Add a baseline method

A baseline method (or control method) is used to test the relative performance of all other methods, and also as a quality control for the pipeline as a whole. A baseline method can either be a positive control or a negative control. The positive control and negative control methods set a maximum and minimum threshold for performance, so any new method should perform better than the negative control methods and worse than the positive control method.

This guide will show you how to create a new Viash component. In the following we will show examples for both Python and R. Note that the Label Projection task is used throughout the guide, so make sure to replace any occurrences of "label_projection" with your task of interest.

Tip

Make sure you have followed the “Getting started” guide.

Step 1: Create a new component

Use the create_component component to start creating a new baseline method.

viash run src/common/create_component/config.vsh.yaml -- \
  --task label_projection \
  --type control_method \
  --name my_method_py \
  --language python

This creates a new folder at src/label_projection/control_methods/my_method_py containing a Viash config and a script.

tree src/label_projection/control_methods/my_method_py
    ├── script.py         Script for running the method.
    ├── config.vsh.yaml   Config file for method.
    └── ...               Optional additional resources.
viash run src/common/create_component/config.vsh.yaml -- \
  --task label_projection \
  --type control_method \
  --name my_method_r \
  --language r

This creates a new folder at src/label_projection/control_methods/my_method_r containing a Viash config and a script.

tree src/label_projection/control_methods/my_method_r
    ├── script.R          Script for running the method.
    ├── config.vsh.yaml   Config file for method.
    └── ...               Optional additional resources.
  • A config file contains metadata of the component and the dependencies required to run it. In steps 2 and 3 we will fill in the required information.
  • A script contains the code to run the method. In step 4 we will edit the script.
Tip

Use the command viash run src/common/create_component/config.vsh.yaml -- --help to get information on all of the parameters if the create_component component.

Step 2: Fill in metadata

The Viash config contains metadata of your method, which script is used to run it, and the required dependencies.

Generated config file

This is what the config.vsh.yaml generated by the create_component component looks like:

Contents of config.vsh.yaml
# The API specifies which type of component this is.
# It contains specifications for:
#   - The input/output files
#   - Common parameters
#   - A unit test
__merge__: ../../api/comp_control_method.yaml

functionality:
  name: my_method_py

  # Metadata for your component (required)
  info:
    pretty_name: My Method Py
    summary: 'FILL IN: A one sentence summary of this method.'
    description: 'FILL IN: A (multiline) description of how this method works.'
    preferred_normalization: log_cpm

  # Component-specific parameters (optional)
  # arguments:
  #   - name: "--n_neighbors"
  #     type: "integer"
  #     default: 5
  #     description: Number of neighbors to use.

  # Resources required to run the component
  resources:
    # The script of your component
    - type: python_script
      path: script.py
platforms:
  - type: docker
    image: python:3.10
    # Add custom dependencies here
    setup:
      - type: python
        pypi: anndata~=0.8.0
  - type: nextflow
    directives:
      label: [midmem, midcpu]
Contents of config.vsh.yaml
# The API specifies which type of component this is.
# It contains specifications for:
#   - The input/output files
#   - Common parameters
#   - A unit test
__merge__: ../../api/comp_control_method.yaml

functionality:
  name: my_method_r

  # Metadata for your component (required)
  info:
    pretty_name: My Method R
    summary: 'FILL IN: A one sentence summary of this method.'
    description: 'FILL IN: A (multiline) description of how this method works.'
    preferred_normalization: log_cpm

  # Component-specific parameters (optional)
  # arguments:
  #   - name: "--n_neighbors"
  #     type: "integer"
  #     default: 5
  #     description: Number of neighbors to use.

  # Resources required to run the component
  resources:
    # The script of your component
    - type: r_script
      path: script.R
platforms:
  - type: docker
    image: eddelbuettel/r2u:22.04
    # Add custom dependencies here
    setup:
      - type: apt
        packages:
          - libhdf5-dev
          - libgeos-dev
          - python3
          - python3-pip
          - python3-dev
          - python-is-python3
      - type: python
        pypi: anndata~=0.8.0
      - type: r
        cran: anndata
  - type: nextflow
    directives:
      label: [midmem, midcpu]

Required metadata fields

Please edit functionality.info section in the config file to fill in the necessary metadata.

functionality.name

A unique identifier for the method. Must be written in snake case. Example: my_new_method.

functionality.info.pretty_name

A label for the method used for visualisations and documentation. Example: "My new method".

functionality.info.subtype

Whether the method is a "positive_control" or a "negative_control".

functionality.info.summary

A one sentence summary of the method. Used for creating short overviews of the components in a task.

functionality.info.description

An explanation for how the method works. Used for creating reference documentation of a task.

functionality.info.preferred_normalization

Which normalization method a component prefers. Possible values are l1_sqrt, log_cpm, log_scran_pooling, sqrt_cpm. Each value corresponds to a normalization component in the directory src/datasets/normalization.

