Traces of ancient duplications of extensive chromosomal regions are being discovered within the human genome. For example, the MHC (major histocompatibility complex) gene region on chromosome 6 (6p21.3) encodes a cluster of genes that are homologous to those on chromosome 9 (9q33-q34) . These regions appear to have arisen from regional duplication around the time of vertebrate emergence. To elucidate the evolution of mammalian genomes, it is crucial to search for duplicated chromosomal regions in the mouse genome. In this study, we searched for long duplicated chromosomal regions in the mouse genome using the map information derived from the Mouse Genome Database and the numerous homologous gene pairs from GenBank. To identify the regions, we searched for homologous gene pairs among the genomic sequences extracted from GenBank (r.118) . We defined the candidates of duplicated regions as those having more than two homologous gene pairs located within 5cM on each chromosome. We conducted a statistical test considering the distribution of homologous gene pairs and tandem repeats in the mouse genome. Twenty-seven pairs of duplicated regions were found in this study. Seven procollagen genes, three Hox gene clusters, six integrin genes, three fibroblast growth factor receptor genes and three eye absent homolog genes were located in the duplicated regions.
A reproducibility of experimental data is important for application of differential display method to the comparison of mRNA expression level among many samples. First, we developed the optimum method to evaluate size (BP) and yield (IOD) of PCR products by analyzing an electrophoretic image. BP (A), which is a way to calculate size of each product by using 3 molecular markers loaded in extreme and central lanes of gel as references, was optimal for size evaluation. Further, IOD (A) and IOD (P), respectively, which are way to use a measured absolute value of signal intensity of each product, and to use a relative value of signal intensity of each product as proportion of total sum of all products in an identical lane, as yield, were determined as the suboptimum and optimum evaluation methods. Next, we performed an analysis of variance of intergel, interlane, and interreaction effects, when using the optimum method. We detected statistically significant effects for intergel and interlane, but not for interreaction, when using BP (A) method. As for IOD (A) and IOD (P), significant effects for intergel and interreaction were observed, but not for interlane. Finally, we evaluated the replicate pattern that is suitable for a largescale gene expression profiling, by examining coefficient of variation of combined effects for BP and IOD obtained by 4 different replicate patterns.