Geometry Analysis for Protein Secondary Structures Matching Problem

Abstract

De novo modeling is a promising computational approach to model the structure of proteins using Cryo-Electron Microscopy data. Not all data produced is within a resolution that enables us to visualize the atomic-structure of the molecule. However, information like secondary structure locations is detectable. At an intermediate step in de novo modeling, the matching between these locations and the amino acid sequence segments that underlie these secondary structure elements needs to be addressed. Many tools have been developed to address this matching problem including DP-TOSS. In this paper, we propose a recast of DP-TOSS that incorporates a geometry-based analytical function to improve accuracy. A test of 15 proteins shows that our analytical function has a positive impact on the accuracy of DP-TOSS.