This repository provides R scripts for applying Partial Least Squares (PLS) regression to identify associations between brain connectivity features (independent variables) and continuous clinical outcomes (dependent variables, such as pain severity).
PLS regression is particularly well-suited for neuroimaging and clinical data because:
-
Handles high-dimensional, collinear predictors
Brain connectivity matrices often contain hundreds of connections, many of which are strongly correlated. PLS can effectively reduce dimensionality while preserving meaningful variance. -
Integrates brain–behavior associations
Unlike univariate tests, PLS simultaneously considers the entire connectivity pattern, capturing distributed neural synchrony linked to symptom burden. -
Compatible with resampling-based inference
The method integrates well with permutation and bootstrap tests, providing empirical p-values that strengthen interpretability in small or heterogeneous clinical samples.
reformat.R: Converts connectivity data between arrays, data frames, and matrices.regress.R: Residualizes predictors (connectivity) and the outcome (a clinical variable).pls_model.R: Fits the PLS model, extracts coefficients, and calculates resampling p-values.main_run.R: The primary script to run the entire analysis pipeline from start to finish.README.md: This documentation file.
If you use this code, please cite:
Saberi et al. (2025). Neural synchrony in the pain connectome is associated with pain severity and its interactions with mental health outcomes: a transdiagnostic study using magnetoencephalography and multivariate modeling.