Abstract
DSSP is the most commonly used method to assign protein secondary structure. It is based on a hydrogen-bond definition with an energy cutoff. To assess whether hydrogen bonds defined in a parameter-free way may give more generality while preserving accuracy, we examine a series of hydrogen-bond definitions to assign secondary structure for a series of proteins. Assignment by the strongest-acceptor bifurcated definition with provision for unassigned donor hydrogens, termed the SABLE method, is found to match DSSP with 95% agreement. The small disagreement mainly occurs for helices, turns, and bends. While there is no absolute way to assign protein secondary structure, avoiding molecule-specific cutoff parameters should be advantageous in generalizing structure-assignment methods to any hydrogen-bonded system.
Highlights
The assignment of secondary structure is an important tool in simplifying and characterizing the structure of a protein
76.8 agreement in hydrogen bonds (HBs) is comparable for all three variables ECOHN, FCOHN, and FOH but gets worse as more bifurcation is included going from Strongest Acceptor (SA) to Strongest Acceptor plus Bifurcation (SAB) to Strongest Two Acceptors (STA)
The results show that parameter-free HB definitions can be applied to the classification of protein secondary structure, yielding results that are comparable to DSSP with its energycutoff HB definition
Summary
The assignment of secondary structure is an important tool in simplifying and characterizing the structure of a protein. Secondary structure motifs such as helices and beta sheets are distinguished by particular patterns of backbone hydrogen bonds (HBs),[1−5] the backbone atoms,[6,7] the Cα atoms,[1,3,8−11] or Cα contacts.[12] The most widely used algorithm for the assignment of the secondary structure of a protein from its 3D structures is DSSP, which makes use of HBs.[2] HBs, like any intermolecular interaction, are strong at short-range and diminish continuously out to zero at infinity. To be useful in a simple classification scheme they must be treated as stepfunctions, either on or off. This introduces some arbitrariness as to where to put the cutoff. The standard approach is to use fixed cutoff parameters in some measure of HB strength such as energy or geometry,[1−3] which assumes that the cutoff parameters are suitable for all proteins and all conditions
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