__merge__

The file specified in this field contains information regarding the input and output arguments of the component, as well as a unit test to ensure that the component is functioning properly. Normally you don’t need to change this if you gave the right arguments to the create_component component.

Step 3: Add dependencies

Each component has it’s own set of dependencies, because different components might have conflicting dependencies.

In the platforms section of the config file update the setup definition that describes the packages that need to be installed in the Docker image and are required for your method to run. Note that both anndata~=0.8.0 and pyyaml are necessary Python package dependencies.

Please check out this guide for more information on how to add extra package dependencies.

Note

Tip: After making changes to the components dependencies, you will need to rebuild the docker container as follows:

viash run src/label_projection/control_methods/my_method_py/config.vsh.yaml -- \
  ---setup cachedbuild
[notice] Building container 'ghcr.io/openproblems-bio/label_projection/control_methods/my_method_py:dev' with Dockerfile
output
[notice] Building container 'ghcr.io/openproblems-bio/label_projection/control_methods/my_method_py:dev' with Dockerfile

Step 4: Edit script

A component’s script typically has five sections:

  1. Imports and libraries
  2. Argument values
  3. Read input data
  4. Generate results
  5. Write output data to file

Generated script

This is what the script generated by the create_component component looks like:

Contents of script.py
import anndata as ad

## VIASH START
par = {
  'input_train': 'resources_test/label_projection/pancreas/train.h5ad',
  'input_test': 'resources_test/label_projection/pancreas/test.h5ad',
  'input_solution': 'resources_test/label_projection/pancreas/solution.h5ad',
  'output': 'output.h5ad'
}
meta = {
  'functionality_name': 'my_method_py'
}
## VIASH END

print('Reading input files', flush=True)
input_train = ad.read_h5ad(par['input_train'])
input_test = ad.read_h5ad(par['input_test'])
input_solution = ad.read_h5ad(par['input_solution'])

print('Preprocess data', flush=True)
# ... preprocessing ...

print('Train model', flush=True)
# ... train model ...

print('Generate predictions', flush=True)
# ... generate predictions ...

print("Write output AnnData to file", flush=True)
output = ad.AnnData(
  obs={
    'label_pred': obs_label_pred
  },
  uns={
    'dataset_id': input_train.uns['dataset_id'],
    'normalization_id': input_train.uns['normalization_id'],
    'method_id': meta['functionality_name']
  }
)
output.write_h5ad(par['output'], compression='gzip')
Contents of script.R
library(anndata)

## VIASH START
par <- list(
  input_train = "resources_test/label_projection/pancreas/train.h5ad",
  input_test = "resources_test/label_projection/pancreas/test.h5ad",
  input_solution = "resources_test/label_projection/pancreas/solution.h5ad",
  output = "output.h5ad"
)
meta <- list(
  functionality_name = "my_method_r"
)
## VIASH END

cat("Reading input files\n")
input_train <- anndata::read_h5ad(par[["input_train"]])
input_test <- anndata::read_h5ad(par[["input_test"]])
input_solution <- anndata::read_h5ad(par[["input_solution"]])

cat("Preprocess data\n")
# ... preprocessing ...

cat("Train model\n")
# ... train model ...

cat("Generate predictions\n")
# ... generate predictions ...

cat("Write output AnnData to file\n")
output <- anndata::AnnData(
  obs = list(
    label_pred = obs_label_pred
  ),
  uns = list(
    dataset_id = input_train$uns[["dataset_id"]],
    normalization_id = input_train$uns[["normalization_id"]],
    method_id = meta[["functionality_name"]]
  )
)
output$write_h5ad(par[["output"]], compression = "gzip")

Required sections

Imports and libraries

In the top section of the script you can define which packages/libraries the method needs. If you add a new or different package add the dependency to config.vsh.yaml in the setup field (see above).

