Standard Data Preprocessing Pipeline for ABCD Tabulated Data

Kunru Song 2024-07-09

• Introduction
• Citation and Suggestions
• Change Logs

Introduction

SDPP-ABCD-TabDat (Standard Data Preprocessing Pipeline for the Adolescent Brain Cognitive Development study Tabulated Data) is an open-source tool aiming to provide an user-friendly data preprocessing pipeline for tabulated data from the ABCD study. This tool mainly depends on R. I hope this open-source tool could improve the accuracy of the secondary data analysis with the ABCD study and also provide an easy way to reproduce results and findings related to the ABCD data.

Author: Kunru Song, PhD Candidate at BNU, supervised by Dr.Jin-tao Zhang.

Kunru’s Researchgate Profile

Kunru’s Google Scholar Profile

Dr. Jin-tao Zhang’s Lab (Researchgate)

First Created Date: 2022.06

Data Version: ABCD 5.1 Data Release (tabulated data)

ABCD Links

• ABCD 5.1 Tabulated Release Data
• ABCD 5.1 Data Dictionary
• ABCD 5.1 Release Notes & Update Logs

Citation and Suggestions

1. If you would like to use any codes in this open-source tool (licensed by GNU GPL v3), please kindly follow terms and conditions in GNU GPL v3 or later and cite our published article used this tool (below is a APA-format Citation):

   Song, K., Zhang, J. L., Zhou, N., Fu, Y., Zou, B., Xu, L. X., Wang, Z., Li, X., Zhao, Y., Potenza, M., Fang, X., & Zhang, J. T. (2023). Youth Screen Media Activity Patterns and Associations With Behavioral Developmental Measures and Resting-state Brain Functional Connectivity. Journal of the American Academy of Child and Adolescent Psychiatry, 62(9), 1051–1063. https://doi.org/10.1016/j.jaac.2023.02.014

2. Please refer to SDPP-ABCD-TabDat User Manual when you have any questions about using this pipeline. User Manual was wrote by R Markdown syntax and output as a html file for convenient to be read. Users could find the manual in this repository: SDPP_ABCD_TabDat_UserManual.html.

   In addition, find here for online html file preview.(Supported by htmlpreview.github.io)

3. Any bugs found in this pipeline can be reported in Github Issues of this repository.

4. If you would like to participate the collaboration in developing this pipeline, please kindly create a pull requests in relevant page of this repository

Change Logs

• 2024.07.09

  • Update README and User Manual.
  • Create a new User Tutorial, named as SDPP_UserTutorial_Example.Rmd in root directory.
  • Completed Step 15, further test are needed.
  • Add more code annotations to SDPP_ParaSet.R and SDPP_DebugTool.R

• 2023.12.10

  • Completed SDPP_Batch, implement click-to-run interface.
  • Separate user-defined necessary arguments to a new SDPP_UserDefined_Input.R script.
  • TO DO: 1) check the full pipeline and fix bugs; 2) write a new usage notes section in user mannual; 3) write some basic code examples in README.

• 2023.12.9

  • Completed Step 13 & 14
  • Drafting the first version of SDPP_Batch
  • refractoring some basic functions in SDPP_subfunctions, add more high-level warps for convenient callback.
  • Reduce the complexity of SDPP_Paraset.
  • Add sMRI variable names mapping table in /.github folder.
  • TO DO: 1) more flexible SDPP_Batch; 2) integrate SDPP_Batch with SDPP_Paraset; 3) check the efficacy of SDPP-ABCD-TabDat on ABCD data release V5.1

• 2023.12.7

  • Completed Step 12.
  • Start Step 13 for sMRI tabulated data preprcossing, including Destrieux (aparc.a2009s) and subcortical auto-segmentation (aseg) atlas.
  • Add new ggseg look up table in /.github folder.

• 2023.12.6

  • Add compatability of nomenclature for ggseg label based on ABCD ROI names
  • Add new MRI ROI Look Up Table (LUT) for Desikan-Killiany Atlas in /.github folder.

• 2023.11.28

  • Start Step 13 (extract sMRI-related variables).
  • Add an EXCEL template: SDPP-MRI-Look-up-table (.github/SDPP_MRI_ROI_LookUpTable.xlsx)
  • Update README & User Manual for ABCD MRI tabulated data.

• 2023.11.27

  • Update README and User Manual for Step 11.
  • Fixed some bugs in SDPP_ParaSet.R and SDPP_Concatenate.R
  • Start & completed Step 12.
  • Add some paragraphs to clarify the rationale of mutliple imputaion startergy used in this pipeline.
  • TO DO: 1) re-construct some basic warp functions in SDPP_subfunction.R; 2) Check the robustness of Step_3; 3) Integrate neuroCombat into this pipeline.

• 2023.11.25

  • Completed Step 11.
  • Update User Manual for Step 11.

• 2023.11.24

  • Completed Step 10.
  • Update User Manual for Step 10.
  • Update README
  • Start Step 11.

• 2023.11.23

  • Update README and User Manual for Step 10.
  • Change license to GNU GPL v3.

• 2023.11.22

  • find some bugs and start codes re-constructing to improve readability of all codes in SDPP-ABCD-TabDat.
  • Add new features: filter columns by manually modified ABCD Data Dictionary file.
  • Update Step 10.
  • Update User Manual for Step 9.

