Glossary | CAP

Getting Started

Creating a Profile Browsing Cells and Datasets Downloading Datasets Exploring Cell Labels

Publishing Your Work

Creating Draft Publications and Uploading Datasets Anndata Requirements for Upload Entering Cell Annotation Metadata Collaborating on the Workspace Viewing Cell Relationships Publishing Workspaces

Molecular Data

Molecular Data Page Overview Creating New Cell Annotation Set Running Differential Expression

Background

Glossary

Dataset

The contents of a single file, such as AnnData, which includes cell annotations, metadata, embeddings and molecular matrices. Datasets can be either single or integrated:

Single dataset: The file includes a single dataset whereby “dataset” means a single library or a set of technical replicates that can be combined without having to worry about the associated batch effect.
Integrated dataset: The file includes an integrated collection of datasets.

User Roles

CAP is open to the public and anyone can search and view publications without registering to be a user. In addition to non-registered users, there are several specific user types:

Owner: a user who created a publication draft and has the ability to grant access to the workspace to other users and make changes to the datasets.
Admin: a user with the same level of access to a workspace as the owner, except they cannot create or remove other admins.
Editor: a user with the ability to make changes to datasets in the publication draft page.
Viewer: a user with view-only access.

Publication Draft

Location where users can organize collections of datasets for versioned publications. Users can collaboratively edit annotations and associated metadata here.

Publication

Location where dataset(s) and associated information such as raw data and journal links and author contact information are viewable but not editable. Publication pages are versioned so previous versions of the publication can be accessed by users.

Molecular Data

Location where users can view embeddings by cluster or gene expression values, interact with a gene expression heatmap, run differential expression analysis and create and edit cell labels.

Cell Relationships

Shows the relationship between cell labels within a single dataset. This information is displayed only if a dataset has more than one cell labelset and does not show relationships for cell labels across datasets. The relationships are not hierarchical but represent the relationship between the cell labelsets; see the example below.

Label Metadata

Synonyms: Terms that have the same, or nearly the same, meaning as the selected cell type. For example, “T cell” and “T lymphocyte”.
Categories: Classes of which the cell type is a subclass. For example, “T lymphocyte” could be a category of “CD8+ T cell”. Although categories will be broader than the selected cell type, it is recommended to select the most specific category that fits your cell type.

Evidence for Annotations

Users can provide evidence for the annotation for each cell label. The evidence can consist of marker genes or free-text provided by the user.

Marker genes: Users can provide a list of up to 20 gene names which were used as evidence to assign this cell annotation. These were the genes explicitly used as evidence supporting the assignment of cell annotations for these cells during the data analysis performed, e.g. highly variable genes, differential expression, etc. NOTE: This field is a fundamentally different concept than "canonical marker genes".
Canonical marker genes: These are the gene names considered by the researcher to be canonical markers for the biological entity used in this cell annotation. While this list may be the same as the marker genes used to annotate cells ("marker genes"), this field "canonical marker genes" details what the cell annotation terms mean, not the analysis performed to assign cell annotations.
Rationale: A free-text explanation where users can provide justification or evidence for their cell annotations. All references cited should be listed using DOIs under rationale_dois.
Rationale DOI: A list of valid publication DOIs cited by the author to support or provide justification or evidence for the cell annotation.

For example: a Beta cell in the human pancreas may be indicated with the marker genes MAFA, INS, and PFKFB2. The supporting Rationale could be provided as follows:

Classical beta cell markers such as, INS, MAFA and PDX1, were detected in addition to PFKFB2 a gene linked to insulin regulation which had not been previously reported in human beta cells. See publication for more detail: https://doi.org/10.1016/j.cels.2016.09.002.