Argument block

The Viash code block is designed to facilitate prototyping, by enabling you to execute directly by running python script.py (or Rscript script.R for R users). Note that anything between “VIASH START” and “VIASH END” will be removed and replaced with a CLI argument parser when the components are being built by Viash.

Here, the par dictionary contains all the arguments defined in the config.vsh.yaml file (including those from the defined __merge__ file). When adding a argument in the par dict also add it to the config.vsh.yaml in the arguments section.

Read input data

This section reads any input AnnData files passed to the component.

Generate results

This is the most important section of your script, as it defines the core functionality provided by the component. It processes the input data to create results for the particular task at hand.

Write output data to file

The output stored in a AnnData object and then written to an .h5ad file. The format is specified by the API file specified in the __merge__ field in the config file.

Step 5: Try component

Your component’s API file contains the necessary unit tests to check whether your component works and the output is in the correct format.

You can test your component by using the following command:

viash test src/label_projection/control_methods/my_method_py/config.vsh.yaml
Output
Running tests in temporary directory: '/tmp/viash_test_majority_vote6584456763614973106'
====================================================================
+/tmp/viash_test_majority_vote6584456763614973106/build_executable/majority_vote ---verbosity 6 ---setup cachedbuild
[notice] Building container 'ghcr.io/openproblems-bio/label_projection/control_methods/majority_vote:test' with Dockerfile
[info] Running 'docker build -t ghcr.io/openproblems-bio/label_projection/control_methods/majority_vote:test /tmp/viash_test_majority_vote6584456763614973106/build_executable -f /tmp/viash_test_majority_vote6584456763614973106/build_executable/tmp/dockerbuild-majority_vote-9ToMtz/Dockerfile'
Sending build context to Docker daemon  41.98kB

