Writing Applications

⚡️Isabl Applications enable you to systematically deploy data science tools across thousands of Experiments in a metadata driven approach. Learn how to build them here.


Isabl Applications enable you to systematically deploy data science tools across thousands of Experiments in a metadata driven approach. The most important things to know about applications are:

  • Applications are agnostic to the underlying tools being utilized.

  • Applications can submit analyses to multiple compute environments (local, cluster, cloud).

  • Results are stored as analyses for which uniqueness is a function of the experiments used.

  • Once implemented, applications can be deployed across any subset of experiments in the database.

Isabl applications are not Workflow Management Systems (see what Isabl is not trying to solve). Instead they use metadata to systematically build and deploy any type of execution commands across thousands of experiments.

How Does an Application Look Like?

During this tutorial we will build a hello world application that show cases the functionalities and advantages of processing data with Isabl. Here is a really simple example of an Isabl application that echoes an experiment's sample identifier and it's raw data.

from isabl_cli import AbstractApplication
from isabl_cli import options

class HelloWorldApp(AbstractApplication):

    VERSION = "1.0.0"
    cli_options = [options.TARGETS]

    def get_command(self, analysis):
        experiment = analysis.targets[0]  
        return f"echo {experiment.sample.identifier} {experiment.raw_data} "

This application can now be executed system-wide using:

isabl apps hello-world --filters projects 102

cli_options enabled us to run the app across multiple experiments using RESTful API filters (i.e. --filters). We will learn more about how to link experiments with analyses later.

Analyses and Results

Results produced by applications are stored as analyses. The uniqueness of an analysis is determined by the experiments associated with it. Specifically, analyses can be linked to multiple targets and references experiments (e.g. tumor-normal pairs). The possibility of linking analyses to multiple experiments allow for a wide variety of experimental designs:

  • Single target analyses (e.g. quality control applications).

  • Tumor-normal pairs (e.g. variant calling applications).

  • One target vs. a pool of references (e.g. copy number applications).

  • Multiple targets agains multiple references (e.g. all vs. all contamination testing).

Importantly, if someone tries to run the same application over the same experiments, a new analysis won't be created and but the existing one will be retrieved.

Getting Started with Isabl Applications

You should store your applications and custom Isabl logic in your own python package. Cookiecutter Apps will help you bootstrap your own Isabl project:

# first make sure you have cookiecutter installed
pip install cookiecutter

# now lets bootstrap your project
cookiecutter https://github.com/isabl-io/cookiecutter-apps

# finally install you project in a new virtual environment
cd <project-name> && pip install -r requirements.txt

An example of a project generated with Cookiecutter Apps its available here. This project, and every project created with Cookiecutter Apps includes the hello world application described in this tutorial, check it out here. Now let's learn about writing apps!

Registering Applications

To make sure your applications are available when running isabl --help, make sure you add them to the client setting INSTALLED_APPLICATIONS:


Creating Applications

All Isabl Applications inherit from isabl_cli.AbstractApplicationand are configured using a class based approach. Your role is to override attributes and methods to drive the behavior of your app.

Applications are represented both with a python class and a database object. The database object is created and updated automatically when the application is run.

Versioning Applications

Applications are uniquely versioned by setting the NAME and VERSION attributes. The version of an application is not necessarily the version of the underlying tool being executed:

class HelloWorldApp(AbstractApplication):

    NAME = "Hello World"
    VERSION = "1.0.0"

A good strategy to version applications is to ask the question: are results comparable across experiments? An optimization (or bug fix) that doesn't change outputs might not require a version change.

