Tutorials on configuring `cazy_webscraper` to retrieve GTDB taxonomic classifications

cazy_webscraper can be configured to retrieve the latest taxonomic classifications from the Genome Taxonomy Database (GTDB) for user specified sets of CAZymes in a local CAZyme database. Many of the same configuration options apply to the retrieval of protein data from CAZy, UniProt, GenBank and PDB.

The retrieved taxonomic classifications are stored in the local CAZyme database GtdbTaxs table. The child prteins for each taxonomy record in the GtdbTaxs table is identified by the including a ncbi_tax_id from the GtdbTaxs table in the respecitve Genomes table records.

Note

As in the GTDB database, GTDB taxonomic classifications are retrieved and associated with genomes stored in the local CAZyme database. To retrieve GTDB taxonomic classifications the genomic data for the proteins of interest must be listed in the local CAZyme database.

cazy_webscraper can be configured via the command line and/or via a YAML configuration file.

This page runs through examples of how to combine the various ‘filters’ that can be applied, to fully customised the retrieval of taxonomic classifications from GTDB. These tutorials are designed for those with less experience using command-line tools.

Note

If you installed cazy_webscraper using bioconda or pip to invoke cazy_webscraper to retrieve UniProt data call it using cw_get_gtdb_taxs - this is the method used in this tutorial. If you installed cazy_webscraper from source then you will need to invoke cazy_webscraper from the root of the repo using the command python3 cazy_webscraper/expand/genbank/taxonomy/get_ncbi_taxs.py.

From this point on, we will be discussed the cw_get_gtdb_taxs, which is the entry point for retrieving data from GTDB Taxonomy. We also presume you are comfortable configuring cazy_webscraper for the scraping of data from CAZy.

Configuration via the command line

cw_get_gtdb_taxs has two required arguments: * The path to the local CAZyme database created using cazy_webscraper * Source kingdoms. Accepts ‘archaea’ and/or ‘bacteria’

cw_get_gtdb_taxs cazy/cazyme_db.db archaea bacteria

When no optional arguments are provided, the default behaviour is invoked. The default behaviour is to: Retrieve the latest taxonomic classification from GTDB for all proteins in the local CAZyme database which do not currently have GTDB taxonomy data listed in local database, and the taxonomic information (i.e. the higher lineage classifications in the local database) are not updated.

Selecting source kingdoms

GTDB catalogues the taxonomic lineages of archaea and bacterial species. Defining the source taxonomic kingdom(s) (the second positional argument for cw_get_gtdb_taxs) determines which datafiles are retrieved from GTDB, and thus which taxonomic lineages are added to the local CAZyme database. This is separate to the --kingdoms filter which is used to define CAZymes of interest by the taxonomic classification retrieved from CAZy.

To add only archaeal lineages retrieved from GTDB use only archaea:

cw_get_gtdb_taxs cazy/cazyme_db.db archaea

To add only bacterial lineages to the local CAZyme database, using only bacteria:

cw_get_gtdb_taxs cazy/cazyme_db.db bacteria

To add both archaeal and bacterial lineages from GTDB to the local CAZyme database using both archaea and bacteria (in any order), separated with a singel space:

cw_get_gtdb_taxs cazy/cazyme_db.db bacteria archaea

Options configurable at the command line

CAZymes of interest can be defined via providing:

A set of GenBank accessions
A set of UniProt accessions
CAZy classes
CAZy families
Taxonomic kingdoms
Genera
Species
Strains
EC numbers (if previously retrieved from UniProt)

Here you can find a full list of the command-line flags and options.

Retrieving taxonomy classifications for specific CAZy classes and families

The --classes and --families flags from scraping data from CAZy are applied in the extact same way for retrieving taxonomy data from GTDB.

For instance, if instead of retrieving protein data for all CAZymes in your local CAZyme database, you want to retrieve protein data for CAZymes in specific CAZy classes then add the --classes flag followed by the classes you want to retrieve protein data for.

Tip

To list multiple classes, separate the classes with a single comma.

For example, if you want to retrieve protein data for all CAZymes from Glycoside Hydrolase and Carbohydrate Esterases then use the command:

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --classes GH,CE

OR

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --classes Glycoside Hydrolases,Carbohydrate Esterases

Retrieving protein data for proteins from specific specific CAZy families is achieved using the --families flag. For example, to retrieve protein data for all proteins in PL1, PL2 and PL3 in the local CAZyme database, use the following command:

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --families PL1,PL2,PL3

Warning

cw_get_gtdb_taxs only accpets families written in the proper CAZy family syntax. GH1 is accepted. gh1 and GlycosideHydrolases1 are not accepted.

