector of each synthetic drug, combined with the types of drug ingredients contained in each synthetic drug, the information importance of each synthetic drug is obtained, and the local importance of the drug development process is determined by the information importance of all synthetic drugs in the drug development process; The data global importance analysis module determines the global importance of the drug development process based on the difference between the information importance of each synthetic drug and other synthetic drugs, combined with the average level of similarity between the information vectors of each synthetic drug and other synthetic drugs; The data dimension reduction storage module obtains the perplexity adjustment value according to the local importance and the global importance and the default perplexity value in the t-SNE algorithm, performs dimension reduction processing on the drug information data, and stores the drug information data after dimension reduction.

8. A drug information storage medium for drug development, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.