This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ PrePrints) and either DOI or URL of the article must be cited.
In this thesis, a protein structure determination using chemical shifts is presented. The method is implemented in the open source PHAISTOS protein simulation framework. The method combines sampling from a generative model with a coarse-grained force field and an energy function that includes chemical shifts. The method is benchmarked on folding simulations of five small proteins. In four cases the resulting structures are in excellent agreement with experimental data, the fifth case fail likely due to inaccuracies in the energy function. For the Chymotrypsin Inhibitor protein, a structure is determined using only chemical shifts recorded and assigned through automated processes. The CA-RMSD to the experimental X-ray for this structure is 1.1 Å. Additionally, the method is combined with very sparse NOE-restraints and evolutionary distance restraints and tested on several protein structures >100 residues. For Rhodopsin (225 residues) a structure is found at 2.5 Å CA-RMSD from the experimental X-ray structure, and a structure is determined for the Savinase protein (269 residues) with 2.9 Å CA-RMSD from the experimental X-ray structure.