Go fetch

Simple python script to fetch organelle or ribosomal reference sequences from NCBI for a given taxonomy.

The script was originally designed to format reference data for GetOrganelle (i.e. seed and gene format). You can specify if you want sequences formatted for GetOrganelle using the --getorganelle parameter.

Dependencies

The script requires getorganelle, biopython and trf. These can be installed in a conda environment:

conda create -n go_fetch -c bioconda getorganelle biopython trf

Input

Below is a summary of the main input parameters.

  -h, --help            show this help message and exit
  --taxonomy TAXONOMY   Taxonomy of rank to search for e.g. "Arabidopsis"
  --target TARGET       Target sequence type. Options = [chloroplast, mitochondrion, ribosomal, ribosomal_complete].
  --db DB               Database to search. Options = [refseq, genbank].
  --min MIN             Minimum number of target sequences to download.
  --max MAX             Maximum number of target sequences to download. 
  --seed SEED           Seed used for subsampling.
  --output OUTPUT       Output directory.
  --overwrite           Overwrite output directory.
  --getorganelle        Format seed and gene database for get organelle.
  --email EMAIL         Email for Entrez.
  --api API             API key for NCBI. Optional.

How it works

The user provides a taxonomy as a taxonomic name (e.g. "Arabidopsis thaliana") or NCBI taxonomy ID (e.g. "3702"). However, it is reccomended to use NCBI taxonomy IDs, as some unrelated groups share names, for example the name "Drosophila" is used for both a genus of fruit flies and basidiomycete fungi.

The script will count the number or target sequences available for the user given taxonomy. If there are not enough sequence available based on the --min parameter, the script will count the number of sequences available for the parent rank, until the minimum threshold is reached.

If the maximum number of sequences is exceeded, based on the --max parameter, the script will subsample the children lineages, aiming for an even sampling across children.

Note that the --target option ribosomal will search for sequences with any of the ribosomal annotations (28S/25S, 18S, 5.8S), whereas ribosomal_complete will try to find sequences with all annotations.

Example usage

Below are simple examples of how to use the script.

# Arabidopsis thaliana chloroplast
./go_fetch.py \
   --taxonomy 3702 \
   --target chloroplast \ 
   --db genbank \
   --min 5 --max 10 \
   --output arabidopsis_chloroplast \
   --overwrite \
   --getorganelle \
   --email [email protected]

# Arabidopsis thaliana complete ribosomal
./go_fetch.py \
   --taxonomy "Arabidopsis thaliana" \
   --target ribosomal_complete \
   --db genbank \
   --min 5 --max 10 \
   --output arabidopsis_ribosomal_complete \
   --overwrite \
   --getorganelle \
   --email [email protected]

# Drosophila melanogaster mitochondrion
./go_fetch.py \
   --taxonomy 7227 \
   --target mitochondrion \
   --db genbank \
   --min 5 --max 10 \
   --output drosophila_mitochondrion \
   --overwrite \
   --getorganelle \
   --email [email protected]

# Drosophila melanogaster ribosomal
./go_fetch.py \
   --taxonomy 7227  \
   --target ribosomal \
   --db genbank \
   --min 5 --max 10 \
   --output drosophila_ribosomal \
   --overwrite \
   --getorganelle \
   --email [email protected]