Structural alignment using the generalized Euclidean distance between conformations

TL;DR

This paper introduces a novel approach for structural alignment of extended objects like polymers using a generalized Euclidean distance as a cost function, revealing differences from traditional RMSD-based methods.

Contribution

It pioneers the use of generalized Euclidean distance for aligning extended structures, comparing its effectiveness with existing metrics like RMSD and MRSD.

Findings

Minimal distance often yields different global alignments than RMSD.

Generalized Euclidean distance is effective for aligning extended structures.

Different cost functions influence the resulting structural alignment.

Abstract

The usual Euclidean distance may be generalized to extended objects such as polymers or membranes. Here, this distance is used for the first time as a cost function to align structures. We examined the alignment of extended strands to idealized beta-hairpins of various sizes using several cost functions, including RMSD, MRSD, and the minimal distance. We find that using minimal distance as a cost function typically results in an aligned structure that is globally different than that given by an RMSD-based alignment