Establishment of iSNP database and its application to genetic study of type 2 diabetes mellitus.

为了了解2型糖尿病的发病机制，重要的是了解胰岛中mRNA的表达谱。在本研究中，我们已经确定了大量的非冗余cDNA，大大增加了人类内分泌胰腺中表达的基因目录。从人胰岛肿瘤cDNA文库中随机选择122，000个克隆，测定其部分序列（表达序列标签，EST），并与公共数据库中的序列进行比较。对21，267个通过cDNA插入片段的3 '端测序获得的EST进行聚类分析，产生6，157个非冗余序列，包括2，323个组和3，834个单例。分别使用3 '-和5'-EST搜索核苷酸和肽数据库，显示这些EST中的3，103和58分别代表已知的人类序列或在其他物种中鉴定的基因的人类同源物和结构相关家族的新成员。序列还与公共EST数据库（dbEST和EPConDB）进行了比较，包括胰岛，胰岛素瘤和胎儿胰腺的EST。结果，新发现3，384个基因，包括胰岛特有的587个基因，在人类胰岛中表达。根据编码的推定蛋白质功能对这些序列进行分类，并通过基因组数据库分析将其分配到相应的染色体。此外，在胰岛特有的每个基因中鉴定了多个内含子或基因内SNP（iSNP）。大量的人类胰岛相关EST和iSNP遗传标记的收集将为内分泌胰腺组织特异性功能的分子研究以及糖尿病的遗传学研究提供更好的基因组来源。

In order to understand the pathogenesis of type 2 diabetes, it is important to understand the expression profile of the mRNAs in pancreatic islets. In the present study, we have identified a large number of non-redundant cDNAs, considerably increasing the catalog of genes expressed in human endocrine pancreas. The partial sequences (expressed sequence tags, ESTs) of 〜22,000 clones randomly selected from a cDNA library of human pancreatic islet tumors were determined and compared with those deposited in the public databases. Clustering analysis of 21,267 ESTs obtained by 3'-end sequencing of cDNA inserts generated 6,157 non-redundant sequences comprising 2,323 groups and 3,834 singletons. Nucleotide and peptide database searches using the 3'- and 5'-ESTs, respectively, show that 3,103 and 58 of these ESTs represent known human sequences or human homologs of genes identified in other species and new members of structurally related families, respectively. The sequences also were compared with the public EST databases (dbEST and EPConDB) including ESTs from pancreatic islet, insulinoma, and fetal pancreas. As the result, 3,384 genes, including 587 unique to the islets, are newly found to be expressed in human pancreatic islets. These sequences were classified on the basis of the putative protein functions encoded, and were assigned to the respective chromosome by genome database analysis. In addition, multiple intron or intragenic SNPs (iSNPs) were identified in each of the genes unique to pancreatic islets. The larger collection of human pancreatic islet-related ESTs and iSNP genetic markers should provide a better genome source for molecular studies of tissue-specific functions of endocrine pancreas as well as for genetic studies of diabetes mellitus.