Construction Of Knowledge Base In The Field Of Drug Interactions

With the aging of the global population,the simultaneous use of two or more drugs due to co-morbidities is also increasing,resulting in an increasing risk of drug interactions.Beneficial drug interactions can enhance the therapeutic effects of drugs and antagonize adverse drug reactions.However,adverse drug interactions will reduce the efficacy of drugs,prolong the course of the disease,and even produce toxicity that is life-threatening,seriously threatening the life and health of the public,and also increasing the personal and social economic burden.Therefore,drug interaction has become an important research issue that has attracted many attentions from the government,medical organizations,scientific research scholars,and the public.To reduce the occurrence of adverse reactions due to drug interactions,the collaboration of clinical doctors,the general public,and scientific researchers is needed.On the one hand,we must accurately,comprehensively and systematically master the knowledge of the known drug interaction field,and on the other hand,we must constantly Mining and discovering potential drug interaction domain knowledge,so as to make reasonable medication decisions when multiple drugs are used at the same time.At present,knowledge in the field of drug interactions has the characteristics of wide sources,huge volume,and diverse types,which makes it difficult for people to obtain knowledge in the field of drug interactions: First,the knowledge in the field of drug interactions is scattered and distributed in different carriers,and a large amount of knowledge in the field of drug interactions still exists in an unstructured form,lacking a standardized representation,and it is difficult for users to quickly obtain and apply;Second,the knowledge in the field of drug interactions of different sources,structures,dimensions,and granularities lacks a reasonable organization,and it is difficult for users to quickly jump and use collaboratively between different knowledge resources when acquiring knowledge;Third,the huge volume of drugs,the thousands of people and diverse combinations of drug treatments make it less efficient to obtain knowledge in the field of drug interactions through traditional channels such as drug experiments and doctors’ clinical experience,and it is difficult to obtain an efficient and comprehensive field of drug interactions.knowledge.In summary,users lack efficient,integrated,systematic,and comprehensive knowledge acquisition channels in terms of obtaining knowledge in the field of drug interactions.With the development of knowledge base related theories and the emergence of technologies and methods such as information extraction,semantic web,and machine learning,it is possible to solve the above problems from the perspective of medical informatics.As a knowledge collection that stores,organizes,processes knowledge and provides knowledge support for decision-making,knowledge base plays an important role in knowledge management and knowledge services,and is an effective way to solve the problem of insufficient knowledge acquisition in the field of drug interactions.Therefore,this article launched the research work on the construction of the knowledge base in the field of drug interactions.On the theoretical level,based on the analysis of the needs of target users,combined with the theory of drug interactions,the knowledge system structure of the knowledge base in the field of drug interactions is constructed.And put forward the knowledge hierarchical structure and construction process of the knowledge base in the field of drug interactions.Based on this,the strategies for key links such as knowledge acquisition,knowledge association,and knowledge reasoning in the process of building a knowledge base in the field of drug interactions are respectively proposed.And under the guidance of each strategy,the specific realization process of each key link is constructed.At the same time,an in-depth study of the knowledge acquisition and knowledge reasoning methods of the knowledge base was carried out to lay the foundation for the feasibility and operability of the construction of the knowledge base.At the practical level,the process and method were verified,and on this basis,the prototype system of the knowledge base in the field of drug interaction was designed and implemented.The specific content of this article is as follows:⑴ Research on the organizational structure of the knowledge base in the field of drug interactions.First of all,starting from the needs of target users,by analyzing the knowledge needs of the general public,clinical doctors,and scientific researchers,three types of potential users are analyzed to clarify the construction principles and knowledge categories of the knowledge base in the field of drug interactions;Secondly,under the guidance of the drug interaction theory,classify and summarize the knowledge in the drug interaction field,construct the knowledge base knowledge system structure in the drug interaction field,and guide the acquisition and organization of knowledge base resources in the drug interaction field;Third,construct the knowledge hierarchical structure of the knowledge base in the field of drug interactions,decompose the construction process of the knowledge base in the field of drug interactions into three levels: knowledge resource layer,knowledge correlation layer,and knowledge application layer,and explain the relationship between the three.Guide each link in the construction of a knowledge base in the field of drug interactions.⑵ Research on knowledge acquisition of knowledge base in the field of drug interactions.A multi-source,multi-dimensional and multi-granularity knowledge acquisition strategy in the field of drug interactions is proposed,and the knowledge acquisition process in the field of drug interactions is constructed under the guidance of the strategy.Take structured knowledge source Drug Bank drug database and unstructured knowledge source Medline literature database as knowledge sources,and use natural language processing,deep learning and other technical methods as support for empirical research.Among them,it focuses on the research of multi-granularity knowledge extraction methods on the experimental data of drug interaction metabolic kinetics in scientific literature,which meets the knowledge needs of different users and lays the foundation for the deep organization of knowledge in the knowledge base of drug interactions.⑶ Knowledge association research on knowledge base in the field of drug interactions.At the theoretical level,combined with the characteristics of knowledge in the drug interaction domain,we propose different sources,dimensions,and granularity of knowledge association strategies in the domain of drug interactions,and build a knowledge association process based on linked data in the domain of drug interactions under the guidance of the strategies.Under the guidance of this process,conduct empirical research on knowledge resources obtained from knowledge acquisition.Construct the semantic concept model of the knowledge base in the drug interaction field,and use the D2 R related data publishing tool to classify and organize the drug interaction field knowledge of different sources,dimensions and granularities to form an orderly related domain semantic knowledge network,which is convenient for users to interact with drugs.Effectively retrieve and utilize knowledge in the field of action.⑷ Knowledge inference research on knowledge base in the field of drug interactions.By analyzing the knowledge reasoning goal of the knowledge base in the drug interaction domain,the knowledge reasoning strategy of the drug interaction domain knowledge base is proposed,and the knowledge reasoning process of the drug interaction domain knowledge base is condensed.Then use this process as a guide to propose two deep learning-based graph representation learning algorithms,which use domain knowledge of different dimensions such as drug chemical structure,drug targets,drug signaling pathways,etc.The domain knowledge in the knowledge base is improved.⑸ Research on the application of knowledge in the knowledge base in the field of drug interactions.On the basis of theoretical and methodological research,a prototype knowledge base system in the field of drug interactions was designed and implemented.The architecture and functions of the prototype system are designed by combining the usage scenarios and knowledge requirements of different users,and the system is divided into three parts: the back-end task management module,the data storage module and the front-end system interface.The functional modules include five functions: drug interaction retrieval,drug interaction characterization retrieval,drug interaction mechanism retrieval,drug attribute retrieval,and drug interaction visualization.From simple description to system research,from appearance to internal mechanism,so as to meet the knowledge needs of different user groups.