The CRISPOR docker container is based on Ubuntu 26 (Phusion, for docker) and includes Apache, Sqlite3, Python 3.9, a ton of Python dependencies for the efficiency scoring algorithms (keras, numpy, numba, scikit-learn, etc), a cronjob to clean up the temp files and CRISPOR itself. The software is installed under /data/www/crispor, which is also the htdocs directory for Apache. The scoring daemon is started on boot under /etc/system.d/crispor by initd.

To download and start the container and map the port 8080 on your machine to the container:

 docker run -d -p 8080:80 --name crispor-container maximilianh/crispor

You should then be able to access the container via http://localhost:8080 from the machine. There is no genome yet.

To download an existing genome from crispor.gi.ucsc.edu into the container:

 docker exec -it crispor-container /data/www/crispor/tools/crisporDownloadGenome hg38

To add a new genome to the container, either via NCBI accession or from a FASTA/GFF file:

 docker exec -it crispor-container /data/www/crispor/tools/crisporAddGenome ncbi GCA_052724335.1
 docker exec -it crispor-container /data/www/crispor/tools/crisporAddGenome fasta GWHBOWM00000000.genome.fasta.gz --gff GWHBOWM00000000.gff.gz --desc 'faAtrBel|Atropa belladonna|Belladonna deadly nightshade|CNCB GWHBOWM00000000'

A few more examples

 crisporAddGenome ucsc sacCer3
 crisporAddGenome ncbi 

 wget https://hgdownload.soe.ucsc.edu/goldenPath/sacCer2/bigZips/sacCer2.fa.gz
 wget https://hgdownload.soe.ucsc.edu/goldenPath/sacCer2/bigZips/genes/sacCer2.ensGene.gtf.gz
 crisporAddGenome fasta --fasta sacCer2.fa.gz --desc 'faSacCer2|Saccharomyces cerevisiae|yeast|sacCer2 from UCSC' --gff sacCer2.ensGene.gtf.gz

You can make fasta files from the host available for adding by including a bind mount when you execute docker run. For example, you can run

 docker exec -d -p 8080:80 -v /path/to/genome.fa:/tmp/genome.fa --name crispor-container maximilianh/crispor

When you run crisporAddGenome via docker exec, the genome will be available at /tmp/genome.fa. If you wish to include multiple genomes at the same time, you can use a directory with multiple fasta files instead: /path/to/genomes_dir:/tmp/genomes_dir. Then every file in genomes_dir will be available at /tmp/genomes_dir/genome_name.fa

The syntax for the --desc option is: "internalName|latinName|commonName|assembly version or description'

To get a linux shell in the running docker container:

 docker exec -it crispor-container /bin/bash

From outside the container, from a clone of the Github repo, I run this command to build the container and push it as a multi-architecture build, you do not have to do this, but it may be interesting:

 docker buildx build . --platform linux/amd64,linux/arm64 -t maximilianh/crispor:latest --push

Then I tag the container with the version. Since it's a multi-arch container, I cannot use the tag comment, but need to use buildx:

 docker buildx imagetools create --tag maximilianh/crispor:v5.2 maximilianh/crispor:latest

The last command automatically pushes the new tag to Docker hub.