Optionally you can also set ASSEMBLY and SPECIES to version the application as a function of a given genome assembly. This is particularly useful for NGS applications as often results are only comparable if data was analyzed against the same version of the genome:

class HelloWorldApp(AbstractApplication):

    NAME = "Hello World"
    VERSION = "1.0.0"
    ASSEMBLY = "GRCh37"

You can add additional metadata to be attached to the database object, such as an application description and URLs (or comma separated URLs):

class HelloWorldApp(AbstractApplication):
    application_description = "An App to show case different Isabl functionalities."
    application_url = "https://docs.isabl.io/writing-applications"

Application Settings

Applications can depend on multiple configurations such as paths to executables, references files, compute requirements, etc. These settings are explicitly defined using the application_settings dictionary:

class HelloWorldApp(AbstractApplication):

    application_import_strings = {"sym_link"}
    application_settings = {
        "echo_path": "echo",
        "default_message": "Hello World",
        "sym_link": "isabl_cli.utils.force_symlink"

If a setting is meant to be imported, include it in application_settings_import_strings.

Optional settings can be set to None whilst required but not yet defined settings can be set to theNotImplemented python object. Settings defined in the application python class are considered to be the default settings, yet they can be overridden using the database application field settings.

> HelloWorldApp().application.settings
   # settings as a function of Isabl CLI client's primary key
   1: {"echo_path": "/usr/local/bin/echo"}  

Note that application.settings are a function of the client's primary key. This enables you to run the sample application in different compute architectures. You can configure application.settings using the Django Admin site or the application method patch_application_settings:


If the value of a setting is a dictionary, the schema (i.e. keys) of that setting will be validated unless the dictionary contains "skip_check": True.

Validate Application Settings

You can make sure applications are properly configured by performing settings validation. To do so, simply define validate_settings and raise an AssertionError if something is not set properly:

from shutil import which

class HelloWorldApp(AbstractApplication):

    def validate_settings(self, settings):
        assert which(settings.echo_path), f"{settings.echo_path} not in PATH"

Theapplication_settings dictionary defines default settings, but during execution your app may have different settings for clients or environments. For example, you may have a small test reference file for testing and the real one for production. That's why you can define NotImplemented by default, but validate that it's in fact implemented on execution.

Running Applications

Applications can be launched from both the command line and from python (we will learn more about the latter in the operational automations guide).

Command Line Configuration

To support CLI capabilities you have to tell the application how to link analyses to experiments using command line options:

from isabl_cli import options

class HelloWorldApp(AbstractApplication):

    cli_help = "This is the Hello World App - a way to learn Isabl applications."
    cli_options = [options.TARGETS]

The attribute cli_options is set to a list of Click Options that will be used to retrieve experiments from the API and link them to new analyses. Out of the box, Isabl supports the following CLI options to retrieve experiments:

When these options are not adequate for your experimental design, you can implement get_experiments_from_cli_options. This function takes the evaluated cli_options and must return a list of tuples: one tuple per analysis, each tuple with 2 lists: the target experiments and the reference experiments. Here is a an example of an application that creates only one analysis linked to all Whole Genome experiments in a project:

from isabl_cli import api
import click

class HelloWorldApp(AbstractApplication):

    cli_options = [
        click.option("--project", help="project key", required=True),
        click.option("--method", help="technique method", default="WG"),

    def get_experiments_from_cli_options(self, **cli_options):
        project = cli_options["project"]
        method = cli_options["method"]
        filters = dict(projects=project, technique__method=method)
        targets = api.get_instances("experiments", **filters)
        return [(targets, [])]  # will result in only one analysis

By default, applications come with --force to remove and start analyses from scratch, --restart to run failed analyses again without trashing them, --local to run analyses locally, one after the other:

class HelloWorldApp(AbstractApplication):

    cli_allow_force = True
    cli_allow_restart = True
    cli_allow_local = True

As such, the CLI configuration for our Hello World app will result in the following help message:

$ isabl apps hello-world-1.0.0 --help

Usage: isabl apps hello-world-1.0.0 [OPTIONS]

  This is the Hello World App - a way to learn Isabl applications.

  --targets-filters       API filters for target experiments [required] <TEXT TEXT>...
  --commit                Submit application analyses.  [default: False]
  --force                 Wipe unfinished analyses and start from scratch.
  --restart               Attempt restarting failed analyses from previous checkpoint.
  --help                  Show this message and exit.

