Dysregulation of mi R-124 has been reported to be involved in the pathophysiology of depression. Chaihu-Shugan-San, a traditional Chinese medicine, has antidepressive activity; however, the underlying mechanisms remai...Dysregulation of mi R-124 has been reported to be involved in the pathophysiology of depression. Chaihu-Shugan-San, a traditional Chinese medicine, has antidepressive activity; however, the underlying mechanisms remain unclear. In this study, to generate a rodent model of depression, rats were subjected to a combination of solitary confinement and chronic unpredictable mild stress for 28 days. Rats were intragastrically administered Chaihu-Shugan-San（2.835 m L/kg/d） for 4 weeks, once a day. Real-time reverse-transcription quantitative polymerase chain reaction, mi RNA microarray, western blot assay and transmission electron microscopy demonstrated that ChaihuShugan-San downregulated mi R-124 expression and upregulated the m RNA and protein levels of mitogen-activated protein kinase 14（MAPK14） and glutamate receptor subunit 3（Gria3）. Chaihu-Shugan-San also promoted synapse formation in the hippocampus. The open field test, sucrose consumption test and forced swimming test were used to assess depression-like behavior. After intragastric administration of Chaihu-Shugan-San, sucrose consumption increased, while the depressive behaviors were substantially reduced. Together, these findings suggest that Chaihu-Shugan-San exerts an antidepressant-like effect by downregulating mi R-124 expression and by releasing the inhibition of the MAPK14 and Gria3 signaling pathways.展开更多
大语言模型(large language model,LLM),包括ChatGPT,在理解和响应人类指令方面表现突出,对自然语言问答影响深远。然而,由于缺少针对垂直领域的训练,LLM在垂直领域的表现并不理想。此外,由于对硬件的高要求,训练和部署LLM仍然具有一定...大语言模型(large language model,LLM),包括ChatGPT,在理解和响应人类指令方面表现突出,对自然语言问答影响深远。然而,由于缺少针对垂直领域的训练,LLM在垂直领域的表现并不理想。此外,由于对硬件的高要求,训练和部署LLM仍然具有一定困难。为了应对这些挑战,以中医药方剂领域的应用为例,收集领域相关数据并对数据进行预处理,基于LLM和知识图谱设计了一套垂直领域的问答系统。该系统具备以下能力:(1)信息过滤,过滤出垂直领域相关的问题,并输入LLM进行回答;(2)专业问答,基于LLM和自建知识库来生成更具备专业知识的回答,相比专业数据的微调方法,该技术无需重新训练即可部署垂直领域大模型;(3)抽取转化,通过强化LLM的信息抽取能力,利用生成的自然语言回答,从中抽取出结构化知识,并和专业知识图谱匹配以进行专业验证,同时可以将结构化知识转化成易读的自然语言,实现了大模型与知识图谱的深度结合。最后展示了该系统的效果,并通过专家主观评估与选择题客观评估两个实验,从主客观两个角度验证了系统的性能。展开更多
基金supported by the National Natural Science Foundation of China,No.81503415,81574038,81603671the China Postdoctoral Science Foundation Grant,No.2016M600709+1 种基金a grant from the Science and Technology Planning Project of Guangdong Province of China,No.2014A020221062a grant from the Science and Technology Planning Project of Shenzhen City of China,No.JCYJ20150401170235349,JCYJ20160428105749954
文摘Dysregulation of mi R-124 has been reported to be involved in the pathophysiology of depression. Chaihu-Shugan-San, a traditional Chinese medicine, has antidepressive activity; however, the underlying mechanisms remain unclear. In this study, to generate a rodent model of depression, rats were subjected to a combination of solitary confinement and chronic unpredictable mild stress for 28 days. Rats were intragastrically administered Chaihu-Shugan-San（2.835 m L/kg/d） for 4 weeks, once a day. Real-time reverse-transcription quantitative polymerase chain reaction, mi RNA microarray, western blot assay and transmission electron microscopy demonstrated that ChaihuShugan-San downregulated mi R-124 expression and upregulated the m RNA and protein levels of mitogen-activated protein kinase 14（MAPK14） and glutamate receptor subunit 3（Gria3）. Chaihu-Shugan-San also promoted synapse formation in the hippocampus. The open field test, sucrose consumption test and forced swimming test were used to assess depression-like behavior. After intragastric administration of Chaihu-Shugan-San, sucrose consumption increased, while the depressive behaviors were substantially reduced. Together, these findings suggest that Chaihu-Shugan-San exerts an antidepressant-like effect by downregulating mi R-124 expression and by releasing the inhibition of the MAPK14 and Gria3 signaling pathways.
文摘大语言模型(large language model,LLM),包括ChatGPT,在理解和响应人类指令方面表现突出,对自然语言问答影响深远。然而,由于缺少针对垂直领域的训练,LLM在垂直领域的表现并不理想。此外,由于对硬件的高要求,训练和部署LLM仍然具有一定困难。为了应对这些挑战,以中医药方剂领域的应用为例,收集领域相关数据并对数据进行预处理,基于LLM和知识图谱设计了一套垂直领域的问答系统。该系统具备以下能力:(1)信息过滤,过滤出垂直领域相关的问题,并输入LLM进行回答;(2)专业问答,基于LLM和自建知识库来生成更具备专业知识的回答,相比专业数据的微调方法,该技术无需重新训练即可部署垂直领域大模型;(3)抽取转化,通过强化LLM的信息抽取能力,利用生成的自然语言回答,从中抽取出结构化知识,并和专业知识图谱匹配以进行专业验证,同时可以将结构化知识转化成易读的自然语言,实现了大模型与知识图谱的深度结合。最后展示了该系统的效果,并通过专家主观评估与选择题客观评估两个实验,从主客观两个角度验证了系统的性能。
