This pull requests introduced a new workflow into cytosnake known as dp_timelapse_process,where it processing morphological features extracted by Deep Profiler from time lapse images.
When adding this feature, also some changes had to occur in order to maintain the current usage experience of cytosnake
Module Updates!
New modules added!
setup_init.py → Handles different variations on how to setup files for workflows.
config_utils.py → handles interactions like (loading, writing) configs with cytosnake
feature_utils.py → Normally functions that focuses on dealing features is normally done in pycytominer, however, functions that are not implemented pycytominer will be placed in this module.
Module changes
errors.py → new errors added WorkflowNotFoundError and BaseFileNotFound
Updated configs!
New configs added!
dp_aggregator_config.yaml → responsible for aggregating Deep Profiler features
dp_data_configs.yaml → responsible for creating the DeepProfiler data type in order to use Deep Profiler specific functions and methods.
changes in configs
configuration.yaml → includes the newly added configs
Scripts
New scripts added!
dp_aggregate.py →aggregates deep profiler features
dp_normalize.py →normalize deep profiler features
build_dp_consnsus.py →builds a consensus from deep profiler features
Environment Files
New env file added!
dp_process.yaml → used for processing deep profiler features
Additional Notes
Issue: Repetitive code in the newly added scripts
New scripts contain repetitive functions calls to load config files.
Since cytosnake is not pip install-able (yet). The modules cannot be imported into script workflows since we are declaring different environments in snakemake.
Therefore, the solution to this repetitive code is to make cytosnakepip install-able and add it to our workflow environment files.
also, @axiomcura, can you please update the title of this PR to correspond to the updated workflow name. (it's no longer timelapse specific, i believe)
This pull requests introduced a new workflow into
cytosnake
known asdp_timelapse_process,
where it processing morphological features extracted by Deep Profiler from time lapse images.When adding this feature, also some changes had to occur in order to maintain the current usage experience of
cytosnake
Module Updates!
New modules added!
setup_init.py
→ Handles different variations on how to setup files for workflows.config_utils.py
→ handles interactions like (loading, writing) configs withcytosnake
feature_utils.py
→ Normally functions that focuses on dealing features is normally done inpycytominer
, however, functions that are not implementedpycytominer
will be placed in this module.Module changes
errors.py
→ new errors addedWorkflowNotFoundError
andBaseFileNotFound
Updated configs!
New configs added!
dp_aggregator_config.yaml
→ responsible for aggregating Deep Profiler featuresdp_data_configs.yaml
→ responsible for creating theDeepProfiler
data type in order to use Deep Profiler specific functions and methods.changes in configs
configuration.yaml
→ includes the newly added configsScripts
New scripts added!
dp_aggregate.py
→aggregates deep profiler featuresdp_normalize.py
→normalize deep profiler featuresbuild_dp_consnsus.py
→builds a consensus from deep profiler featuresEnvironment Files
New
env
file added!dp_process.yaml
→ used for processing deep profiler featuresAdditional Notes
Issue: Repetitive code in the newly added scripts
New scripts contain repetitive functions calls to load config files.
Since
cytosnake
is notpip
install-able (yet). The modules cannot be imported into script workflows since we are declaring different environments insnakemake
.Therefore, the solution to this repetitive code is to make
cytosnake
pip
install-able and add it to our workflow environment files.