| layout | title | parent | nav_order |
|---|---|---|---|
default |
Organize Your Data Upload |
Step 1 - Upload Your Data |
3 |
Centers may use a hierarchical file structure within their cloud storage location. However, each top-level folder (also referred to as the root folder) and all of its subfolders must contain data of the same type (see details below). This will make annotating metadata with the data curator app easier for users since one metadata template will be submitted for each folder.
Top level folders correspond to the assay and level metadata you will be submitting. For example, we have scRNAseq levels 1, 2, 3, 4 and in each RFC the corresponding attribute table is converted into a schema and generates a google spreadsheet for the Data Curator App. See below for examples.
The DCC will create empty, common top-level folders for you to get started.
.
├── clinical_tier_1_diagnosis
├── clinical_tier_1_demographics
├── biospecimen_tier_1
│ ├── file1.txt
│ └── file2.txt
└── single_cell_RNAseq_level_1
├── fileA.fq
└── fileB.fq
Your files should be reasonably descriptive in stating the assay type and level and be consistently prefixed with the assay type.
.
├── biospecimen_tier_1_experiment_1
├── biospecimen_tier_1_experiment_2
├── single_cell_level_1_batch_1
└── single_cell_level_1_batch_2
Note that this could create many top level folders, please see the section below about subfolders.
In the hierarchical case, you would fill in one manifest including all files in experiment/batches; in the flattened case, you would fill in one manifest for each top level folder.
Subfolders must be of the same data type and level as the root folder they are contained in. For example you can not put a biospecimen Tier 1 and clinical demographics Tier 1 subfolder within the same folder.
.
├── clinical_tier_1_diagnosis
├── clinical_tier_1_demographics
├── biospecimen_tier_1
│ ├── experiment_1
│ └── experiment_2
└── single_cell_level_1
├── batch_1
└── batch_2