Frequently Asked Questions

You can search for participant data in several ways:

1. Global Search: Use the Search Center (accessible from the top navigation) and enter the CMMI ID (e.g., CMMI-001). This will show all data types associated with that participant.

2. Quick Lookup: In the "Unified Data" dropdown menu, there's a "Look Up Participant" field where you can enter a CMMI ID and jump directly to their profile.

3. Data Type Specific: Navigate to any data type page (CBC, Metabolomics, etc.) and use the participant filter in the sidebar to view only that participant's results.

All participant data is linked by the CMMI ID, making it easy to track across different data types.

The CMMI-DCC database contains multiple types of clinical and omics data:

Clinical Data: - CBC (Complete Blood Count) - Blood test results - HLQ (Health and Lifestyle Questionnaire) - Participant health information - Interpretive Results - Clinical interpretations

Omics Data: - Metabolomics - Small molecule measurements from biofluids - Proteomics - Protein abundance data - Metagenomics - Microbial genomic sequencing data - Microbiome - Gut microbiome composition and diversity

You can browse each data type from the "Lab Data" menu or search across all types using the Search Center.

To upload data to CMMI-DCC:

1. Navigate to Upload Data: Click on "Upload Data" in the top navigation menu

2. Select Data Type: Choose the appropriate data type (CBC, Metabolomics, etc.)

3. Download Template: Click "Download Example" to get the correct CSV template for your data type

4. Prepare Your Data: Fill in the template with your data, ensuring all required columns are complete

5. Upload: Click "Choose File" and select your completed CSV file

6. Preview & Confirm: Review the preview of your data to ensure it was parsed correctly, then click "Confirm" to submit

The system will validate your data and notify you of any errors. For omics data, you may need to provide additional metadata after the initial upload.

If your data upload fails validation, follow these steps:

1. Review Error Messages: The system will display specific error messages indicating which rows or columns have issues

2. Common Issues:

   • Missing required columns (CMMI ID, measurement values, etc.)
   • Invalid CMMI IDs (participant not found in database)
   • Incorrect date formats (use YYYY-MM-DD)
   • Duplicate entries
   • Values outside expected ranges

3. Fix Your Data: Download your original file, correct the errors, and re-upload

4. Use the Template: Always start with the official template from the "Download Example" button to ensure correct formatting

5. Contact Support: If you continue to have issues, contact the CMMI-DCC team with your error messages and data file

To run machine learning analyses in CMMI-DCC:

1. Navigate to ML Pipelines: Click "Analysis" then "ML Pipelines" in the top navigation

2. Create New Pipeline: Click "New ML Pipeline"

3. Select Data:

   • Choose your data types (Metabolomics, Proteomics, CBC, etc.)
   • Select specific studies or cohorts
   • Choose which features (columns) to include

4. Configure Analysis:

   • Select task type (Classification or Regression)
   • Choose target variable (what you want to predict)
   • Select algorithm (Random Forest, XGBoost, etc.)
   • Configure preprocessing options (scaling, feature selection, outlier detection)

5. Submit Job: Review your configuration and click "Submit". The analysis will run in the background

6. View Results: Once complete, you can view performance metrics, feature importance, and download results

Results include model performance metrics, visualizations, and the ability to export predictions.

CMMI-DCC provides several options for exporting data:

Individual Data Types: - Navigate to any data type page (CBC, Metabolomics, etc.) - Use filters to select the data you want - Click the "Download" or "Export" button - Choose CSV format for compatibility with Excel, R, Python, etc.

Combined Export: - Go to "Lab Data" then "Export" - Select multiple data types to combine - Choose participants, studies, or cohorts - Download as a single integrated CSV file

Search Results: - Use the Search Center to find specific data - Click "Export" to download search results

Pre-generated Downloads: - Visit "Lab Data" then "Downloads" for pre-packaged datasets - These include complete datasets for common use cases

All exports maintain data integrity and include metadata for proper interpretation.

"Out of Range" indicates that a CBC test result falls outside the normal reference range for healthy individuals.

Reference Ranges: - Each CBC component (HGB, HCT, WBC, RBC, PLT) has established normal ranges - These ranges vary by age, sex, and sometimes ethnicity - Results are flagged as "High" (above range) or "Low" (below range)

What It Means: - Not Always Abnormal: Some healthy individuals naturally have values outside typical ranges - Context Matters: A single out-of-range result may not indicate disease - Trends Important: Changes over time are often more meaningful than single values

In CMMI-DCC: - Out-of-range results are highlighted in red in data tables - You can filter to show only out-of-range results - The "Status" column shows whether results are High, Low, or Normal

For detailed information about reference ranges, see the Help Center page on "Reference Range Status".