Transcriptional regulation of protein complexes and biological pathways
Transcriptional regulation of protein complexes and biological pathways
复制标题
DOI：
10.1007/s00335-002-2260-x
复制
发表时间：
2003-09-01
期刊：
影响因子：
2.5
通讯作者：
Levy, S
中科院分区：
文献类型：
--
作者：
Hannenhalli, S;Levy, S
基因的顺式元件谱（cis-element profile）是指调控基因转录的转录因子结合位点（transcription factor binding sites， TFBS）的集合。各种已发表的研究试图通过模式检测算法发现共表达基因附近的顺式元件，其基础是一个隐含的假设，即共表达基因与其顺式谱之间存在相关性。在这项研究中，我们发现顺式相似性（定义为两个顺式元件谱之间共享TFBS的比例）对于功能连接的相互作用蛋白以及信号转导途径的成员来说更高。在一系列代谢途径中催化相邻底物转化为产物的酶的类似分析没有显示出更高的顺式相似性。该分析基于三个不同的公开数据来源，即：1)相互作用蛋白的BIND数据库，2)NMDAR蛋白复合物中已知的相互作用，3)凋亡途径和9个与辅助因子和维生素代谢相关的途径，均来自KEGG。此外，我们分析了NMDAR复合体谷氨酸受体（GR）亚复合体中所有基因的顺式元件谱，以检测与大多数基因相邻的一组顺式元件。通过与已发表的生物医学文献进行比较，我们发现大多数相应的转录因子都参与了GR调控。此外，我们能够通过搜索与GR检测到的转录本共享相同的调控信号来检测基因产物与GR相关的转录本。这为构建高阶基因调控模型和检测共调控基因产物提供了一种新的计算方法。
The cis-element profile (or cis-profile) of a gene refers to the collection of transcription factor binding sites (TFBS) regulating the transcription of the gene. Underlying the various published studies that attempt to discover cis-elements in the vicinity of co-expressed genes via pattern detection algorithms, there is an implicit assumption that a correlation exists between co-expressed genes and their cis-profiles. In this study, we show that the cis-similarity, defined as the proportion of shared TFBS between two cis-element profiles, is higher for functionally linked interacting proteins as well as for members of a signal transduction pathway. A similar analysis of the enzymes catalyzing the conversion of adjacent substrates to products in a collection of metabolic pathways, did not reveal higher cis-similarity. The analysis is based on three distinct sources of publicly available data, namely, 1) the BIND database of interacting proteins, 2) known interactions in NMDAR protein complex, 3) the apoptosis pathway and nine pathways related to metabolism of cofactors and vitamins all from KEGG. Additionally, we analyze the cis-element profiles of all the genes in the glutamate receptor (GR) sub-complex of NMDAR complex to detect a set of cis-elements that occur adjacent to a majority of the genes. We show that most of the corresponding transcription factors are known to be involved in GR regulation by comparing our findings with the published biomedical literature. In addition, we were able to detect transcripts whose gene products associate with GR by searching for transcripts that share the same regulatory signals as those detected for GR. This suggests a novel computational methodology for constructing high-order gene regulatory models and detecting coregulated gene products.