/viroseqs

abricate formatted database to detect phage genes. Messily constructed from the databases used in Prophet

viroseqs

viroseqs db was obtained by installing ProphET and then converting the nucleotide tmp files into an abricate formatted database

https://github.com/jaumlrc/ProphET

I use this to detect phage genes in bacterial assemblies using abricate

Notes:

  1. I removed some seemingly redundant sequences. These had the same accession but different lengths, I kept the longer of the two
  2. There is probably a more appropriate way to format this database for abricate.
  3. There were over 60,000 sequences in the original database, I ran cd-hit on the originals (default settings) to cluster and reduce the size. I think there are 40,000 sequences in there now

How to setup this database for use with abricate:

  1. Download and unzip the viroseqs_90.fasta.gz file from this repo.
  2. Ceate a 'viroseqs' directory where abricate looks for databases. Can run the abricate --help command to find this path. It should be listed in the DATABASES section, next to the --datadir option.
  3. Move the decompressed fasta into this new directory with the name sequences.
  4. Run abricate --setupdb.

For example, this is how I configured it on my system:

cd ~/reference/
git clone https://github.com/Jtrachsel/viroseqs.git
cd viroseqs
gunzip viroseqs_90.fasta.gz
cd /home/julian.trachsel/miniconda3/db # wherever your abricate install looks for dbs
mkdir viroseqs
cp ~/reference/viroseqs/viroseqs_90.fasta ./viroseqs/sequences
abricate --setupdb