Step 1/7 : FROM python:3.10
 ---> fc98d03e6037
Step 2/7 : RUN pip install --upgrade pip &&   pip install --upgrade --no-cache-dir "anndata~=0.8.0" "pyyaml"
 ---> Running in 340965c0af9d
Requirement already satisfied: pip in /usr/local/lib/python3.10/site-packages (23.0.1)
Collecting pip
  Downloading pip-23.1.2-py3-none-any.whl (2.1 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.1/2.1 MB 25.3 MB/s eta 0:00:00
Installing collected packages: pip
  Attempting uninstall: pip
    Found existing installation: pip 23.0.1
    Uninstalling pip-23.0.1:
      Successfully uninstalled pip-23.0.1
Successfully installed pip-23.1.2
WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv
Collecting anndata~=0.8.0
  Downloading anndata-0.8.0-py3-none-any.whl (96 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 96.1/96.1 kB 5.1 MB/s eta 0:00:00
Collecting pyyaml
  Downloading PyYAML-6.0-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl (682 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 682.2/682.2 kB 23.8 MB/s eta 0:00:00
Collecting pandas>=1.1.1 (from anndata~=0.8.0)
  Downloading pandas-2.0.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (12.3 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 12.3/12.3 MB 199.1 MB/s eta 0:00:00
Collecting numpy>=1.16.5 (from anndata~=0.8.0)
  Downloading numpy-1.24.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (17.3 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 17.3/17.3 MB 227.1 MB/s eta 0:00:00
Collecting scipy>1.4 (from anndata~=0.8.0)
  Downloading scipy-1.10.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (34.4 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 34.4/34.4 MB 228.2 MB/s eta 0:00:00
Collecting h5py>=3 (from anndata~=0.8.0)
  Downloading h5py-3.8.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (4.6 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 4.6/4.6 MB 208.5 MB/s eta 0:00:00
Collecting natsort (from anndata~=0.8.0)
  Downloading natsort-8.3.1-py3-none-any.whl (38 kB)
Collecting packaging>=20 (from anndata~=0.8.0)
  Downloading packaging-23.1-py3-none-any.whl (48 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 48.9/48.9 kB 189.3 MB/s eta 0:00:00
Collecting python-dateutil>=2.8.2 (from pandas>=1.1.1->anndata~=0.8.0)
  Downloading python_dateutil-2.8.2-py2.py3-none-any.whl (247 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 247.7/247.7 kB 291.5 MB/s eta 0:00:00
Collecting pytz>=2020.1 (from pandas>=1.1.1->anndata~=0.8.0)
  Downloading pytz-2023.3-py2.py3-none-any.whl (502 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 502.3/502.3 kB 309.6 MB/s eta 0:00:00
Collecting tzdata>=2022.1 (from pandas>=1.1.1->anndata~=0.8.0)
  Downloading tzdata-2023.3-py2.py3-none-any.whl (341 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 341.8/341.8 kB 305.2 MB/s eta 0:00:00
Collecting six>=1.5 (from python-dateutil>=2.8.2->pandas>=1.1.1->anndata~=0.8.0)
  Downloading six-1.16.0-py2.py3-none-any.whl (11 kB)
Installing collected packages: pytz, tzdata, six, pyyaml, packaging, numpy, natsort, scipy, python-dateutil, h5py, pandas, anndata
Successfully installed anndata-0.8.0 h5py-3.8.0 natsort-8.3.1 numpy-1.24.3 packaging-23.1 pandas-2.0.1 python-dateutil-2.8.2 pytz-2023.3 pyyaml-6.0 scipy-1.10.1 six-1.16.0 tzdata-2023.3
WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv
Removing intermediate container 340965c0af9d
 ---> 4e5a35173a10
Step 3/7 : LABEL org.opencontainers.image.description="Companion container for running component label_projection/control_methods majority_vote"
 ---> Running in f64dd24cb407
Removing intermediate container f64dd24cb407
 ---> ab2d077ee88e
Step 4/7 : LABEL org.opencontainers.image.created="2023-05-06T00:07:29Z"
 ---> Running in d14053f986a0
Removing intermediate container d14053f986a0
 ---> 5386e2356d45
Step 5/7 : LABEL org.opencontainers.image.source="https://github.com/openproblems-bio/openproblems-v2"
 ---> Running in 8dd70ff5fa24
Removing intermediate container 8dd70ff5fa24
 ---> 37f45798010e
Step 6/7 : LABEL org.opencontainers.image.revision="9438b8ad0cdd9cd2ed3ba6a01d0b4f075c059d64"
 ---> Running in 44c8eb828f1f
Removing intermediate container 44c8eb828f1f
 ---> ad7a3108d0b8
Step 7/7 : LABEL org.opencontainers.image.version="test"
 ---> Running in 9947336b8897
Removing intermediate container 9947336b8897
 ---> 223d4aff497b
Successfully built 223d4aff497b
Successfully tagged ghcr.io/openproblems-bio/label_projection/control_methods/majority_vote:test
====================================================================
+/tmp/viash_test_majority_vote6584456763614973106/test_check_method_config/test_executable
Load config data
Check general fields
Check info fields
All checks succeeded!
====================================================================
+/tmp/viash_test_majority_vote6584456763614973106/test_run_and_check_adata/test_executable
>> Checking whether input files exist
>> Running script as test
Load data
Compute majority vote
Create prediction object
Write output to file
>> Checking whether output file exists
>> Reading h5ad files and checking formats
Reading and checking input_train
  AnnData object with n_obs × n_vars = 346 × 419
    obs: 'label', 'batch'
    var: 'hvg', 'hvg_score'
    uns: 'dataset_id', 'normalization_id'
    obsm: 'X_pca'
    layers: 'counts', 'normalized'
Reading and checking input_test
  AnnData object with n_obs × n_vars = 154 × 419
    obs: 'batch'
    var: 'hvg', 'hvg_score'
    uns: 'dataset_id', 'normalization_id'
    obsm: 'X_pca'
    layers: 'counts', 'normalized'
Reading and checking input_solution
  AnnData object with n_obs × n_vars = 154 × 419
    obs: 'label', 'batch'
    var: 'hvg', 'hvg_score'
    uns: 'dataset_id', 'normalization_id'
    obsm: 'X_pca'
    layers: 'counts', 'normalized'
Reading and checking output
  AnnData object with n_obs × n_vars = 154 × 419
    obs: 'batch', 'label_pred'
    var: 'hvg', 'hvg_score'
    uns: 'dataset_id', 'method_id', 'normalization_id'
    obsm: 'X_pca'
    layers: 'counts', 'normalized'
All checks succeeded!
====================================================================
SUCCESS! All 2 out of 2 test scripts succeeded!
Cleaning up temporary directory

Visit “Run tests” for more information on running unit tests and how to interpret common error messages.

You can also run your component on local files using the viash run command. For example:

viash run src/label_projection/control_methods/my_method_py/config.vsh.yaml -- \
  --input_train resources_test/label_projection/pancreas/train.h5ad \
  --input_test resources_test/label_projection/pancreas/test.h5ad \
  --output output.h5ad

Next steps

If your component works, please create a pull request.