Code for Coll et al., The neural signature of the decision value of future pain

You can execute the line below to reproduce all analyses/figures in the manuscript. The BIDS formatted data root folder is the $DATAPATH folder and this repository is the $CODEPATH folder.

The system parameters (e.g. number of cpus) should be changed in each script. Bootstrapping/permuting can take a long time.

The emotion dataset (need to download data at http://neurovault.org/collections/503) and the pain risk dataset (need to download data at https://openneuro.org/datasets/ds001814) should be in the $DATAPATH/external/pines and the $DATAPATH/external/ds_001814 respectively.

The NPS signature can be obtained on request from Tor Wager. The SIIPS signature and the Canlab parcellation are available at https://github.com/canlab/Neuroimaging_Pattern_Masks.

The striatum mask is availalbe at https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Atlases.

The templates used for plotting can be obtained from https://www.templateflow.org/.

Running the analyses

Preprocessing with fmriprep

You need to sort out paths and add freesurfer license.txt

docker run -it --rm -v $DATAPATH:/data -v $DATAPATH/derivatives:/out  poldracklab/fmriprep:latest /data /out participant --output-spaces MNI152NLin2009cAsym:res-2 --fs-license-file /out/license.txt --fs-no-reconall

Using the container

All python scripts can be executed using the docker container. You can also run all the scripts locally if you install all necessary packages (see requirements.txt),

To pull the container:

docker pull mpcoll2/coll_painvalue_2021:latest

After pulling the container, all scripts can be executed within the container using this command:

docker run -it --rm -v $DATAPATH:/data -v $CODEPATH:/code mpcoll2/coll_painvalue_2021:latest python $SCRIPTPATH

Behavioural analyses and figures

./code/behavioural/behavioural.py

Note that there are no R kernels in the container so the R script should be executed locally.

Rscript $CODEPATH/behavioural/behavioural_linearmodels.R

Create the group mask

./code/preprocessing/make_group_mask.py

GLMs first level

Pain and money first part offer

./code/glms/model_offer1.py

Shocks

./code/glms/model_shocks.py

Anticipation

./code/glms/model_anticipation.py

Single trials decisions

./code/glms/model_offer2_strials.py

MVPA + univariate analyses

Offer 1 pain level

./code/mvpa/pain_level.py

Offer 1 money level

./code/mvpa/money_level.py

Shock level

./code/mvpa/shock_level.py

Collect data for single trial deicsions

./code/mvpa/decision.py

Anticipation level

./code/mvpa/anticipation_level.py

Apply patterns to data with cross validation

./code/mvpa/cv_pattern_expressions.py

Predict participants' choices

./code/mvpa/predict_choice.py

Run analyses on emotion dataset

./code/external_datasets/pines/rating_level.py

Run whole brain cross-prediction searchlights with bootstraps

./code/mvpa/sl_crosspainmoney.py

./code/mvpa/sl_crosspainemo.py

./code/mvpa/sl_crosspainshock.py

Run analyses on pain risk dataset

Preprocessing

docker run -it --rm -v $DS001814PATH:/data -v $DS001814PATH/derivatives:/out  poldracklab/fmriprep:latest /data /out participant --output-spaces MNI152NLin2009cAsym:res-2 --fs-license-file /out/license.txt --fs-no-reconall

GLM

./code/external/ds_001814/glms/model_risk_strials.py

Apply pattern and figure for risk dataset

./code/external/ds_001814/mvpa/risk_level_strials.py

Make all remaining plots (panels assembled manually for manuscript)

./code/figures/figures_pvp.py

./code/figures/figures_mvp.py

./code/figures/figures_sip.py

./code/figures/figures_crosspred.py

./code/figures/figures_decision.py

Run additional supplementary analyses

Computational model and individual differences

./code/beahvioural/computational.py

Note that there are no Matlab kernels in the container so the Matlab script should be executed locally. The VBA toolbox is available at https://mbb-team.github.io/VBA-toolbox/

./code/beahvioural/vba_model_comparison.m

./code/mvpa/indiff_computational_suppmat.py

./code/mvpa/pain_complevel_suppmat.py