Skip to content

Autism Brain Imaging Data Exchange II (ABIDE II)

From the ABIDE II webpage:

ABIDE II was established to further promote discovery science on the brain connectome in ASD. To date, ABIDE II has aggregated over 1000 additional datasets with greater phenotypic characterization, particularly in regard to measures of core ASD and associated symptoms. In addition, two collections include longitudinal samples of data collected from 38 individuals at two time points (1-4 year interval). To date, ABIDE II involves 19 sites - ten charter institutions and seven new members - overall donating 1114 datasets from 521 individuals with ASD and 593 controls (age range: 5-64 years). These data have been openly released to the scientific community on June 2016.

Steps already done to produce the preliminary data dictionary saved here

  1. Went to the ABIDE II webpage.
  2. Scrolled down to the ABIDE II Downloads >> Phenotypic data section
  3. Clicked the ABIDE II Phenotypic Data Legend link and downloaded the PDF directly to this ABIDE_II/ subfolder.
  4. Did not rename the PDF file from its originally downloaded name.
  5. Installed and used the Tabula software to pull the tables out of the PDF pages and save the output as a TSV.
  6. "Manually" edited, as necessary, to eliminate multi-line entries and to match each PDF table's intent.
  7. Renamed the filename from tabula-ABIDEII_Data_Legend.tsv to more simply ABIDEII_Data_Legend.tsv.
  8. Committed the final ABIDEII_Data_Legend.tsv TSV here to Git version control.

Steps to produce this study's data dictionary

  1. Install the required Python 3 library with the following line of code:

    shell python3 -m pip install --user pandas

  2. Run the following line of code within the ABIDE_II/ subfolder:

    shell python3 dictionary.py

Notes about this data dictionary

  1. The BIDS Description is mostly composed here of the comma-separated list of VARIABLE TYPE then MIN then MAX because the unique combination of these three describes the data in that field well.
  2. The entries for FIQ, VIQ, and PIQ were "manually" adjusted to have simplified Description and no more Levels.
  3. The entries for FIQ_TEST_TYPE, VIQ_TEST_TYPE, and PIQ_TEST_TYPE were "manually" adjusted to include better values in each level of the Levels and a shorted Description.
  4. You can review the exact manual fixes toward the bottom of the dictionary.py script.