Curating Metadata

First: Reach out to the Data management team ([e-mail]) to set up a brief meeting to overview metadata curation. Below is a summary of what he will discuss in the meeting.

We structure metadata as a relational database. A very brief overview:

• We architecture sample metadata through provenance: starting at generation of the sample (PAT: Patient, PAV: Patient Visit, and TIS: Tissue) along with metadata surrounding the data file (DNA: DNA Library, D.SEQ: Sequencing Data File, A.GEX: Gene Expression Analysis File).

• Each sample type (PAT, PAV, TIS, DNA, D.SEQ, A.GEX) will have different sets of metadata included. Different labs will be responsible for different sample types as well.

For example: The Mayo Clinic will handle the PAT, PAV, and TIS metadata: IE everything about the Patients (PAT), what happened to them / when they came in for visits (PAV) and then what tissues were extracted (TIS)

• Since our system is not HIPPA/DSL4 compliant, these metadata attributes for PAT/PAV will be binary (0 if it doesnt exist, 1 if it does exist) with a contact email to someone at the mayo clinic to obtain specific values

• Then, the lab responsible for the sequencing (IE, the lab that got the Tissues from the mayo clinic), will handle the DNA/D.SEQ/A.GEX sample types.

An in depth description of AssaySheets exists here. You only need to fill out the Samples Page. The instructons page includes references/notes on what each column contains.

And the AssaySheet Templates Exist Below:

29KB

PAT-PAV-TIS_AssaySheet_CSBC.xlsx

PAT/PAV/TIS Combined Assay Sheet. Reach out to the fairdata team with any questions

21KB

DNA-D.SEQ_AssaySheet_CSBC.xlsx

DNA/D.SEQ Combined Assay sheet. Reach out to the fairdata team with any questions