识别药物作用表型谱（治疗作用和不良反应）的高通量计算方法及研究策略是后基因组时代药物研发的热点问题,也是药物基因组信息学领域具有挑战性课题。本项目以网络药理学为研发理念,整合药物的化学亚结构,不良反应,治疗作用和蛋白domain数据，采用相关性分析方法识别domain与亚结构及作用表型的关联关系；构建亚结构-domain-不良反应、亚结构-domain-治疗作用复合网络，并基于亚结构与作用表型与domain的关联一致性程度，进一步优化亚结构与作用表型的关联关系；进而结合亚结构与药物的已知关系，生成药物-亚结构-作用表型关联网络。最后构建药物信息学在线预测平台。实现只需要亚结构信息就可以预测上市药物或者新药候选物的作用表型谱；实现上市药物重定位、药物作用表型机制预测、新药候选物作用表型及作用表型特异亚结构预测，为新药结构优化和进一步开发提供针对性指导，也为药物研发引入新思想和开辟新途径。

It is a very meaningful and challenging pharmacogenomics task to make high-throughput computation methods for identification of drug reaction phenotype s(therapeutic effects and adverse reactions). Firstly，we will integrate the data information with substructure , adverse reactions, therapeutic effects of approved drugs and domains of approved drug targets(including therapeutic effect target proteins and drug adverse reaction related proteins), and construct multidimensional two element matrix models,respectively, then identificate the relationship between chemical substructures and domains，adverse reactions and domains ，therapeutic effects and domains by correlation analysis methods.Second, we will construct a complex network composed of substructures-domains-adverse reactions，substruction-domains-therapeutic effects.and , further identificate strong correlation of substructures- drug reaction phenotypes based on relational consistency measure,and combining with the known relationship of drug -substructure, construct drug–substructures-drug effect phenotypes network(DSPN) . So that we can achieve drug repositions, explan the drug action mechanism and find the risk drug combination and Associated with accompaning occuranc adverse reactions . At the same time, based on the tree structure relationship, we also can make predictions for new drug candidates about not only phenotype and mechanism, but also substructural features of these new drug candidates action phenotypes. Finally, we will create a query and prediction platform for drugs and new drug candidates based on the developing models and methods, and integrated data resources.