defaults.toml

[workflow]
# Only print the final merged config and a list of stages to be submitted.
# Will skip any communication with Metamist, Hail Batch, and Cloud Storage, so
# the code can be run without permissions.
#dry_run = true

# Show a workflow graph locally or save to web bucket.
# Default is false, set to true to show the workflow graph.
#show_workflow = true

# Cohorts to use as inputs.
#input_cohorts = []

# Datasets to load inputs. If not provided, datasets will be determined

# Skip these stages:
#skip_stages = []

# Skip all other stages:
#only_stages = []

# Start from this stage:
#first_stages = []

# Finish at this stage:
#last_stages = []

# Force stage rerun
#force_stages = []

# Name of the workflow (to prefix output paths)
#name =

# Description of the workflow (to display in the Batch GUI)
#description =

# Suffix the workflow outputs location (`get_workflow().prefix`) with this string.
# By default, the hash of all input paths will be used.
#output_version = "0.1"

# Limit to data of this sequencing type
#sequencing_type = 'genome'

# Check input file existence. When they are missing,
# the `skip_sgs_with_missing_input` option controls whether such
# sequencing groups should be ignored, or it should raise an error.
check_inputs = true

# For the first (not-skipped) stage, if the input for a target does
# not exist, just skip this target instead of failing. E.g. if the first
# stage is Align, and `sequencing_group.alignment_input` for a sequencing group do not exist,
# remove this sequencing group, instead of failing. In other words, ignore sequencing groups
# that are missing results from skipped stages.
skip_sgs_with_missing_input = false

# Within jobs, check all in-job intermediate files for possible reuse.
# If set to False, will overwrite all intermediates. Used by `utils.can_reuse(path)`.
check_intermediates = true

# Before running a stage, check if input (i.e. expected outputs from required stages)
# already exist. If it exists, do not submit stage jobs.
check_expected_outputs = true

# Calling intervals (defauls to whole genome intervals)
# TODO: Remove from cpg_flow and move to cpg_utils, then remove from defaults.toml
#intervals_path =

# Map of stages to lists of sequencing groups, to skip for specific stages
# [workflow.skip_stages_for_sgs]
#CramQC = ['CPGaaa']

[hail]

# This is different from the cpg_flow Workflow dry_run option.
# This will create Hail Batch jobs, but will enable Hail Batch's dry_run option.
dry_run = false

# Delete temporary directories with intermediate files.
delete_scratch_on_exit = false

[resource_overrides]
# Override default resource requirements for unusually large seq data without
# demanding higher resources for all operations as standard. Examples below

# The value should be an int representing the memory in GB
# This is true for both memory and storage overrides
# job_mem_override = 50

# Other common overrides could be for number of cpus or GBs of storage
# job_cpu_override = 4
# job_storage_override = 100

# To use this override in the job python file do something like this:
# def my_job(
#     b: hb.Batch,
#     input_file: hb.ResourceFile,
#     job_attrs: dict | None = None,
#     output_path: Path | None = None,
#     second_output_path: Path | None = None,
#     fasta_reference: hb.ResourceGroup | None = None,
#     overwrite: bool = False,
# ) -> Job | None:
#     """
#     My super awesome custom job
#     """
#     job = b.new_job(job_name, job_attrs)
#     ...
#
#     # check for a memory override for this job
#     memory_override = get_config()['resource_overrides'].get('job_mem_override')
#     assert isinstance(memory_override, (int, type(None)))
#
#     resource = HIGHMEM.request_resources(ncpu=4, mem_gb=memory_override)
#     # Any other resource logic or modifications...
#
#     # This line sets the resource for the job created above
#     resource.set_to_job(j)
#     ....
#
#     return job

# Adding custom options for workflow stages and jobs
# You are more than welcome to add your own custom options, however they should not
# live in the defaults.toml file. Instead, create a new toml file in the same directory
# That you pass to the analysis runner. Any options you create following the toml syntax
# will be available to you in the workflow and job python files.
# For example, if you create a file called custom_options.toml with the following content:

# [custom]
# my_custom_option = "my_custom_value"

# You can access this value in your workflow or job python file like this:
# get_config().get('custom', {}).get('my_custom_option')