As with scraping data from CAZy, the --classes and --families flags can be combined. To retrieve protein data for all CAZymes in PL1, PL2, PL3 and all of GH and CE both:

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --families PL1,PL2,PL3 --classes GH,CE

AND

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --classes GH,CE --families PL1,PL2,PL3

are accepted.

Applying taxonomic

The --kingdoms, --genera, --species and --strains flags can be used to refine the dataset of proteins to retrieve protein data by taxonomy. These flags are applied in the exact same way as they are used for the scraping of data from CAZy. Only proteins in the local CAZyme database and matching at least on of the provided taxonomy criteria will have data retrieved from GTDB taxonomy.

For example, if you want to retrieve data for all CAZymes in a local CAZyme database from bacterial and eukaryotic species, then use the command

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --kingdoms bacteria,eukaryota

Warning

The kingdoms must be spelt the same way CAZy spells them, for example use ‘eukaryot**a**’ instead of ‘eukaryot**e**’.

Note

The kingdoms are not case sensitive, therefore, both bacteria and Bacteria are accepted.

Note

You can list the kingdoms in any order. Thus, both bacteria,eukaryota and eukaryota,bacteria are accepted.

You can combine any combination of the optional flags, including combining the taxonomic filters. For example, you may wish to retrieve taxonomic data for all CAZymes in a local CAZyme database that are derived from all viral species, Aspergillus species, Layia carnosa, Layia chrysanthemoides, Trichoderma reesei QM6a and Trichoderma reesei QM9414. To do this we would combine the respective flags for a single cw_get_gtdb_taxs command. The command we would use would be:

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --kingdoms viruses --genera Aspergillus --species Layia carnosa,Layia chrysanthemoides --strains Trichoderma reesei QM6a,Trichoderma reesei QM9414

Note

The order that the flags are used and the order taxa are listed does not matter, and separate multiple taxa names with a single comma with no spaces.

Warning

Use the standard scientific name formating. Captialise the first letter of genus and write a lower case letter for the first letter of the species.

Aspergillus niger is correct

asepergillus niger is incorrect

ASPERGILLUS NIGER is incorrect

Warning

When you specify a species cw_get_gtdb_taxs will retrieve taxonomic data from all strains of the species.

Applying EC number filter

The retrieval of taxonomic data from GTDB can also be limited to proteins in a local CAZyme database that are annotated with specific EC numbers.

Having previously retrieved EC number annotations from UniProt and adding them to the local CAZyme database, you may wish to retrieve protein data for CAZymes annotated with specific EC numbers. To do this add the --ec_filter flag to the command, follwed by a list of EC numbers.

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --ec_filter "EC1.2.3.4,EC2.3.4.5"

Note

Provide complete EC numbers. Both dashes (‘-’) and asterixes (‘*’) are accepted for missing digits in EC numbers.

EC1.2.3.- and EC1.2.3.* are accepted. EC1.2.3. and EC 1.2.3 are not accepted.

Note

The ‘EC’ prefix is not necessary. EC1.2.3.4 and 1.2.3.4 are accepted.

Warning

If using dashes to represent missing digits in EC numbers, it is recommended to bookend the entire EC number list in single or double quotation marks. Some terminals may misinterpret EC1.2.-.- as trying to invoke the options ‘.’

Note

cw_get_gtdb_taxs will retrieve the GTDB taxonomic classification for all proteins in the local CAZyme database that are annotated with at least one of the given EC numbers. Therefore, if multiple EC numbers are given this does not mean taxonomic data will only be retrieved for CAZymes annotated for all provided EC numbers.

--ec_filter is based upon EC number annotations stored within the local CAZyme database. For example, if protein A is annotated with the EC1.2.3.4, but this annotation is not stored in the local CAZyme database, using --ec_filter EC1.2.3.4 will not cause cw_get_gtdb_taxs to retrieve data for protein A. This is because cw_get_gtdb_taxs does not know protein A is annotated with EC1.2.3.4, because this annotation is not within its database.

Warning

If --ec_filter is used along side --ec, cw_get_gtdb_taxs will retrieve all EC number annotations from UniProt for all proteins in the local CAZyme database that are associated with at least one of the EC numbers provided via --ec_filter within the CAZyme database.