• 2023.9.21

  • Update Step 9.

• 2023.8.7

  • Start Step 11.

• 2023.8.5

  • Completed Step 9.
  • Update User Manual for Non-NIH toolbox Neurocognition Tasks data re-coding.
  • Start Step 10.

• 2023.7.31

  • Update ReadMe and User Manual, auto-load EXCEL file to display tables
  • Start Step 9

• 2023.7.18

  • Completed Step 8.
  • Add more information about Neurocognition and Mental Health to User Manual.
  • Update the Interactive Data Dictionary Excel File.

• 2023.7.17

  • Completed Step 6, add a new subfunction to re-code nm and nt values in ABCD Summary Data File.
  • Updated coding scheme table, more columns (e.g. coding details about factor levels collapse will be added in future version).
  • Completed Step 7.

• 2023.7.10

  • Updated README and User Manual.

• 2023.7.9

  • Completed Step 5, added some subfunctions to perform knn-imputation for missing values in SMA-related behavioral measures.
  • Added coding scheme table.
  • Fixed some bugs in calling MATLAB Bioinformatics Toolbox.

• 2023.07.07

  • Completed Step 4 and write Usage Notes in README, which provide a comprehensive overview for using SDPP-ABCD-TabDat.
  • Update some codes to generate more elegant code running logs, which would be saved into ‘Res_1_Logs’ folders under your project directory. SDPP_ParaSet will generate a log with postfix 0. The numeric postfix is correspond to the step number.
  • Separate SDPP-ABCD-TabDat User Manual from README.

• 2023.07.06

  • Modifed Step 4 for ABCD V5.0. Added a new section in README to illstruate some details in Step 4.
  • Screen Use Time now will be re-coded into 7-level oridinal variable, data type is numeric.

• 2023.07.05

  • Modifed Step 3 for ABCD V5.0.
  • Created a new script ‘SDPP_ParaSet.R’, which is the all in one parameters setting script for SDPP-ABCD-TabDat. ‘SDPP_ParaSet.R’ should be executed before any steps of SDPP-ABCD-TabDat to source all common parameters and libary all common packages. Log for ‘SDPP_ParaSet.R’ will be save into Log_SDPP-ABCD-TabDat_0.txt

• 2023.07.03

  • Errors in measures of height and weight that noted in ‘ABCD 5.0 Changes and Known Issues’ (Page 9) were processed in Step 1 R script. please see attached file ‘ABCD5.0_BMI_NotesTable.xlsx’, where Sheet ‘NotesTable’ was directly copied from ‘ABCD 5.0 Changes and Known Issues.doc’ (Page 9 & 10), sheet ‘ManualCorrectiveTable’ was manually generated by Kunru Song (2023.07.03) to correct the value with each subject in it. The second sheet could be used to replace (‘correct’) the measured values in /physical-health/ph_y_anthro.csv (that is what Step 1 would do). Please download and place it within the folder where SDPP-ABCD-TabDat scripts were storage.
  • At 4-year follow-up, the EHIS was measured again. To keep consistent, youth’s handedness indicated by EHIS was determined by the baseline measure, which would be seen as a time-invariant variable.(Section 8 in Step 1)
  • Add function to automatedly save logs.
  • Merge Intersex-Male and Male into Male. Because the number of intersex-male subjects is too small.

• 2023.07.01

  • Add some codes to re-code rel_group_id into a new numeric variable, which could describe the order of youths from same family. A new variable named GroupID will be created within execution of Step1, range from 0 to 3. For single-child family, the value would be set to 0. For multi-child family (sibling, twin or triplet), the corresponding value indicates the order of youths in their family.
  • Old Variables with prefix ‘GeneInfo’ now were replaced by ‘GI’.
  • The top five Principal component of genetic ancestry would be preserved in the preprocess data frame. I did this as a conventional choice such as aCompcor and tCompcor, where we usually extract the top five principal components. It should be noted that such conventional choice did not have particular reason. Usually, the top five PCs are enough to capture the variations in raw data. Because I did not find proper justification for ABCD gentic ancestry PCs, such choice may be an optional and practical process for the genetic ancestry data. Any comments are welcome to address this issue.

• 2023.06.29

  • Revising all codes for ABCD V5.0 Tabulated Data (NDA Collection ID: 2147)
  • In ABCD 5.0 tabulated data, ‘src_subject_id’ would be used as participant identifier.

• 2023.06.06

  • To improve compatability of CSV files across different environment (R, Python, MATLAB, LibreOffice, Excel), vaiable names (i.e., Column Names) have been modified. All ‘.’ were replaced by ’_’.

• 2023.06.05

  • Add additional parameter in write.csv with fileEncoding=‘UTF-8’. Now all CSV-files generated by SDPP-ABCD-TabDat will be formated in UTF-8 encoding type.

• 2023.06.03

  • To keep consistent with NDA platform, the variable ‘subjectkey’ will be used as Subject ID to merge data. Because ‘subjectkey’ is the ‘NDAR Global Unique Identifier (GUID) for research subject’. Anyone using ABCD data could easily apply GUID filter to create their own NDA Study, which is the requirements of ABCD DUC terms and condition.
  • Added more comments into R script. Added code blocks to make everything easy to read.

---