hurst

This repository contains the code and data for the analyses in the eLife 2020 paper 'Intrinsic excitation-inhibition imbalance affects medial prefrontal cortex differently in autistic men versus women'. It implements a method to compute the Hurst exponent (H) from fMRI data as a proxy for the excitation-inhibition (E:I) balance in neural circuits.

Key Features

• E:I Ratio Modeling — Simulates local field potentials (LFP) by manipulating excitation-inhibition ratios using the Gao et al. (2017) model.
• Hurst Exponent Computation — Calculates the Hurst exponent (H) from fMRI time series as a measure of long-range temporal correlations and E:I balance.
• Multi-modal Analysis Pipeline — Integrates in-silico neural modeling, in-vivo DREADD mouse experiments, and human resting-state fMRI analyses.
• Preprocessing & Denoising — Includes scripts for fMRI preprocessing using AFNI, FSL, and wavelet denoising via the Brain Wavelet Toolbox.
• Statistical & Genomic Integration — Performs univariate statistics, partial least squares (PLS) analysis, and gene expression enrichment tests.

Philosophy

The project adopts a multi-scale, cross-species approach to validate H as a biomarker for E:I imbalance, linking computational models with experimental and clinical neuroimaging data.