Combining all filters

The --classes, --families, --ec_filter, --kingdoms, --genera, --species and --strains flags can be used in any combination to define a specific subset of proteins in the local CAZyme database for whom taxonomic data will be retrieved from GTDB.

Below we run through 3 example commands of combining these flags, and the resulting behaviour.

Example 1: To add taxonomic data for all CAZymes in GH, GT, CE1, CE5 and CE8, and which are derived from baceterial species, we use the command:

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --classes GH,CE --families CE1,CE5,CE8 --kingdoms bacteria

Example 2: To add taxonomic data for all CAZymes in GH and which are derived from Aspegillus and Trichoderma species, we use the command:

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --classes GH --genera Aspegillus,Trichoderma

Example 3: To add taxonomic classifications for all CAZymes in GH,CE and CBM which are derived from baceterial species and are annotated with at least one of EC3.2.1.23, EC3.2.1.37 and EC3.2.1.85, we use the command:

cw_get_gtdb_taxs cazy/cazyme.db archaea bacteria --classes GH,CE,CBM --kingdoms bacteria --ec_filter "3.2.1.23,3.2.1.37,3.2.1.85"

Example 4: To add bacterial taxonomic classifications for all CAZymes in GH,CE and CBM which are derived from baceterial species and are annotated with at least one of EC3.2.1.23, EC3.2.1.37 and EC3.2.1.85, we use the command:

cw_get_gtdb_taxs cazy/cazyme.db bacteria --classes GH,CE,CBM --kingdoms bacteria --ec_filter "3.2.1.23,3.2.1.37,3.2.1.85"

Providing a list of accessions

Instead of retrieving taxonomic data for all CAZymes matching a defined set of criteria, cw_get_gtdb_taxs can retrieve taxonomic data for a set of CAZymes defined by their GenBank and/or UniProt accession.

The flag --genbank_accessions can be used to provide cw_get_gtdb_taxs a list of GenBank accessions to identify the specific set of CAZymes to retrieve taxonomic data for.

The flag --uniprot_accessions can be used to provide cw_get_gtdb_taxs a list of UniProt accessions to identify the specific set of CAZymes to retrieve taxonomic data for.

In both instances (for --genbank_accessions and --uniprot_accessions) the list of respective accessions are provided via a plain text file, with a unique protein accession of each line. The path to this file is then passed to cw_get_gtdb_taxs via the respective --genbank_accessions and --uniprot_accessions flag.

--genbank_accessions and --uniprot_accessions can be used at the same time to define all CAZymes of interest.

Warning

--genbank_accessions and --uniprot_accessions take president over the filter flags.

When either --genbank_accessions or --uniprot_accessions is used, cw_get_gtdb_taxs will not retrieve any CAZymes from the local database matching a set of criteria.

Therefore, if --genbank_accessions and --classes are used, cw_get_gtdb_taxs will ignore the --classes flag and only taxonomic classifications for the proteins listed in the file provided via the --genbank_accessions.

Providing GTDB datafiles

By default --cazy_webscraper retrieves the latest GTDB datafiles from the GTDB website. However, you can provide your own GTDB datafiles.

Specifically, these are the avaialble from GTDB release page.

The filenames must use the same filename format as GTDB to enable the correct extraction of the GTDB release number from the filename.

To provide a previously downloaded archaea datafile use the --archaea_file flag followed by the path point to the target data file.

Similarly, to provide a previously downloaded bacteria datafile use the --bacteria_file flag followed by the path point to the target data file.

Note

cazy_webscraper excepts GTDB datafiles in the compressed (.gz) tab-separated file format (.tsv).

Updating genomic classifications

By default cw_get_gtdb_taxs only adds links to GTDB lineages to genomes in the local CAZyme database that are not already linked to a GTDB lineage. To update which GTDB lineage genomes in the local CAZyme database are linked to add the --update_genome_lineage flag.

cw_get_gtdb_taxs cazy/cazyme.db bacteria \
   --classes GH,CE,CBM \
   --kingdoms bacteria \
   --update_genome_lineage

Tutorials on configuring cazy_webscraper to retrieve GTDB taxonomic classifications

Configuration via the command line

Selecting source kingdoms

Options configurable at the command line

Retrieving taxonomy classifications for specific CAZy classes and families

Applying taxonomic

Applying EC number filter

Combining all filters

Providing a list of accessions

Providing GTDB datafiles

Updating genomic classifications

Tutorials on configuring `cazy_webscraper` to retrieve GTDB taxonomic classifications