The --force flag will not completely remove the analyses, but it will move them to a temporary trash directory within the BASE_STORAGE_DIRECTORY. You may want to clean this location periodically using crontab -e:

crontab -e
# Clears trash directory.
0 0 * * *   source ~/.bash_profile &> /dev/null; rm -rf <replace with BASE_STORAGE_DIRECTORY>/.analyses_trash/* &> /dev/null;

You can usecli_options to include any other argument your app may need in order to successfully build and deploy data processing tools.

Validate Experiments Before Creating Analyses

Some of the advantages of metadata-driven applications is that we can prevent analyses that don't make sense, for example running a variant calling application on imaging data. Simply raise an AssertionError if something doesn't make sense, and the error message will be provided to the user:

class HelloWorldApp(AbstractApplication):

    def validate_experiments(self, targets, references):
        assert len(targets) == 1, "only one target experiment per analysis"
        assert targets[0].raw_data, "target experiment has no linked raw data"
        self.validate_dna_only(targets)  # multiple validators are readily available

Analyses are understood to be unique if their targets, references, and application are the same (as well as previously linked dependencies). If you need custom get or create logic, you can override the get_or_create_analyses method.

AbstractApplicationcomes with readily available validators that you may want to use. Here are some examples of commonly used ones:

  • validate_dna_only Check technique category isDNA.

  • validate_same_technique Validate experiments have same experimental technique.

  • validate_same_individual Check experiments come from same individual.

Building Commands Using Metadata

Now that we know how to link analyses to experiments, lets dive into creating data processing commands. Our only objective is to use the analysis and settings objects to build a shell command and return it as a string (ignore inputs for now, we will learn more about it when specifying application dependencies).

from os.path import join
import click

class HelloWorldApp(AbstractApplication):
    cli_options = [options.TARGETS, click.option("--message")]
    def get_command(self, analysis, inputs, settings):
        echo = settings.echo_path
        target = analysis.targets[0]
        message = settings.run_args.message or settings.default_message
        output_file = join(analysis.storage_url, "output.txt")
        input_file = join(analysis.storage_url, "input.txt")
        settings.sym_link(target.raw_data[0].file_url, input_file)
        return (
            f"bash -c '{echo} Sample: {target.sample.identifier} > {output_file}' && "
            f"bash -c '{echo} Message: {message} >> {output_file}' && "
            f"bash -c '{echo} Data: >> {output_file}' && "
            f"bash -c 'cat {input_file} >> {output_file}' "

All options passed in cli_options are available during get_command using the settings attribute run_args. In this simple example, we allowed the user to pass a custom --message.

This also means we can perform actions when restarting an analysis, which can be useful for cleaning up files:

from os.path import join, exists
from os import remove

class HelloWorldApp(AbstractApplication):
    cli_options = [options.TARGETS]

    def get_command(self, analysis, inputs, settings):
        if settings.restart:
            to_rm_path = join(analysis.storage_url, "run.lock")
            if exists(to_rm_path):
        return "echo 'Hello World!'"

Isabl is agnostic to compute architecture, get_command does not need to worry about HPC schedulers, or cloud architecture (e.g. LSF, AWS), its only role is to return a shell command.

Submitting Analyses to Compute Architectures

Isabl is agnostic of the compute infrastructure you're working on and can be configured to work with different batch systems (e.g. local, HPC, cloud). Currently, Isabl supports local, LSF, SGE, and Slurm submissions, how ever you can create a submitter for other schedulers.

Importantly Isabl is not a workflow management system or language like Toil, Bpipe, CWL, etc. Isabl however, can submit head jobs per analysis to a compute infrastructure.

