Contributing

We welcome contributions to the Cell Painting Gallery.

Contributing to Documentation

If you would like to contribute to our documentation, please make a Pull Request. We particularly welcome contributions to our list of publications using data from the Cell Painting Gallery and to workflows accessing the Cell Painting Gallery.

To ask a question that is not covered by our documentation, you are also welcome to create an Issue in the Cell Painting Gallery repository. Please note that the Cell Painting Gallery is not a place to ask dataset-specific questions. Instead, please direct such questions to the respective dataset repository linked in Complete Datasets or if no dataset repository, to the authors of any associated publication.

Contributing Data to the Gallery

Contributions can be in the form of complete datasets (can be images and metadata or images and a complete workspace) or additions to extant datasets (e.g. segmentations or deep-learning generated profiles). Please contact @erinweisbart or @shntnu to initiate discussion of a data contribution.

For new datasets, please include the following details in your contact:

1. assay used (standard Cell Painting or describe the variation. If you would like to contribute data from a derivative assay it must be useable for morphological profiling in that it stains/labels multiple cellular compartments/organelles.)

2. approximate data size

3. components you wish to contribute (all are described in data structure) (major components: images, analysis, backend, load_data_csv, profiles. optional components: pipelines, qc, etc.). Note that images, metadata and load_data_csv are required.

4. institutional identifier to use for data (e.g. broad, anonymous)

5. suggested top level project tag. We will assign the cpg number but are happy to take suggestions for the tag appended to the cpg number. The tag is typically a 1 word summary of the project (e.g. cpg0011-lipocyteprofiler, cpg0016-jump, cpg0022-cmqtl) and sometimes also includes the last name of the first author (e.g. cpg0010-caie-drugresponse, cpg0028-kelley-resistance, cpg0031-caicedo-cmvip)

For existing datasets, please the following details include in your contact:

1. top-level project tag that your data corresponds to (e.g. cpg0016-jump)

2. approximate data size

3. components you wish to contribute

After approval, we will assign you a project identifier and create a new Github Discussion using our discussion template to provide next steps and track data deposition.

Preparing for data deposition

In preparation for transferring data, please perform all of the following steps:

Remove special characters in folder names

To the maximum extent possible, please avoid the following in your folder names (delete them if present):

• Whitespaces

• Special characters other than _ and -

Prepare project-specific naming

Reference data structure for comprehensive information on folder structure and naming. Your data must strictly comply with the data structure we have laid out. Additionally it must include all, unblinded metadata - this means that the minimum dataset contribution is images, metadata, and load_data_csv.

Validate your data

Currently, validation is performed by Erin/Shantanu after data is uploaded to staging. In the future we plan on having contributors first validate their data before any upload.

Create a pull-request

Any data contributions to Cell Painting Gallery must be accompanied by a pull request to the Cell Painting Gallery repository with updates to the README to add your dataset to Available datasets. If your dataset is associated with a publication, please also edit Publications.

Acknowledging CPG maintainers’ role in deposition

Most datasets require a handful of hours of time for a maintainer of the CPG to assist in data deposition, performing tasks such as handling upload credentials, transferring data from staging to production, answering questions, confirming data structure and metadata completeness, and generating harmonized metadata. We are happy to provide these hours as a service to the scientific community. We ask that you acknowledge our volunteered time in your paper’s acknowledgements.

In the rare occasion that you don’t have someone in your team that has the skills or time to prepare your data appropriately for deposition, and your data requires more substantial cleaning/reorganizing/preparation for deposition, we are available to help but ask that we be compensated. This can be through hourly fee-for-service and/or by making the person on our team who spends the time handling the data an author on your paper describing the data. This is consistent with the Contributor Role Taxonomy (CRediT) that acknowledges “data curation” as one of the diverse contributions behind published research.