BIG DATA for DISCOVERY SCIENCE
infographic
The Big Data for Discovery Science Center (BDDS) - comprised of leading experts in biomedical imaging, genetics, proteomics, and computer science - is taking an "-ome to home" approach toward streamlining big data management, aggregation, manipulation, integration, and the modeling of biological systems across spatial and temporal scales.
 
 

Probe ABIDE data set using the BDDS Data Dashboard

The Data Dashboard webapp allows us to navigate this very heterogeneous data (n=1,112, k>2,000), graphically explore deep multivariate relations, associations, and trends. In addition, the dashboard service may be used to generate testable research hypothesis about various ASD traits that can be subsequently evaluated using prospective datasets.

Protocol

  1. Download the BDDS ABIDE BagIt data object from here
  2. Extract the 4 data files (neuroimaging_metrics.csv, neuroimaging_metrics_metadata.csv, neuroimaging_metrics_qc.csv, phenotypes.csv), which contain the relevant (appropriately named) components of the ABIDE dataset.
  3. Start the Data Dashboard app (http://socr.umich.edu/HTML5/Dashboard/)
  4. Select Free—Anchored Data
  5. Enter a new anchor variable subject_ID:
  6. Import → Phenotypes data (Pheno)
  7. Select a variable of interest (e.g., Sex)
  8. Add → Chart (e.g., Pie Chart for Sex) from the Pheno data (Variable Sex)
  9. Add → Chart (e.g., Histogram Chart for BMI) from the Pheno data (Variable BMI) – note that BMI has lots of missing data!!!
  10. Add → Chart (e.g., Pie Chart of DX_Group or DSM_IV_TR) from the Pheno data
  11. Add → Chart (e.g., Scatter Plot: FIQ vs. ADOS_COMM)
  12. Manipulate some of the charts (e.g., select females, BMIs, etc.) to examine the change in the other features (e.g., FIQ score)
  13. Import → Neuroimaging Data (NI), be patient, it will take some time to load the data. After the dialogue box shows up, click “Import”
  14. Add → NI → Select a variable of interest (e.g., Volume or SurfArea, as an example of a feature) click “Continue”
  15. Add → NI → Select a variable of interest (e.g., Fold_Ind, as an example of a feature) click “Continue”
  16. Add → NI → Select a variable of interest (e.g., Primary_Lobe, as an example of a feature) click “Continue”
  17. Manipulate some of the charts to study the impact on the other plots.

References

  1. BagIt/BDBag Data File Format:
    1. Source code and Format: https://github.com/ini-bdds/bdbag and https://github.com/ini-bdds/minid
    2. Docs: http://bd2k.ini.usc.edu/assets/all-hands-meeting/minid_v0.1_Nov_2015.pdf
  2. Data Dashboard
    1. Source code: https://github.com/ini-bdds/Data-Dashboard
    2. Docs: http://wiki.socr.umich.edu/index.php/SOCR_Videos_Dashboard
    3. Videos: http://wiki.socr.umich.edu/index.php/SOCR_Videos_Dashboard#Video_Tutorial
    4. Service: http://socr.umich.edu/HTML5/Dashboard/