dcarrillox edits

Currently, the path to the sqlite database file is hardcoded in CrisprOpenDB_HostID.py (lines 102-103) so it needs to be in CrisprOpenDB/SpacersDB/ as stated in the README.md . :

if self._connection is None:
    self._connection = CrisprOpenDB.CrisprOpenDB(os.path.join("CrisprOpenDB", "SpacersDB", "CrisprOpenDB.sqlite"))

This is a problem since I want to put CrisprOpenDB in a Singularity container and CrisprOpenDB.sqlite is too heavy for that. I modify the code so the sqlite file can be outside of the repo. blastdb and fastadb files should not be a problem since there already options for specifying their paths.

Code modifications:

CL_Interface.py
- Line 16: added option -q so user can specify path to CrisprOpenDB.sqlite
- Line 38, old 37: added args.sqlitedb as argument for PhageHostFinder()
CrisprOpenDB_HostID.py
- Lines 12-17: added sqlite_db to the definition of the Class
- Line 104, old 103: change the hardcoded path to the variable sqlite_db

It worked with this test:

$ python CrisprOpenDB/CL_Interface.py -i  CrisprOpenDB/TestGenomes/KJ489400.fasta -b SpacersDB -q CrisprOpenDB.sqlite
('KJ489400.1', 'Bacillus', 1)

Singularity

I create a Singularity container with the modifications. Definition file CrisprOpenDB.def for building the container is also in the repository. To do so:

$ sudo singularity build crisprpendb.sif CrisprOpenDB.def

Container is also available from library://dcarrillo/default/crispropendb:0.1 .

IMPORTANT: blastdb has to be built with the makeblastdb provided in the container, which is blast 2.10.1. This is to avoid version incompatibility issues. To do so:

$ singularity run library://dcarrillo/default/crispropendb:0.1 makeblastdb -in SpacersDB.fasta -out SpacersDB -dbtype nucl

CrisprOpenDB

This command-line host prediction tool was implemented to run predictions for large numbers of phage genomes and to offer a more customized host prediction process. The method is described in the following paper, that you should cite in publications:

Moïra B Dion, Pier-Luc Plante, Edwige Zufferey, Shiraz A Shah, Jacques Corbeil, Sylvain Moineau, Streamlining CRISPR spacer-based bacterial host predictions to decipher the viral dark matter, Nucleic Acids Research, Volume 49, Issue 6, 6 April 2021, Pages 3127–3138, https://doi.org/10.1093/nar/gkab133

Prerequisites

First, download or clone this repository. The easiest way is using git clone https://github.com/plpla/CrisprOpenDB.git.

Next, we recommend to create a virtual environment to run the tool. You can set up a conda environment using the conda_env.txt file in the initial directory and the following command:

conda create -c conda-forge --name CrisprOpenDB_env --file conda_env.txt

Don't forget to activate the environment to use it: conda activate CrisprOpenDB_env. You can now install the tool: python setup.py install.

To use this program, you need to download the spacer database and the sqlite file (http://crispr.genome.ulaval.ca/dash/PhageHostIdentifier_DBfiles.zip) and unzip the files in the CrisprOpenDB/SpacersDB/ directory. Note that files are quite large. The download size is about 800Mo for the compressed file. Once unzipped, file sizes will be approximately 600Mo for the spacer database and 3.8Go for the sqlite file. If you plan on using Blast to run the program, you need to build the blast database. Instructions for this optional step are at the end of this README file with instructions on how to install Blast if it not already installed on your computer.

Once these steps are complete, don't forget to go back to the initial directory to run the program. You can now try your installation with one of our test genome, as explained below.

Running

To run the program, you must launch the CL_Interface.py file in the working directory. Here is an example of how to run the program using a Salmonella phage genome and a number of mismatches of 2:

python CL_Interface.py -i Salmonella_161.fasta -m 2

Test genomes

We provide test genomes for each level of prediction in the TestGenomes/ directory. These will allow you to make sure the tool has been properly set up before moving on with your personal analyses. The commands you should use to perform the tests as well as the expected results are shown down below.

Level 1 prediction:

python CL_Interface.py -i TestGenomes/KJ489400.fasta

('KJ489400.1', 'Bacillus', 1)

Level 2 prediction:

python CL_Interface.py -i TestGenomes/AY133112.fasta

('AY133112.1', 'Vibrio', 2)

Level 3 prediction:

python CL_Interface.py -i TestGenomes/MT074469.fasta

('MT074469.1', 'Salmonella', 3)

Level 4 prediction (note that level 4 predictions do not return a genus):

python CL_Interface.py -i TestGenomes/MT074470.fasta

('MT074470.1', 'Enterobacterales', 4)

Options and additional information for installation

Alignment can be done using blastn or fasta36. To install Blast inside the Anaconda environment run:

conda activate CrisprOpenDB_env
conda install -c bioconda blast

If using BLAST, you also need to use makeblastdb before running the phage host identification tool. Here is the command line you should use when running makeblastdb from the CrisprOpenDB/SpacersDB/ directory:

makeblastdb -in SpacersDB.fasta -dbtype nucl -out SpacersDB

If you wish, you can also provide your own BLAST or FASTA database to perform the alignment.

If you would like to manually further explore a prediction, you can use the -t, --table option. This will send the table containing both blast results and spacers information extracted from the spacers database to an external CSV file. Since one file per phage genome is issued, we recommend that you do not use this option when first running the tool (especially with large datasets), but rather use it afterwards if you need further information on a specific genome.

dcarrillox/CrisprOpenDB