Trajectories through semantic spaces in schizophrenia and the relationship to ripple bursts

PNAS, October 2023

Matthew M Nour, Daniel C McNamee, Yunzhe Liu, Raymond J Dolan matthew.nour@psych.ox.ac.uk

Main scripts

To reproduce paper results (within tolerance of stochastic procedures like agglomerative clustering, permutation-based z-scoring, and UMAP)

semantic_master_movingAv_repo.m

• Wrapper script for all NLP analyses excluding computational modelling.
• Performed on a sliding-window basis and list-specific permutation-based z-scoring, as in paper.
• Analyses performed separately for category/letter task, and semantic/orthographic similarity (task/similarity metric set in the analysis_options structure)
• Semantic distances recomputed in code using fastText embedding
• Orthographic distances loaded in pre-computed
• Travelling salesman analyses is time-intensive, so I provide precomputed .mat files for this (to plot travelling salesman results using the precomputed .mat files, see plot4paper_travellingSalesman_plotting_module.m in the plotting folder)

semantic_master_modelling_concatTasks.m

• Wrapper script for computational modelling analysis
• Precomputed models already saved in modelling_output folder
• For model comparison: concat_model_comparison_repo.m
• To visualise results using fitted parameters (including correlations with clinical/cognitive/MEG variables): plot4paper_winning_model.m

Note on data_store_repo.xlsx

Note that the participant-level clinical, demographic and MEG data is in the data_store_repo.xlsx, however the participant numbering here matches that in a previous publication. I provide the .m code to reorder these rows to match the PNAS numbering (called automatically in all plotting scripts that relate variables in data_store_repo.xlsx to behavioural language measures).