Construction of Gene Interaction Network with Computational Life

通过构建具有计算寿命的基因互作网络，获得了以下结果。针对基因相互作用网络模型，提出了一种基因功能标注方法。以日本理研基因组科学中心测序的小鼠全长cDNA序列为研究对象，利用序列相似性对其进行聚类，寻找共同保守区的候选序列，剔除InterPro等已知数据库中的基序，得到16个新的候选基序。通过与已知基因序列的同源性、染色体的位置、蛋白质二级结构和跨膜区的预测等分析，对检测到的10个候选基序进行了功能注释。通过基因表达谱聚类，揭示了一组与发育阶段、组织特异性和代谢途径相关的基因。虽然许多研究人员已经用微阵列数据对一组基因进行了功能预测，但利用FANTOM联盟的可靠注释，可以阐明给定基因的功能关系，并找到功能未知的基因。

The following results were obtained by the construction of gene interaction network with computational life. For the model of gene interaction network, a method was developed for the functional annotation of genes. As the target of our research, the mouse full-length cDNA sequences, which were sequenced by RIKEN Genome Science Center, were clustered with their sequence similarity, the candidates for commonly conserved regions were explored, and 16 new motif candidates were obtained by removing the motifs found in known databases such as Inter Pro. Ten of the detected motif candidates were functionally annotated by various analyses, such as homology to the known gene sequences, the location of chromosomes, the predictions of protein secondary structure and transmembrane regions.Databases for expression profiles of the mouse genes at various developmental phases and tissues were developedfor the mouse cDNA microarray, which were developed at RIKEN Genome Science Center. A set of genes related to developmental phases, tissue specificity and metabolic pathway were elucidated by the clustering of gene expression profiles. Although many researchers have already done the functional predictions of a set of genes with microarray data, the elucidation of functional relationships of given genes and the gene finding with unknown functions became able to do using reliable annotations by the FANTOM consortium.