Analyses Batch Submission

Isabl comes with prebuilt logic to submit thousands of analyses to LSF, SGE, and Slurmusing Job Arrays. To do so simply set the Isabl CLI setting SUBMIT_ANALYSES as follows:

// IBM's LSF
"SUBMIT_ANALYSES": "isabl_cli.batch_systems.lsf.submit_lsf",

// Sun Grid Engine
"SUBMIT_ANALYSES": "isabl_cli.batch_systems.sge.submit_sge",

// Slurm
"SUBMIT_ANALYSES": "isabl_cli.batch_systems.slurm.submit_slurm",

This submitter can check for the following configurations in SUBMIT_CONFIGURATION:

The method get_requirements must take the application and a list of targets' technique methods (which are submitted together in the same job array):

def get_lsf_requirements(app, targets_methods):
    if isinstance(app, apps.HelloWorldApp):
        memory = 10 if "WG" in targets_methods else 1
        return f"-n {1} -R 'rusage[mem={memory}]'"

Other Schedulers

You can implement SUBMIT_ANALYSES functions for other schedulers, the function must take a list of tuples, each tuple being an analysis and the analysis head job script.

Applications Run by Multiple Users

Isabl applications can be run by multiple users in the same unix group. However, if applications are run by users different than the ADMIN_USER and are not re-runnable, then analyses will be set to FINISHED instead of SUCCEEDED. isabl process-finished can be run by the ADMIN_USER to copy and own the results and set the permissions to read-only whilst updating analyses status to SUCCEEDED. We recommend you add the following cron task in the ADMIN_USER profile using crontab -e:

crontab -e
# Change analyses permissions and updates them to SUCCEEDED.
*/30 * * * * source ~/.bash_profile &> /dev/null; isabl process-finished &>> ~/moving.log

Running Applications from Python

Apps can programmatically be triggered from python using the run method:

    tuples=[([target_experiment], [])],
    run_args=dict(message="custom message"),

Tip: this is useful when creating operational automations!

Application Results

You can provide an specification your application results using the application_results dictionary. Each key is a result id and the value is a dictionary with specs of the result:

class HelloWorldApp(AbstractApplication):

    application_results = {
        "input": {
            "frontend_type": "text-file",
            "description": "Symlink to experiment raw data.",
            "verbose_name": "Hello World Input",
            "external_link": "https://en.wikipedia.org/wiki/Symbolic_link",
        "output": {
            "frontend_type": "text-file",
            "description": "Sample id, hello world message, and content of raw data.",
            "verbose_name": "Hello World Result",
            "external_link": "https://hello.world/",
        "count": {
            "frontend_type": "number",
            "description": "Count of characters in output file.",
            "verbose_name": "Characters Count",

By default, all applications come with 3 default settings command_script, command_log, and command_err. These point to the standard output, standard error, and analysis head job command, respectively.

Results can be paths to files, strings (e.g. MD5s), numbers, and any other serializable value. Here is a full list of the different specifications a result can have:

  • description Information about the result (required)

  • verbose_name Name displayed for the result in the results list (required)

  • optionalIf Falseand result is missing, an alert will be shown online (optional)

  • external_link URL to a resource that may explain about the result (optional)

  • frontend_type Defines how the result should be displayed in the frontend.

By default, analysis results are protected upon completion (i.e. permissions are set to read only). If you want your application to be re-runnable indefinitely, set application_protect_results = False.

Frontend Result Types

Here is a full list of the result types that are supported for rendering in Isabl Web:

Re-runnable Applications

By default, analysis results are protected upon completion (i.e. permissions are set to read only). If you want your application to be re-runnable indefinitely, set:

class HelloWorldApp(AbstractApplication):

    application_protect_results = False

Databasing Analysis Results

When application_results is defined, you must implement get_analysis_results. This method must return a serializable dictionary of results and its only run after the analysis has been completed successfully. For our example it can be something like:

class HelloWorldApp(AbstractApplication):

    def get_analysis_results(self, analysis):
        output = join(analysis.storage_url, "output.txt")

        with open(output) as f:
            count = sum(len(i) for i in f)

        return {
            "input": join(analysis.storage_url, "input.txt"),
            "output": output,
            "count": count

