♦ G. classic

♦ G. taxon

• G. yeast (analyzing)

♦ G. universal (preparing for analysis)

♦ G. transfer (preparing for analysis)

• Vertical transfer
• Horizontal transfer

♦ Concept

• By the advance of next-generation sequencing, more & more organisms' genomes have been decoded.
• Although the number of these organisms is still limited inclusively, it sufficiently makes performable a phylogenetic approach in a particular kingdom.
  
• The Gcorn database project aims to reveal relationships between speciation & gene divergence based on gene and/or amino acid sequences.
• The Gcorn classic databases provide information on the dynamics of homology groups along with speciation using various thresholds of homology indices.
• The project started with kingdoms in eukaryotes; i.e., plants, fungi, protozoa, invertebrates, & vertebrates..
• After publishing databases of eukaryotes, those of prokaryotes will be planned.
• Gcorn can be applied not only to kingdom level, but to particular genes used for DNA barcoding, & to candidate genes transferred vertically (along with speciation) or horizontally (across speciation).

♦ Database type

• Gcorn classic: Databases of particular kingdoms such as plants & fungi at the genome level.
• Gcorn taxon: Databases of particular taxon such as yeast, including multiple strains or ecotypes.
• Gcorn universal: Databases of genes used for DNA barcoding such as RbcL, 16S rDNA, & COI.
• Gcorn transfer: Databases of genes transferred vertically & horizontally.

• Amino acid or base sequences & information on gene/protein function are obtained from the RefSeq database.
• In the RefSeq database, these sequences are classified into six organism groups (approximately kingdom); i.e., fungi, invertebrate, plant, protozoa, vertebrate_mammalian, vertebrate_others.
• In this approach, each organism group is used for a single database.

♦ BLASTP analysis

• In each organism group, BLASTP analyses are performed between all genes of the species selected through the pre-data analysis.

♦ Calculating homology index

• Using the result of the BLASTP analysis, homology indices (HI) between all pairs of genes are calculated.
• HI is calculated by twice the number of amino acids shared with a pair of genes over the sum of amino acids of the pair; i.e., it is equivalent to a F-measure, harmonic mean of precision & recall, of the pair.
• The index ranges between zero (dissimilar) to one (equivalent).
• For example, when the numbers of amino acids of genes A & B are 100 & 140, respectively & 90 amino acids are shared with these genes, HI is calculated as 90 x 2 / ( 100 + 140 ) = 0.75.
• HI represents a variation rate in a pair of genes.

♦ Grouping homologous genes

• Using various HIs as thresholds, genes are grouped in a manner of correlation network analysis.
• Namely, a single gene can be contained in multiple gene groups on the basis of various HIs.
• Hypothetically, a gene group represents a gene of an ancient organism.
• In the Gcorn database, therefore, the gene group is named “ancient gene.”

♦ Constructing phylogenetic tree for gene

• A phylogenetic tree of a gene group (ancient gene) that contains a gene of interest & around 20 genes.
• In the Gcorn database, a phylogenetic tree that contains many genes is not provided because such trees are difficult to grasp & understand it.
• Instead of such trees, Gcorn provides a line chart that represents information on ancient genes (or “gene evolution”; see the next section).

♦ Depicting line chart of gene evolution

• A line chart that represents the numbers of genes (red), species (blue), & families (green) in a lineage of ancient genes (or gene evolution).
• Namely, this line chart is a digest of a phylogenetic tree (regarding these numbers).
• Vertical & horizontal axes represent these numbers & HI, respectively.
• Horizontal axis is hypothetically along with evolution, although the speed of evolution (i.e., mutation or variation rate) is not necessarily constant.
• The line chart is composed of flat segments & slopes to represent the timings of homologous events.
• A red & flat segment represents no homologous event in this evolutionary period (of cource, a blue line is also flat).
• A red slope represents a homologous event: in the cases of a blue slope & flat segment, the period represents potentially an orthologous or paralogous event, respectively.

♦ Constructing correlation network between organisms

• A correlation network between orgamisms is depicted on the basis of the ratio of homologous genes.
• In the network, nodes represent orgamisms & are connected to other nodes in the cases that show the higher ratio than a threshold determined by each organism group.
• Groups (or “network modules”) in which nodes are tightly connected to each other represent organism groups that are mutually close such as members of a family.
• Red nodes represent organisms that contain genes homologous to a gene of interest on the basis of HI threshold which is changeable (0.8 as default).

♦ Depicting taxonomy tree

• This phylogenetic tree for species is depicted on the basis of information obtained from the NCBI Taxonomy database.
• Several divergence points are based on authorized information or articles (see references for each orgaism group); e.g., the tree for plants is revised by information on APG III.
• Blue & gray nodes represent taxon with & without hyperlinks to the Taxonomy database, respectively.
• Red nodes represent organisms used in the Gcorn databases & species names in red represent those that contain genes homologous to a gene of interest based on a HI threshold.

♦ Depicting orthology tree

• This tree contains genes orthologous to a gene of interest based on a HI threshold.
• When the order of branching in the tree is similar to that in the phylogenetic tree for species, the evolution (or variation) of the gene occurred similarly to speciation.
• On the other hand, when the order is dissimilar, some evolution events may occurred in a different manner to speciation.

♦ Constructing database

• The Gcorn databases are constructed using Perl/CGI & HTML scripts.
• The databases are maintained using a server of Osaka Prefecture University & each database will be updated every one or two years.

♦ Key to analysis of speciation (Kanon)

• Based on the results of the Gcorn database construction, the Kanon database will provide information on genes that hypothetically showed frequent mutations (or variations) at a particular speciation.
• At first, we have a plan to publish Kanon plant.