Annotate Samples¶
Well-annotated data is significantly more valuable than data that lacks structure or context. Investing time in proper annotation ensures that your data can be readily located, interpreted, and reused in the future. High-quality annotations also streamline tasks such as leveraging color-by functionality in visualizations or preparing submissions for public databases.
View and add annotations to samples¶
View annotations
Each sample can have metadata attached to it. This metadata is located in the View Sample Details page, which you can bring up by clicking on a sample name either:
in the results table from Samples tab on the Search for Data page or
in the Samples card from within the View Collection Details page.
Metadata for a sample is displayed in the Sample Annotation Fields card. Once populated, the card is organized into sub-sections based on metadata category. These categories include general sample properties (e.g., species, cell line, annotator, description, QC information) as well as specific subject and disease information, making them suitable for annotating both preclinical and clinical samples.
You can explore the available metadata fields in the Annotation Fields Library section of the Collection Details page. Drag and drop individual or groups of fields into the Current Selection of Annotation Fields. The fields selected will be associated with the Collection and included in the “Annotation template” Excel file when downloaded from the Actions card -> Annotation templates view.
Add annotations¶
Structured metadata
Sample annotation is supported by the Python SDK (ReSDK) library. To add metadata to all or a subset of samples, refer to the Annotate Samples section of the ReSDK tutorial. The web application does not currently support adding or modifying sample metadata.
Unstructured metadata
Metadata information that cannot be readily aligned with the available annotation fields can be uploaded in tab-delimited format to the working Collection using the Actions’s card tool Metadata table (one-to-one). The uploaded file should contain the sample metadata given in columns, and is matched to sample objects by rows. To match the sample objects by their corresponding rows, the table must contain a column with one of the following headers: “Sample ID”, “Sample name” or “Sample slug”.
Metadata uploaded using the Metadata table (one-to-one) tool is combined with the structured metadata information and is presented in a unified table when accessed using the ReSDK Python library RNATables class.