OCD modeling documentation

This project aims to model the changes of functional connectivity observed in obsessive-compulsive disorder (OCD) compared to healthy subjects.

We previously confirmed that OCD is associated with disrupted fronto-striatal connectivity using functional MRI at rest (Naze et al., 2022).

We also conducted a clinical trial to assess the effect of transcranial magnetic stimulation (TMS) targeted to the frontal pole to reduce OCD symptoms (Cocchi et al., 2023).

The frontostriatal regions disrupted in OCD: nucleus accumbens (red), putamen (magenta), orbitofrontal cortex (blue), and lateral prefrontal cortex (green).

Here, we provide a model of frontostriatal dynamics derived from a well-known decision making model. Two of those models are coupled to simulate the ventral and dorsal frontostriatal circuits in-silico. Parameters are optimized to reproduce healthy and OCD functional connectivity using Bayesian inference. Virtual interventions targeting sets of model paramteters are performed to simulate the restoration of healthy functional dynamics from the OCD condition. We generate predictions about best intervention targets, and test those predictions using longitudinal dataset of OCD subjects showing improvement of symptoms over time.

Schematic of the analysis steps and modelling framework.

This documentation exposes this modeling and analysis work along four main axis:

Dynamical system analysis

Development and analysis of the dynamical system derived from the Reduced Wong-Wang model (Deco et al., 2013). This includes phaseplane and bifurcation analysis of the model dynamics in two-dimensions.

Simulation-based inference

Large scale simulation framework performing the Bayesian optimization of model parameters to fit healthy and OCD frontostriatal functional connectivity. This section also introduce how to leverage the outputs of the optimization to generate synthetic data.

Restoration analysis

Virtual intervention framework to assess which targeted changes in sets of parameters bring functional connectivity closer to the healthy regime. This implies the definition of intervention efficacy and parameter contribution measures.

Digital twins validation

Validation framework using the concept of digital twins: pairing OCD subjects from the empirical dataset to virtual subjects from the simulated dataset. This pairing permits to associate the improvement of symptoms in the empirical data to the changes of parameters in the model, and test the predictions generated by the restoration analysis.

The package is composed of three major modules:

Models

Classes of dynamical systems derived from the Reduced Wong-Wang model of perception (Wong & Wang, 2006), tailored to modeling OCD, with recording, scoring and plotting functions.

Analysis

Analytical and numerical analysis of the two-populations model. Phase portrait, bifurcation analysis, symbolic analysis and piece-wise linear approximations.

MCMC

Framework for Bayesian optimizations using Approximate-Bayesian Computation with Sequential Monte-Carlo sampling (ABC-SMC) via the pyABC package. Parameters optimizations are performed to fit both patients and controls cohorts. Key parameters are extracted for optimized potential future treatment through inference and analysis of the resulting dynamics.

Workflow

• Installation
• Dynamical system analysis
• Simulation-based inference
• Restoration analysis
• Digital twins validation

API

• API Reference
  • OCD_modeling
  • OCD_modeling.models
  • OCD_modeling.analysis
  • OCD_modeling.mcmc
  • OCD_modeling.utils

Indices and tables

• Index

• Module Index

• Search Page

References

Clewley, Robert. “Hybrid Models and Biological Model Reduction with PyDSTool.” PLoS Comput Biol 8, no. 8 (August 9, 2012): e1002628. https://doi.org/10.1371/journal.pcbi.1002628.

Cocchi, Luca, Sebastien Naze, Conor Robinson, Lachlan Webb, Saurabh Sonkusare, Luke J. Hearne, Genevieve Whybird, et al. “Effects of Transcranial Magnetic Stimulation of the Rostromedial Prefrontal Cortex in Obsessive–Compulsive Disorder: A Randomized Clinical Trial.” Nature Mental Health 1, no. 8 (August 2023): 555–63. https://doi.org/10.1038/s44220-023-00094-0.

Deco, Gustavo, Adrián Ponce-Alvarez, Dante Mantini, Gian Luca Romani, Patric Hagmann, and Maurizio Corbetta. “Resting-State Functional Connectivity Emerges from Structurally and Dynamically Shaped Slow Linear Fluctuations.” Journal of Neuroscience 33, no. 27 (2013): 11239–52. https://www.jneurosci.org/content/33/27/11239.short.

Schälte, Y., Klinger, E., Alamoudi, E., Hasenauer, J. “pyABC: Efficient and robust easy-to-use approximate Bayesian computation.” Journal of Open Source Software. (2022). https://doi.org/10.21105/joss.04304.

Naze, Sebastien, Luke J Hearne, James A Roberts, Paula Sanz-Leon, Bjorn Burgher, Caitlin Hall, Saurabh Sonkusare, et al. “Mechanisms of Imbalanced Frontostriatal Functional Connectivity in Obsessive-Compulsive Disorder.” Brain 146, no. 4 (April 3, 2023): 1322–27. https://doi.org/10.1093/brain/awac425.

Wong, Kong-Fatt, and Xiao-Jing Wang. “A Recurrent Network Mechanism of Time Integration in Perceptual Decisions.” Journal of Neuroscience 26, no. 4 (January 25, 2006): 1314–28. https://doi.org/10.1523/JNEUROSCI.3733-05.2006.