Project and Individual Level Auto-merge

Isabl applications can produce auto-merge analyses at a project and individual level. For example, you may want to merge variants whenever new results are available for a given project, or update quality control reports when a new sample is added to an individual. A newly versioned analysis will be created for each type of auto-merge and your role is to take a list of succeeded analysis and implement the merge logic.

class HelloWorldApp(AbstractApplication):

    application_project_level_results = {
        "merged": {
            "frontend_type": "text-file",
            "description": "Merged output files.",
            "verbose_name": "Merged Output Files",
        "count": {
            "frontend_type": "number",
            "description": "Count of characters for merged output.",
            "verbose_name": "Merged Outth put Characters Count",
    def merge_project_analyses(self, analysis, analyses):
        with open(join(analysis.storage_url, "merged.txt"), "w") as f:
            for i in analyses:
                with open(i.results.output) as output:

    def get_project_analysis_results(self, analysis):
        merged = join(analysis.storage_url, "merged.txt")

        with open(merged) as f:
            count = sum(len(i) for i in f)

        return {"merged": merged, "count": count}

The first argument in merge_project_analyses, is the project level analysis, which is unique per project and application. The second argument is a list of all completed analyses of this application for a given project. Your role is to merge analyses output into the project level analysis directory. We need to define similar methods for the Individual level auto merge. Lets say that our project-level merge logic is the same for individuals, then we can simply do:

class HelloWorldApp(AbstractApplication):

    # reuse the project merge logic at the individual level
    application_individual_level_results = application_project_level_results
    merge_individual_analyses = merge_project_analyses
    get_individual_analysis_results = get_project_analysis_results

If at any arbitrary time you want to test the auto-merge logic, use any of these two commands:

# for project level auto merge
isabl merge-project-analyses --project <project id> --application <application id>

# for individual level automerge
isabl merge-individual-analyses --individual <individual id> --application <application id>

Please note that merged output will always be stored in the same analysis for a given project or individual and application. Furthermore, you can validate analyses before running the merge operation by implementing validate_project_analyses, and validate_individual_analyses.

Submitting Merge Analysis to A Compute Architecture

Merge operations are triggered automatically when the last analysis that is meant to be merged finish running. By default, the merge operation will be conducted right after the analysis is patched to SUCCEEDED. However, you can define how merge analyses are submitted using Isabl CLI setting SUBMIT_MERGE_ANALYSIS. For example in LSF:

import subprocess

def submit_merge_analysis_to_lsf(instance, application, command):
    """Submit project merge to LSF."""
    command = ["bsub", "-n", "1", "-W", "40000", "-M", "32"] + command.split()
    print("Submited merge analysis using: " + " ".join(command))

Here is an example for SGE:

import subprocess

def submit_merge_analysis_to_sge(instance, application, command):
    """Submit project merge to SGE."""
    command = f"qsub -l h_vmem=[32G] << EOF\n{command}\nEOF\n"
    subprocess.check_call(command, shell=True)
    print(f"Submited project level merge with: {command}")

Optional Functionality

This section lists additional optional functionality supported by Isabl applications. Particularly, dependencies on other applications, after completion analyses status, and unique analyses per individual.

Analyses Inputs and Dependencies on Other Applications

Application inputs are analysis-specific settings (settings are the same for all analyses, yet inputs are potentially different for each analysis). Inputs can be formally defined using application_inputs, inputs set to NotImplemented are considered required and must be resolved using get_dependencies:

from isabl_cli import utils

class HelloWorldApp(AbstractApplication):
    application_inputs = {"previously_generated_result": NotImplemented}

    # get dictionary of inputs this function is run before get_command.
    # must return a tuple: (list of analyses dependencies primary keys, inputs dict)
    def get_dependencies(self, targets, references, settings):
        result, analysis_key = utils.get_result(  # return result value and key of analysis that generated it
            experiment=targets[0],                # get result for the first target experiment
            application_key=123,                  # primary key of previous app in database
            result_key="a_result_name",           # name of result in the app definition

        return [analysis_key], {"previously_generated_result": result}

The main objective of application_inputs and get_dependenciesis to retrieve results and analyses that should be linked to the newly created analysis. Linked analyses are accessible from the analysis detail frontend view.

