V.A.L.L.Y.: An Ultra-Fast Computational Framework Integrating Elastic Network Models and Informed Heuristics for Accelerating Drug Discovery. Software Version: v1.5 (Academic Edition).

The high computational cost of molecular simulation constitutes a critical bottleneck in drug discovery. To address this, we developed VALLY-Scan, a computational framework that offers speed and accessibility. Our method performs an ultra-fast Vibrational Analysis of protein structures using Anisotropic Network Models (ANM) to capture essential collective dynamics. Validated against experimental crystallographic data, the framework yielded a significant Pearson correlation of r=0.67 for the SARS-CoV-2 main protease (Mpro, PDB: 6LU7). Our results identify key residues with high dynamic impact (GLY 278, ASN 277, ARG 279), distinct from the catalytic sites, suggesting potential allosteric sites for drug design. This work demonstrates a powerful, first-tier methodology for rapidly generating therapeutic hypotheses, thereby democratizing computational biophysics.