Kipoi (pronounce: kípi; from the Greek κήποι: gardens) is an API and a repository of ready-to-use trained models for genomics. It currently contains 2201 different models, covering canonical predictive tasks in transcriptional and post-transcriptional gene regulation. Kipoi's API is implemented as a python package, and it is also accessible from the command line.
In this tutorial example, we will run a genomics model on Bacalhau.
To get started, you need to install the Bacalhau client, see more information here
To run locally you need to install kipoi-veff2. You can find out the information about installing and usage here
In our case this will be the following command:
To run Genomics on Bacalhau we need to set up a Docker container. To do this, you'll need to create a Dockerfile and add your desired configuration. The Dockerfile is a text document that contains the commands that specify how the image will be built.
We will use the kipoi/kipoi-veff2:py37 image and perform variant-centered effect prediction using the kipoi_veff2_predict tool.
See more information on how to containerize your script/app here
The docker build command builds Docker images from a Dockerfile.
Before running the command replace:
hub-user with your docker hub username. If you don’t have a docker hub account follow these instructions to create a Docker Account, and use the username of the account you created
repo-name with the name of the container, you can name it anything you want
tag this is not required but you can use the latest tag
In our case:
Next, upload the image to the registry. This can be done by using the Docker hub username, repo name or tag.
After the repo image has been pushed to Docker Hub, we can now use the container for running on Bacalhau. To submit a job for generating genomics data, run the following Bacalhau command:
In this example, a model from github.com is downloaded during the job execution. In order to do this, use the --network full flag when describing the job, and --job-selection-accept-networked when starting the compute node on which the job will be executed.
Note, that in the demo network, nodes do not accept jobs that require full network access. Consider creating your own private network.
Let's look closely at the command above:
bacalhau docker run: call to Bacalhau
jsacex/kipoi-veff2:py37: the name of the image we are using
kipoi_veff2_predict ./examples/input/test.vcf ./examples/input/test.fa ../outputs/output.tsv -m "DeepSEA/predict" -s "diff" -s "logit": the command that will be executed inside the container. It performs variant-centered effect prediction using the kipoi_veff2_predict tool
./examples/input/test.vcf: the path to a Variant Call Format (VCF) file containing information about genetic variants
./examples/input/test.fa: the path to a FASTA file containing DNA sequences. FASTA files contain nucleotide sequences used for variant effect prediction
../outputs/output.tsv: the path to the output file where the prediction results will be stored. The output file format is Tab-Separated Values (TSV), and it will contain information about the predicted variant effects
-m "DeepSEA/predict": specifies the model to be used for prediction
-s "diff" -s "logit": indicates using two scoring functions for comparing prediction results. In this case, the "diff" and "logit" scoring functions are used. These scoring functions can be employed to analyze differences between predictions for the reference and alternative alleles.
When a job is submitted, Bacalhau prints out the related job_id. We store that in an environment variable so that we can reuse it later on.
The same job can be presented in the declarative format. In this case, the description will look like this:
The job description should be saved in .yaml format, e.g. genomics.yaml, and then run with the command:
Job status: You can check the status of the job using bacalhau job list.
When it says Published or Completed, that means the job is done, and we can get the results.
Job information: You can find out more information about your job by using bacalhau job describe.
Job download: You can download your job results directly by using bacalhau job get. Alternatively, you can choose to create a directory to store your results. In the command below, we created a directory (results) and downloaded our job output to be stored in that directory.
To view the file, run the following command:
If you have questions or need support or guidance, please reach out to the Bacalhau team via Slack (#general channel).