Skip to content

Latest commit

 

History

History
84 lines (70 loc) · 3 KB

ISC_Group_Analyses.md

File metadata and controls

84 lines (70 loc) · 3 KB

How to Do ISC Analyses - Within and Between Subjects


Pipeline to complete group analysis

  • Maximum time to complete the processing (per subprocess) as well as what type of job to run is included
  • These times might increase depending on how many subjects you are analyzing


Warp the original files to the 1st subject's template: swarm (1 hr)

bash warping_group_template.sh orig_warped sub-??


Mask the warped originals: sbatch (10 mins)

bash GroupMask_MH.sh masking_warped_files


Within-Subject Correlations: swarm (2 hrs)
Correlates all of the runs across subjects by datatype (2nd echo, OC, TedDN, combined regressors)

bash ISC_correlations.sh movie_A_x_movie_B
bash ISC_correlations.sh movie_A_x_resp_A1
bash ISC_correlations.sh movie_B_x_resp_A1
bash ISC_correlations.sh resp_A1_x_resp_A2


Between-Subject Correlations: swarm (< 12 hrs)
Correlates all of the runs across subjects by datatype (2nd echo, OC, TedDN, combined regressors)

bash ISC_correlations.sh movie_A_x_movie_B_between
bash ISC_correlations.sh movie_B_x_movie_A_between
bash ISC_correlations.sh resp_A1_x_resp_A1_between
bash ISC_correlations.sh resp_A2_x_resp_A2_between


Blur the Between-subject Correlations: swarm (~24hrs)

bash GroupMask_MH.sh blurring_between_correlations


Create the .txt files that contain the table for the 'blurred' between-subject correlations (< 1 min)

bash Create_ISC_datatables.sh isc_dataframe all


Group T-test for Within-Subject Correlations: swarm (10 mins)

bash GroupStats_Corrs.sh Ttest movie_A_x_movie_B all
bash GroupStats_Corrs.sh Ttest movie_A_x_resp_A1 all
bash GroupStats_Corrs.sh Ttest movie_B_x_resp_A1 all
bash GroupStats_Corrs.sh Ttest resp_A1_x_resp_A2 all


Group ISC for Between-Subject Correlations: swarm (3 hrs)

bash GroupStats_Corrs.sh ISC movie_A_x_movie_B all
bash GroupStats_Corrs.sh ISC movie_B_x_movie_A all
bash GroupStats_Corrs.sh ISC resp_A1_x_resp_A1 all
bash GroupStats_Corrs.sh ISC resp_A2_x_resp_A2 all


(Optional): If you'd like to compare performance across certain groups of subjects (i.e., good task performance or low motion or 'sepcial_group' - low motion and task compliant), you can call "motion" or "task_compliant" or "special_group"
(Example)

bash Create_ISC_datatables.sh isc_dataframe motion
bash Create_ISC_datatables.sh isc_dataframe task_compliant
bash Create_ISC_datatables.sh isc_dataframe special_group
Ttest movie_A_x_movie_B task_compliant
Ttest movie_A_x_movie_B motion
ISC movie_A_x_movie_B motion
ISC movie_A_x_movie_B task_compliant


Once everything is done:
Calculate the Average Stat files for Within & Between group map analyses: command line

bash GroupMeanStatTxt.sh extract_means within_dir
bash GroupMeanStatTxt.sh extract_means between_dir


The above Average Stats can be visualized with ../../Visualizations/OHBM2023Poster/Summarization_Plots_KappaRho_Histograms.py