Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Subjects: | |
| Online Access: | https://doi.org/10.5281/zenodo.20116106 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <p class="MsoNormal"><strong><span>VALLY-Scan v6.2.0: A Sovereign Physics-Informed Machine Learning Framework for High-Resolution Allosteric Mapping through Nuclear Vibrational Coupling</span></strong></p> <p class="MsoNormal"><strong><span> </span></strong></p> <p class="MsoNormal"><span class="ng-star-inserted1"><strong><span>Author:</span></strong></span><span><span class="ng-star-inserted1"><span> Eng. Lionell E. Nava Ramos</span></span></span><span><br><span class="ng-star-inserted1"><strong><span>Affiliation:</span></strong><span> Lead Researcher, Structural Intelligence Unit. Centro de Investigación en Informática (CII-UPTMA), Venezuela. Project V.A.L.L.Y.</span></span></span></p> <p class="MsoNormal"><span>https://orcid.org/0000-0001-9349-3534<br><span class="ng-star-inserted1"><strong><span>Date:</span></strong><span> May 10, 2026</span></span><br><span class="ng-star-inserted1"><strong><span>DOI:</span></strong><span> 10.5281/zenodo.19443677 (v2.0.0)</span></span></span></p> <p class="MsoNormal"><span> </span></p> <h3><span><span class="ng-star-inserted1"><strong><span>Abstract</span></strong></span></span></h3> <p class="MsoNormal"> </p> <p class="MsoNormal"><span class="ng-star-inserted1"><span>Traditional drug discovery and allosteric mapping protocols depend heavily on computationally expensive Molecular Dynamics (MD) simulations or third-party black-box solvers, creating significant barriers for real-time pandemic response and regional technological sovereignty. We present </span><strong><span>VALLY-Scan v6.2.0</span></strong><span>, a standalone Physics-Informed Machine Learning (PIML) framework designed for high-resolution biophysical characterization. Building upon the validated v4.4.0 baseline, this version introduces a paradigm shift from atomic-static to </span><strong><span>nuclear-resonant modeling</span></strong><span>.</span> The framework’s core innovation is a </span><strong><span>Sovereign Manual Matrix Inversion engine</span></strong><span> that eliminates dependencies on external C++ solvers and incorporates a density-dependent stabilization filter (</span><span class="mord"><em><span>ρ</span></em></span>). By integrating Benoît Champagne’s Sum-Over-States (SOS) expressions for vibrational polarizability, VALLY-Scan v6.2.0 successfully couples nuclear fluctuations (calculated via manual GNM/ANM) with induced electronic responses. This allows for the precise identification of allosteric hotspots in highly flexible and anharmonic domains, such as the Receptor Binding Domain (RBD) of viral glycoproteins.<span> <span class="ng-star-inserted1">Benchmarking against large-scale multimeric systems yielded a Pearson correlation of </span></span><span class="mord"><strong><em><span>r</span><span> </span></em></strong></span><span class="mrel"><strong><span>=</span></strong></span><strong><span><span> </span></span></strong><span class="mord"><strong><span>0.9522</span></strong></span> for the SARS-CoV-2 Spike protein (<strong><span>6VSB</span></strong><span>, 2,905 residues) and </span><span class="mord"><strong><em><span>r</span><span> </span></em></strong></span><span class="mrel"><strong><span>=</span></strong></span><strong><span><span> </span></span></strong><span class="mord"><strong><span>0.9102</span></strong></span> for the Andes Virus nucleocapsid (<strong><span>6I2N</span></strong><span>). These results demonstrate the framework's ability to effectively isolate functional signaling from experimental Cryo-EM noise. VALLY-Scan v6.2.0 provides a computationally efficient, hardware-agnostic alternative for large-scale structural biology and rapid drug triage, operating entirely on consumer-grade hardware (Intel i3) without compromising scientific fidelity.</span></p> <p class="MsoNormal"><span class="ng-star-inserted1"><span> </span></span></p> <p class="MsoNormal"><span class="ng-star-inserted1"><span> </span></span></p> <p class="MsoNormal"><span class="ng-star-inserted1"><span> </span></span></p> <p class="MsoNormal"><span class="ng-star-inserted1"><span> </span></span></p> <p class="MsoNormal"><span> </span></p> <p class="MsoNormal"><span class="ng-star-inserted1"><strong><span>Keywords:</span></strong></span><span><span class="ng-star-inserted1"><span> Physics-Informed Machine Learning (PIML), Allosteric Hotspots, Nuclear Polarizability, Sovereign Technology, Structural Biology, Virus Andes, SARS-CoV-2 Spike, Multi-Scale Simulation.</span></span></span></p>