Get After Completion Status

In certain cases you don't want your analyses to be marked as SUCCEEDED after completion, as you may want to flag them for manual review or leave them to know that you need to run an extra step on them. For these cases, you may want to set the after-completion status to IN_PROGRESS:

class HelloWorldApp(AbstractApplication):
    def get_after_completion_status(self, analysis):
        return "IN_PROGRESS"

Unique Analysis Per Individual

It is possible to create applications that are unique at the individual level. To do so set unique_analysis_per_individual = True. A good example of a unique per individual application could be a patient centric report that aggregates results across all samples. If you are interested on how analyses for these applications are created, give a look at AbstractApplication.get_individual_level_analyses.

Individual Level Auto-Merge and Unique Analyses Per Individual are different concepts. Applications that require a unique analysis per individual don't support individual level auto-merge.

Get Notified When Analyses Fail

You can configure Isabl API to periodically check if any analysis has failed and send you email notifications. To do so, head to the admin site at /admin/django_celery_beat/periodictask/add/ and in Task (registered) select isabl_api.tasks.report_status_change_task, then create a 1 hour interval, and provide the following Keyword arguments {"status": "FAILED", "seconds": 3600} (i.e. every hour check how many analyses failed in the past hour):

Testing Applications

Our goal is to make it extremely easy to test your applications. Ideal apps can be tested locally, with fake/dummy data using factory-created database instances. Isabl CLI and Cookiecutter Apps come with a range of utilities to help you test your applications.

Useful Pytest Fixtures

If you created your project using Cookiecutter Apps, the following pytest fixtures are available to you:

Look at the test example to learn how use these fixtures.

Creating Fake Data and Metadata

Isabl CLI comes with a full set of factories that facilitate the creation of fake metadata. Here is an example of how to create two experiments for the same sample:

from isabl_cli import api
from isabl_cli import factories

meta_data = factories.ExperimentFactory()
meta_data["sample"]["individual"]["species"] = "HUMAN"
experiment = api.create_instance("experiments", **meta_data)
assert experiment.pk > 0, "failed to create database instance"

We recommend limiting use of these factories to development instances of Isabl API. By default, ISABL_API_URL is set to

Being able to actually run the applications (i.e. passing --commit) during testing might be something valuable to you. In the case of Next Generation Sequencing, for example, you could create fake BAMs and really small reference genomes (few KBs) to test variant calling applications.

Application Test Example

Here is a comprehensive example to test our HelloWorldApp, projects created Cookiecutter Apps will include this test:

def test_hello_world_app(tmpdir, datadir, commit):
    # path to hello_world test data
    hello_world_datadir = join(datadir, "hello_world")
    raw_data = [
            file_url=join(hello_world_datadir, "test.txt"),

    # overwrite default configuration for the default client
    meta_data = factories.ExperimentFactory(raw_data=raw_data)
    meta_data["sample"]["individual"]["species"] = "HUMAN"
    meta_data["storage_url"] = hello_world_datadir
    experiment = api.create_instance("experiments", identifier="a", **meta_data)

    # create files if you may test with real data at some point
    app = HelloWorldApp()
    app.application.settings.default_client = {
        "default_message": "Hello from Elephant Island.",
        "echo_path": tmpdir.docker("ubuntu", "echo")

    # run application and make sure results are reported
        tuples=[([experiment], [])],
        results=["output", "count", "input"],

The actual test is much more comprehensive. It creates two experiments per individual and validates that auto-merge was actually conducted, that the application is re-runnable, and that the command line configuration works well.

Last updated