Quality Control Rules

Version 1:

The following is a list of metrics to measure the quality level of data in FaceBase. A program is run nightly that compares existing data to these metrics and flags discrepancies.

1. Dataset-level Requirements

1.1 Dataset-level required fields:

• Accession (in place)
• Title (in place)
• Description
• Project (in place)
• Requires DOI
• DOI (generated)
• Released (in place)

1.2 Dataset-level required high-level metadata tags:

• Organism
• Experiment Type
• Stage (depending on Species and specimen)
• Anatomy
• Gene

1.3 Dataset-level organizational requirements:

• At least one Experiment AND one Biosample AND one Replicate associating the Experiment and Biosample

2. Experiment-level requirements

• At least one associated Replicate
• Experiment Type (in place)
• Local Identifier
• Strandedness (depending on Experiment Type value)
• Target of Assay (depending on Experiment Type=ChIP-seq)
• Control Assay (depending on Target of Assay value)
• Molecule Type (depending on Experiment Type value)
• Histone Modification (depending on Target of Assay value)
• Chromatin Modifier (depending on Experiment Type value)
• Transcription Factor (depending on Experiment Type value)
• RNA-Seq Selection (depending on Experiment Type value)

3. Biosample-level requirements:

• At least one associated Replicate
• Local Identifier
• Species (in place)
• Specimen (in place)
• Stage (depending on Species and specimen)
• Anatomy
• Collection Date
• Cell Source (depending on Specimen value)
• Cell Characterization (depending on Specimen value)

4. Replicate-level requirements:

• Experiment (in place)
• Biosample (in place)
• Bioreplicate Number (in place)
• Technical Replicate Number (in place)
• Sequencing files (depending on associated Experiment Type)
• Image or Mesh file (depending on associated Experiment Type)