The version 1.0 is the latest assembly, and is mirrored by Ensembl and UCSC genome browser. The other version are provided mainly for archival purposes, and should not be used unless you have special reasons (e.g. you already have your own annotations for an older assembly).
As one of the important targets of the group grant project Genome Science (Grant-in-Aid for Scientific Research on Priority Areas supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan), we started the sequencing of the medaka genome at the Academia Sequencing Center of the National Institute of Genetics (NIG) in mid 2002. The strain we chose was a southern inbred strain, Hd-rR, and sequencing was conducted by the whole-genome shotgun strategy.
The genome of Hd-rR, an inbred medaka strain, was assembled from 13.8 million reads that were obtained from the whole genome shotgun plasmid, fosmid, and bacterial artificial chromosome (BAC) libraries. The total size of the assembled contigs was 700.4 megabases (Mb). Of the 700.4 Mb in the sequenced genome, 50% of nucleotides are covered in scaffolds (or contigs) of length 1.41Mb (9.8 kilobases) that are called N50 values. This contiguity is sufficient to characterize the genomic structures of genes.
The medaka genome sequence data have been released to the public four times to meet urgent requests from the medaka research community. Four versions named 200406, 200506, version 0.9, and version 1.0 have been created to provide users with timely information. The former two versions had shorter scaffolds that were not anchored on the medaka chromosomes because they were built in 2004 and 2005, before genetics markers were available. Versions 0.9 and 1.0 were created in 2006, when comprehensive genetic markers were available, so that about 90% of their scaffolds and ultracontigs were located on the twenty-four medaka chromosomes. Versions 0.9 and 1.0 were built from the identical contigs and scaffolds, but the assembly of version 1.0 is longer than that of version 0.9 because more genetic markers could be used to generate version 1.0. Version 0.9 is left open to the public because most of the data analysis in the medaka genome paper was based on version 0.9. In these two versions, two scaffolds linked by a single BAC are connected into one ultracontig if it is consistent with genetic markers. The N50 value of ultracontigs in version 1.0 amounted to 5.1 Mb, excluding gaps, and therefore, the great continuity of ultracontigs promises to accelerate the task of positional cloning with an ample number of confirmed genetic markers in our database.
In addition to the genome of the Hd-rR medaka inbred strain, the genome of another inbred strain HNI was also sequenced to produce the draft 648-Mb HNI genome. Inbred strains Hd-rR and HNI originated in the southern and northern Japanese populations, respectively. They can mate and produce healthy offspring, although they are estimated to have diverged about 4 million years ago, and their genome sequences have diverged by approximately 3.42%. The alignment of the two medaka genomes identified about 16.4 million single nucleotide polymorphisms (SNPs), from which 2,401 SNPs were selected and mapped genetically onto medaka chromosomes using a backcross panel between these two inbred strains. These genetic SNP markers, together with 140 single sequence length polymorphism (SSLP) and restriction fragment length polymorphism (RFLP) markers, were used to anchor scaffolds on chromosomes, to construct the medaka chromosome map. These confirmed markers should be useful in isolating responsible genes of interest by positional cloning. To enable users to use the markers, our database contains PCR primers for 2,473 markers, with the genetic distances between the markers, and their locations on the chromosomes.
This work has been supported by Grant-in-Aid for Scientific Research on Priority Areas (Grant#12209003) to Shinichi Morishita.
Ramen Assembler Development Team members are indebted to Yuji Kohara and Tadasu Shin-i for their technical discussions on the whole genome shotgun assembly.
Members in the UT Genome Browser Development Team are grateful to Kiyoshi Naruse, Daisuke Kobayashi, and Takanori Narita for their valuable input to improve the functions of the browser in a variety of ways.