By eliminating the rank-inflation bottleneck, this framework extracts un-distorted 3D atomic binding pockets of mutated oncogenic proteins, enabling O(N 2) polynomial-time de novo small-molecule drug design and targeted inactivation.

Deterministic Low-Rank Matrix Completion for Structural Oncology: Resolving the Distance Geometry Problem in 3D Protein Target Modeling.