Overview

Current computational psychiatry models heavily rely on isolated biochemical parameters, obscuring the non-linear realities of topological vulnerability in the human brain. To address this, I have developed the Bioenergetic Stability Index (ISB): a deterministic, multi-scale Ordinary and Partial Differential Equation (ODE-PDE) architecture.

Caption: Spatiotemporal connectome map visualizing the anisotropic diffusion of metabolic stress across the 148-node Destrieux topology.

Operating strictly on a "Zero-Assumption" basis, the neural network is calibrated at a thermodynamic steady-state (d[ATP]/dt=0). The model integrates six-layer empirical matrices—including the 148-node Destrieux connectome, AHBA transcriptomics, Neuromaps PET density, and pan-ancestry gnomAD data. It mathematically demonstrates that major depressive phenotypes emerge from a focal topological phase transition (a Saddle-Node Bifurcation, [ATP] < 0.5 mM) driven by cumulative allostatic load.

Computational Evidence & Phase Transitions

To mathematically validate this architecture, extreme stress simulations were executed across dynamic cohorts.

Caption: Deterministic trajectory of focal bioenergetic depletion, capturing the precise Saddle-Node Bifurcation threshold.

Caption: Kaplan-Meier estimates tracking systemic bioenergetic survival probabilities across diverse ancestral cohorts (N=40,000).

Caption: Thermodynamic contour mapping the bifurcation zone against cumulative allostatic load and age.

Traction & Current Bottleneck

As an independent researcher, I have completed the foundational ISB architecture and executed a massive Pan-Ancestry Monte Carlo simulation (N=40,000 subjects). The full Python codebase is open-sourced on GitHub (GPL v3.0), and the preprint is publicly deposited on Zenodo.

GitHub: https://github.com/cefiyana-clover/ISB_Thermodynamics_Pipeline

Zenodo: https://doi.org/10.5281/zenodo.20297789

​Remarkably, this entire mathematical architecture, including the N=40,000 Monte Carlo execution, was engineered and run strictly from an Android mobile device utilizing the free, 2-core tier of Google Colab.

​I am now advancing the model to its next evolutionary phase. I have successfully integrated 4 out of 5 advanced computational expansions:

1. ​Adaptive Neuroplasticity

2. ​Stochastic Langevin Noise (SDE)

3. ​Anisotropic Diffusion

4. ​Blood-Brain Barrier (BBB) Filtration Delay

​However, the 5th expansion—Dynamic Epigenetic Drift (time-series transcriptomics)—and the integration of Stochastic Differential Equations (SDE) require computational power that drastically exceeds free 2-core cloud limits. Currently, running 4,000 deterministic subjects takes ~49 minutes. Scaling stochastic models across 40,000+ subjects on free tiers results in forced runtime disconnections.

Funding Utilization (Minimum vs. Goal)

• ​Minimum Funding ($1,500): Hardware Acquisition (Unblocking the Bottleneck). This will directly fund the acquisition of a dedicated local computational node (a mobile workstation with a minimum 6-core/12-thread processor, 32GB DDR5 RAM, and 1TB SSD). This hardware will immediately bypass cloud session limits, allowing continuous, multi-threaded execution of massive stochastic datasets and reducing simulation times from days to hours.

• ​Full Funding Goal ($3,500): Cloud Scaling & Research Autonomy. Achieving the full goal will cover the hardware acquisition ($1,500) plus $500 for dedicated enterprise cloud compute credits (e.g., AWS/GCP instances for specialized PDE rendering) and $1,500 as an independent research stipend. This stipend provides the necessary operational runway to focus exclusively on securing in-vivo time-series transcriptomic data and completing the 5th expansion (Dynamic Epigenetic Drift) to finalize the ISB architecture for public release.

​Why Fund This?

​This project represents extremely high leverage. By funding the essential hardware for an independent researcher, you are directly enabling the advancement of a strictly rigorous, open-source computational neuroscience framework that challenges reductionist psychiatry—without the overhead costs of traditional institutional academia.