README.md

Complete analysis source code for:

Replicable Multivariate BWAS with moderate sample sizes

Tamas Spisak, Ulrike Bingel, Tor Wager, Nature, 2022

---

Project webpage:

spisakt.github.io/BWAS_comment

---

How to get the data?

• Option 1: contact the authors for credentials and get it with get_data.ipynb in one minute (this option is for collaborators only).
• Option 2: Get the data manually from ConnectomeDB. See hcp_data/readme.md for details.

---

How to run?

• Run the analyses with 1 click (in the cloud)

  • 
  • 

• Set up everything to run in 5 mins (locally on your computer)

  • install docker
  • run docker run -it -p 8080:8080 -p 8888:8888 tspisak/bwas:latest jupyter notebook
  • copy paste the last link in your browser.

---

What is what in the repository?

Main analysis notebooks:

• get_data.ipynb
• multivariate_BWAS_replicability_analysis_FC.ipynb

Extended analysis: replicability of 52 phenotypes:

• multivariate_BWAS_replicability_analysis_FC_extensive.ipynb

Plotting results:

• plot_results_FC.ipynb

Supplementary analysis notebooks:

• multivariate_BWAS_replicability_analysis_FC_hires.ipynb
• multivariate_BWAS_replicability_analysis_FC_motion.ipynb
• multivariate_BWAS_replicability_analysis_CT.ipynb
• plot_results_CT.ipynb
• sim_effect_size.ipynb

Data: HCP1200 release Data source: ConnectomeDB. See hcp_data/readme.md

---

How to cite?

T. Spisak, U. Bingel, T. Wager, Multivariate BWAS can be replicable with moderate sample sizes in some cases, accepted in Nature