中文摘要：

目的利用开放性基因芯片数据库筛选氨磷汀调控人类造血及免疫相关基因并对其进行生物信息学分析。方法以氨磷汀（Amifostine）为关键词，在互联网开放性数据库包括GEO、SAGE、GeneCard、InterPro、ProtoNet、UniProt和BLOCKS中搜索基因表达数据，对筛选出的数据库进行生物信息学分析，计算差异表达基因，再利用The Database for Annotation，Visualization and Integrated Discovery（DAVID）数据库进行造血及免疫相关基因的功能聚类分析。结果仅在GEO数据库中筛选出1个氨磷汀调控人类基因组表达数据库。分析表明，氨磷汀处理K562细胞后，人类全基因组仅2．14％（460／192000）的基因在转录水平表达有显著差异（P〈0．01）。460个差异基因中有139条为已知功能基因，聚类分析共筛出17个与造血及免疫相关的基因，分属于8大类生物学通路，其中上调基因15个，下调基因2个。结论充分利用开放性基因芯片数据库，可经济、方便和快捷地分析药物调控人类基因表达谱特点，为预测和验证药物新的药理作用提供指导。

英文摘要：

Objective To analyze the amifostine-regulated human hematopoietic and immune-related genes retrieved from open gene chip databases with bioinformatics method. Methods Differently expressed amifostine-regulated human hematopoietic and immune-related genes were retrieved from internet gene expression databases including GEO, SAGE, GeneCard, InterPro, ProtoNet, UniProt, and BLOCKS using amifosfine as the key word and analyzed with bioinformatics method, and their functional clustering was analyzed according to the Database of Annotation, Visualization and Integrated Discovery （DAVID）. Results Only 1 expressed gene was screened from the GEO database. The functional clustering analysis showed that a significant difference was found only in expression of 2.14% of the whole genome （460/192 000） at transcription level after the K562 cells were treated with amifostine. Of the 460 differential genes, 139 were the known genes. Of the 139 known genes, 17 including 15 up-regulated and 2 down-regulated genes were related with hematopoiesis and immune system and divided into 8 categories in the biological pathway. Conclusion Full use of open human gene chip databases can rapidly, conveniently and economically analyze the characteristics of drug-regulated gene expression profiling, thus providing a guidance for predicting and verifying the pharmacological actions of new drugs.