Skip to content

A. Installation

Jump to bottom
Javier Gonzalez-Castillo edited this page Jun 6, 2024 · 11 revisions

Required Packages

  • lemon_pipelines: originally available here. These pipelines had to be modified so they could run on our Biowulf environment. The modified version of the pipeline used in this work can be found at: https://github.com/nimh-sfim/lsd_pipelines_nih

  • mipav Version 5.4.2_amd64: you can download MIPAV from here. At the bottom of the page there is a section titled "Archived MIPAV Binary releases < 10.0.0". You can find the version we need there.

  • jist, jist-cruise, cbs tools for MIPAV: you can download it from here CBS Tools 3.0.8 + JIST-CRUISE 3.2 bundle

  • nipype: for this work, we used version 1.3.2 (to ensure functionality of the lemon pipelines)

  • fsl: version 5.0.10

  • freesurfer: version 5.3.0

  • ANTs: version 2.2.0

MIPAV & JIST Plugins

1. Install MIPAV

Although biowulf provides MIPAV as a module, for this project we need a much older version than those provided by the HPC team. For that reason we will have to install our own local copy. The version we need for this project in 5.4.2_amd64.

  1. You can find the installer for this version in: /data/SFIMJGC_Introspec/2023_fc_introspection/downloads/MIPAV/installMIPAV_5.4.2_amd64.bin.

  2. Open a terminal on an spersist node with graphical access (e.g., in no-machine or VNC)

  3. Run the installer

cd /data/SFIMJGC_Introspec/2023_fc_introspection/downloads/MIPAV/
chmod +x ./installMIPAV_5.4.2_amd64.bin
./installMIPAV_5.4.2_amd64.bin

NOTE: When asked where to install the sofware, please select /data/SFIMJGC_Introspec/2023_fc_introspection/sw/mipav

2. Install JIST, JIST-CRUISE and CBS Tools

  1. Open a terminal on an spersist node with graphical access (e.g., in no-machine or VNC)

  2. Download the JIST + CBS Tool bundle from the link provided above.

  3. Unzip the downloaded file into /data/SFIMJGC_Introspec/2023_fc_introspection/downloads/

  4. Start Mipav

cd /data/SFIMJGC_Introspec/2023_fc_introspection/sw/mipav
./mipav
  1. On the main MIPAV window go to Plugins --> Install Plugins

NOTE: If this is your first time opening mipav, the software will tell you that the plugins folder was missing and that you need to restart MIPAV. Follow those instructions. As a result of this, MIPAV will create a folder on your home directory /home/<your_user_name>/mipav/.

  1. On the Install Plugins window, press on Browse and in the new pop-up menus select the folder /data/SFIMJGC_Introspec/2023_fc_introspection/downloads/ and press ok.

  2. Select the three jar files that were part of the JIST+CBS Tools bundle (namely JIST-CRUISE-2014Dec12-03-37PM.jar,CBSTools-release-3.0.8-160229.jar, and external-scripts.jar) on the left list and press the right pointing arrow on the middle of the install plugin dialog to add it to the empty list on the right.

  3. Press Install Plugins. This operation is quite slow and it will throw some "plugin not found" messages on the terminal. That's ok as long as by the end of the installation process, you see a new menu item named JIST on the Plugins menu from the main MIPAV window.

  4. Once the plugins are installed go to Plugins --> JIST --> JISTLayoutTool

  5. On the new window that opens, go to Project --> Rebuild Library

  6. Once this process ends, go back to the main Mipav window and close the program.

Clone the LEMON Pre-processing Pipelines (Modified version for NIH Biowulf environment)

  1. Open a terminal in an spersist node and create/enter target folder
cd /data/SFIMJGC_Introspec/2023_fc_introspection/code/
  1. Clone the pre-processing pipelines

git clone https://github.com/nimh-sfim/lsd_pipelines_nih.git

4. Create necessary local python environments

In this project we use two different environments. This is mostly becuase some required packages have not yet been updated to work with pandas >=2.0.0. The table below list the environment we used for each notebook, and some differences in packages.

Notebook Environment Comments
S00_OrganizeData fc_introspection_2023b_py310.yml Most common environment
S01_NC_run_structural.CreateSwarm fc_introspection_2023b_py310.yml
S02_NC_run_func_preproc.CreateSwarm fc_introspection_2023b_py310.yml
S03_QA_ExcessiveMotion fc_introspection_2023b_py310.yml
S04_TransformToMNI.CreateSwarm fc_introspection_2023b_py310.yml
S05_SegmentT1 fc_introspection_2023b_py310.yml
S06_NuissanceRegression fc_introspection_2023b_py310.yml
S07_PrepareAtlases_400 fc_introspection_2023_py310.yml Needs pandas < 2.0.0 becuase of pd.append()
S08_Extract_ROI_ts fc_introspection_2023b_py310.yml
S09_ExploreFCmatrices_Schaefer_AAL2 fc_introspection_2023b_py310.yml
S10_SNYCQ_Exploration fc_introspection_2023b_py310.yml
S11_SNYCQ_sbcNNMF_Hyperparameters fc_introspection_2023_py310.yml has scikit_posthocs,nbformat and plotly
S12_SNYCQ_Clustering_Confounds fc_introspection_2023b_py310.yml
S13_NBSprepape_BasicAndSbjAware fc_introspection_2023b_py310.yml
S14_PostNBS_PlotFCdiffs fc_introspection_2023b_py310.yml
S15_NBS_Literature_Search_Results fc_introspection_2023b_py310.yml
S16a_CPM_subject_aware.CreateSwarm fc_introspection_2023b_py310.yml
S17_CPM_View_PredictionResults fc_introspection_2023_py310.yml Needs pandas < 2.0.0, statsmodels
S18_CPM_Dashboard fc_introspection_2023_py310.yml For compatibility with S17
S19_CPM_Literature_Comparison fc_introspection_2023_py310.yml For compatibility with S17
Clone this wiki locally