markerdb 0.2.0

markerdb

DNA Marker Database for Metabarcoding

Accurate species identification relies on the marker sequence's resolution and the quality and comprehensiveness of the reference library.

This repository contains a tool for creating a DNA marker database suitable for metabarcoding studies. The database can be tailored to include various organisms, taxonomic groups, and specific marker sequences.

The tool employs sequence information from BLAST databases and NCBI's taxonomy to create the reference database. It is compatible with sequences conforming to NCBI's taxonomy standards and supports multiple markers such as COI, COII, COIII, 12S, 16S, Cytb, and rbcL. Users can add additional markers by editing the synonyms file.

How does it work?

The tool utilizes sequence titles from BLAST databases to identify marker sequences and uses a synonyms file to obtain alternative names for these markers. The TaxonKit tool is used to fetch taxonomy details.

The different steps involved are

1. Creates a table with sequence attributes from blast databases.
2. Extracts species taxids for all species under each taxa specified in a taxalist file.
3. Parses the synonyms file and the sequence titles obtained in step1 to extract marker-specific details.
4. Gets fasta sequences for the marker accessions obtained in step3.
5. Creates sqlite3 database with marker sequence details.

The entire process typically takes ~45 minutes using NCBI's nt database.

Quick start

How to install the tool?

Step1: Install script

pip install markerdb --upgrade

Step2: Install additional requirements

conda install --file conda-requirements.txt

The tool uses Taxonkit for fetching taxonomy details. See here for the additional requirements before using the Taxonkit for the first time.

How to run

The command below uses NCBI's nucleotide blast databases (nt) to extract fish specific COI marker sequences. nt databases should be in the BLASTDB variable.

markerdb create -m COI -t taxa.txt -b nt -s synonyms.csv

To test the tool, use the blast databases included here.

markerdb create -m COI -t taxa.txt -b blastdb/demo -s synonyms.csv

Inputs

1. Marker name ( eg: COI)
2. Taxa list (eg: taxa.txt)
3. NCBI's nucleotide blast databases (nt)
4. Synonyms file (eg.synonyms.csv)

1.Marker Name: Specify the marker gene for which the sequences must be obtained. The specified gene and its alternative names must be present in the synonyms file. To extract all markers listed in the synonyms file, use 'ALL'.

2.Taxa list: This is a list of taxonomic names for which marker gene sequences are extracted. The list can contain names at the species level or higher taxonomic ranks. Marker genes for all species and subspecies within each taxon are extracted. For instance, specifying 'Viridiplantae' will yield marker genes for all plants. An example taxa list file is

Coelacanthimorpha
Hyperoartia
Hyperotreti
Actinopterygii
Dipnoi
Chondrichthyes

3.Blast databases: Indicate the BLAST database from which to extract marker sequences. If the database is not in the system path, provide its absolute path, including the prefix. For example /export/refs/nt/nt.

4. Synonyms file: A comma separated file with alternative names for the marker genes. This file is provided in this repository. Users can edit the file to add additional markers.

Outputs

The main outputs are

1. A FASTA file containing the marker sequences (marker.fa)
2. SQLite3 database file with sequence details of the marker accessions (marker.db).

Additional files including sequence details from BLAST databases, species list and taxids, and marker accession list are also provided in a separate folder.

Additional details

Getting NCBI blast databases

Users can use the update_blastdb.pl script from the BLAST suite to download the preformatted databases. For example to download nt blast databases use the command

update_blastdb.pl --decompress nt

Using custom blast databases

Users can also create custom blast databases for use in the workflow. The sequences must adhere to NCBI taxonomy. To create custom blast databases, use makeblastdb command from the BLAST suite. A taxid-map file connecting the sequence ids to the NCBI taxid must be provided with the -taxid_map option while creating the custom databases.

makeblastdb -in seq.fa -out blastdb/seq -dbtype 'nucl' -taxid_map taxa_map.txt -parse_seqids

A taxid map file would look like this

#
# cat taxa_map.txt
#
A 217026
B 335001
C 990597
D 990597

Makefile workflow

The tool is also implemented as a Makefile workflow. To run the workflow use the command

make create_db MARKER=COI TAXA=taxa.txt BLAST_DB=blastdb/demo

To get details on the usage of the Makefile, run the command make.

make