Taxallnomy is a taxonomic database based on NCBI Taxonomy that provides taxonomic lineages according to the ranks used on Linnean classification system (e.g. Kingdom, Phylum, Class etc.).

In this package you will find the following files/folder.

* generate_taxallnomy.pl- a script that generates files for Taxallnomy 
                          database;
* get_lineage.pl        - a script that accesses the Taxallnomy database 
                          to retrieve the taxonomic lineage of TaxIDs of interest;
* lib/                  - lib folder containing a PERL module required by the 
                          script 'get_lineage.pl'.

All scripts were developed with Perl and tested in a UNIX environment.

The script generate_taxallnomy.pl generates all necessary file to load the Taxallnomy database in a local MySQL. The execution of this script requires internet connection. To run the script, just type in a UNIX terminal the following command:

> perl generate_taxallnomy.pl

The script will basically download the latest NCBI Taxonomy database and construct the Taxallnomy hierarchical structure from it. After the running, it will generate these files:

* taxallnomy_XXX.sql  - Dump file containing SQL commands to create Taxallnomy 
                        database and the table XXX;
* taxallnomy_XXX.tab  - Tab-delimited file containing contents of Taxallnomy  
                        database of table XXX;

XXX conrresponds to the table name of Taxallnomy database. They are lin, lin_name, tree_balanced, tree_original, tree_all, tax_data or rank. Detailed description of each table is presented below;

To load one of those tables in your local MySQL, go to the path where these files are located and type the following command line:

> mysql -u <username> -p < taxallnomy_XXX.sql

or go to MySQL environment and type:

mysql> source taxallnomy_XXX.sql

Loading Taxallnomy database to MySQL may take several minutes depending on which table is being loaded, so be patient.

After loading a single table, your MySQL should have a database called "taxallnomy" and, in that database, a table named XXX. There is no need to load all tables. Load only those that meet your needs.

This table contains all taxonomic lineages of Taxallnomy database. From this table, you can query for the taxonomic lineage of an organism by its taxonomy ID (primary key). Each taxonomic ranks are represented in a column of this table. Keep in mind that the content of the taxonomic rank columns is not a taxon name, but a taxon code used by Taxallnomy (See the section "Taxallnomy taxon code"). The taxon names can be programmatically generated from the taxon code or by querying on table lin_name. You can also use the script "get_lineage.pl" to retrieve lineages with taxon name from this table.

• lin table content:

This table has the same structure as lin table, but instead of having taxon codes on each taxonomic rank column, it contains the taxon name. This table occupies more space than the lin table. Contents of this table could be retrieved using script get_lineage.pl (see below).

This table provides the balanced hierarchical structure of Taxallnomy database.

• tree_balanced table content:

It has the same structure as the tree_balanced table. In this tree, no rank taxa that were deleted during the generation of the taxallnomy database (because no rank could be assigned to it), are preserved. Thus, the hierarchical strucuture of this table is not balanced.

It has the same structure as the tree_balanced table and the same hierarchical structure as the original database from NCBI Taxonomy.

This table provides information about each txid comprising the NCBI Taxonomy.

* includes txid which has "unpublished", "unidentified", "unclassified", "environmental", "unassigned", "incertae sedis" or "other sequences" in its name.

A table containing some information about the taxonomic ranks comprising the database.

• rank table contents:

Taxallnomy primarily uses the Taxonomy ID provided by NCBI Taxonomy database to identify all nodes comprising its hierarchical structure. However, since Taxallnomy algorithm creates new nodes or assign ranks to existent nodes, we included in the identifier a code to identify them properly.

The Taxallnomy "code" is added to the NCBI taxonomy ID as decimal number of three digits. For example, in the taxon code 8287.071, 8287 is the NCBI Taxonomy ID (Sarcopterygii) and 071 is the code added by Taxallnomy algorithm. In the code, the first two digits indicates the taxonomic rank in which it belongs. It goes through the code "01" to "33", in which the first rank is Superkingdom ("01") and the last one is Forma ("33"). The third digit indicates the type of node and corresponds to the action performed by the algorithm on the node. This can be:

• 1 (type 1) - This taxon was originally unranked on NCBI Taxonomy tree and it was ranked by Taxallnomy algorithm.
• 2 (type 2) - This taxon was created by Taxallnomy algorithm and the name of one of its ranked descendant taxon was used to name it.
• 3 (type 3) - This taxon was created by Taxallnomy algorithm and the name of one of its ranked ascendant taxon was used to name it.

We use the following rules to name the nodes of each type:

• For type 1, we use the abbreviation of the rank name followed by the scientific name of the NCBI Taxonomy ID. E.g. For taxon code 8287.071, we name it as spc_Sarcopterygii;
• For type 2, we use the abbreviation of the rank name followed by the preposition "of" and followed by the scientific name of the NCBI Taxonomy ID. E.g. For taxon code 9605.202, we name it as tri_of_Homo;
• For type 3, we use the abbreviation of the rank name followed by the preposition "in" and followed by the scientific name of the NCBI Taxonomy ID. E.g. For taxon code 9606.273, we name it as sbs_in_Homo sapiens;

Obs.: Taxa originally ranked in NCBI Taxonomy database have the code 000 and we use their own scientific name to name them.

You can use the script 'get_lineage.pl' to programatically retrieve the taxonomic lineage of TaxIDs of interest. To use this script, you have to load only the lin, tax_data and rank tables to taxallnomy database on MySQL. To test if the script is running properly on your system, type the following command:

> perl get_lineage.pl

If all goes well, this will show a message like this:

ERROR Please provide TaxIDs or a file containing a list of TaxIDs.
Usage:
	perl get_lineage.pl -txid 9606,9595

	perl get_lineage.pl -file <txid_list_file>

	Inputs:
		[-txid set_of_taxids] [-file txids_list_file]

	Other parameters:
		[-rank rank_code] [-srank ranks] [-format format_code] [-user
		mysql_user] [-showcode] [-out file_name] [-database database_name]
		[-linTab table_name] [-taxTab table_name] [-rankTab table_name]

	Help:
		[-help] [-man]

		Use -man for a detailed help.                                          

Detailed instruction for running this script can be accessed by typing:

> perl get_lineage.pl -man

This script uses the Perl modules 'Net::Wire10' and 'Term::InKey' located in the folder 'lib/' that accompanies this package. If you want to execute this script in other location, install these modules on your system. The easiest way to install a Perl module is using CPAN. For this, login as root user and type the command:

> cpan

After entering the CPAN environment, type:

cpan> install Net::Wire10
cpan> install Term::InKey

This will install the modules Net::Wire10, Term::InKey and their dependencies.

After the module installations, leave the CPAN environment and test the script.

If you have any suggestion to improve our work or any question, feel free to contact us by these email addresses:

[email protected] (PhD. Tetsu Sakamoto)

[email protected] (PhD. J. Miguel Ortega)

See also the Taxallnomy website at bioinfo.icb.ufmg.br/taxallnomy

Laboratorio de Biodados Universidade Federal de Minas Gerais (UFMG) Belo Horizonte - MG, Brazil