Computational studies and embedding strategies for the characterisation of a novel covalent bond in proteins
by Jin Ye
Date of Examination:2024-09-20
Date of issue:2025-03-25
Advisor:Prof. Dr. Ricardo A. Mata
Referee:Prof. Dr. Kai Tittmann
Referee:Prof. Dr. Burkhard Geil
Files in this item
Name:Thesis_final.pdf
Size:10.0Mb
Format:PDF
Abstract
English
This work offers a comprehensive study on the bond properties of a noval covalent bond between three electronegative atoms nitrogen, oxygen and sulfur (NOS), which is found in proteins as crosslink between lysine and cysteine. At the presence of an additional cysteine the nitrogen atom of NOS can be further oxidized and extends the bond NOS to a SONOS moiety. The formation mechanism of the NOS-bond is systematically explored and the radical reaction mechanism in the presence of singlet oxygen is found to be favored. Based on such mechanism the formation of the SONOS bond is compared with that of disulfide, which was thought to be the only crosslink in proteins for quite a long time. With a conformational sampling for the transition state of both formation reactions the SONOS was proved to have an advantage. Another topic of this work is the application of a charge model based on atomic electronic multipoles in quantum mechanics/molecular mechanics (QM/MM) embed ding. Atomic multipoles can reproduce the electronic density of a molecule with high accuracy but are inconvenient to be integrated in different quantum chemical soft wares. Point charges, on the other hand, are much easier to be included in any kind of computations. With a proper transformation from the atomic multipoles to point charges the performance of the QM/MM calculation can be improved. To solve the problem that the atomic multipoles require information of the electron density from quantum chemical calculations a neural network is trained to predict the atomic mul tipoles based on the dihedral angles in the amino acid side chains. With this set up it is possible to perform QM/MM calculations in a more efficient way.
Keywords: NOS-bond