Dokument: Efficient FRET-assisted computational structural modelling
| Titel: | Efficient FRET-assisted computational structural modelling | |||||||
| Weiterer Titel: | Effiziente FRET-assistierte computergestützte Strukturmodellierung | |||||||
| URL für Lesezeichen: | https://docserv.uni-duesseldorf.de/servlets/DocumentServlet?id=53301 | |||||||
| URN (NBN): | urn:nbn:de:hbz:061-20200615-102749-8 | |||||||
| Kollektion: | Dissertationen | |||||||
| Sprache: | Englisch | |||||||
| Dokumententyp: | Wissenschaftliche Abschlussarbeiten » Dissertation | |||||||
| Medientyp: | Text | |||||||
| Autor: | Dimura, Mykola [Autor] | |||||||
| Dateien: |
| |||||||
| Beitragende: | Prof. Dr. Seidel, Claus A. M. [Gutachter] Prof. Dr. Gohlke, Holger [Gutachter] Prof. Dr. Kleinekathöfer, Ulrich [Gutachter] | |||||||
| Stichwörter: | fluorescence, FRET, computational modeling, simulations, hybrid modeling, integrative modeling, structure prediction | |||||||
| Dewey Dezimal-Klassifikation: | 500 Naturwissenschaften und Mathematik » 540 Chemie | |||||||
| Beschreibung: | Structures of biomacromolecules and their complexes are often key to understanding the molecules’ functions and underlying mechanisms. For large multidomain proteins, biomacromolecular complexes, partially unstructured proteins, and systems with lowly populated conformational states, experimental structure determination remains challenging. Computational structural modelling techniques aim at elucidation of molecular mechanisms in biological systems, but trade-offs between the required computational power and accuracy of obtained models still remain prohibitive. Förster resonance energy transfer (FRET) experiments yield state-specific structural information on complex constructs, including very dynamic systems with short-lived states and for time scales down to microseconds. However, FRET experiments need to be combined with computer simulations to solve the issue that FRET data is usually too sparse to cover all structural details. This hybrid approach opens up a possibility for a rational and formalized experiment design aimed at highest accuracy and detail with minimal possible effort and expense.
To realize the potential of FRET-assisted hybrid modelling approach, in this work, efforts were focused on two areas: avoiding unnecessary experimental work by prioritizing the most informative measurements and using the obtained experimental data as efficiently as possible. Construction of a detailed structural model requires data from multiple FRET experiments. Automated selection of most informative FRET pairs based on the computational structural modelling was implemented in order to minimize the number of measurements. Methods to enhance computational modelling using FRET data were implemented for simulations at different levels of coarse-graining, from FRET-restrained all-atom Molecular Dynamics simulations to FRET‑guided normal mode-based coarse-grained geometric simulations and FRET-restrained rigid body sampling. In order to assess the accuracy of produced models, quantitative quality estimate based on careful cross-validation analysis was implemented for FRET-assisted structures. Thorough propagation of experimental errors onto structural modelling results enabled reliable precision estimation and opened up a possibility to verify error estimation procedures by benchmarking against yardstick molecules like double-stranded RNA. These methods were proven in benchmarks with simulated fluorescence data and in experiments with T4 lysozyme protein, where resolution of ~3 Å was achieved for FRET-selected conformers, and the structural mechanism behind the enzyme’s catalytic function was demonstrated. Study of human guanylate binding protein 1 (hGBP1) allowed to connect its conformational dynamics to immune response mechanism. Investigation of chromatin fibre complex helped understanding the mechanisms behind the gene access regulation. These studies establish a streamlined FRET-assisted computational structural modelling procedure and reliable quality assessment of the results. Synergetic combination of FRET experiments and computational structural modelling techniques reveal mechanisms of biomolecular interactions at an otherwise inaccessible detail and scope. | |||||||
| Lizenz: |  Urheberrechtsschutz | |||||||
| Fachbereich / Einrichtung: | Mathematisch- Naturwissenschaftliche Fakultät » WE Chemie » Physikalische Chemie und Elektrochemie | |||||||
| Dokument erstellt am: | 15.06.2020 | |||||||
| Dateien geändert am: | 15.06.2020 | |||||||
| Promotionsantrag am: | 31.07.2019 | |||||||
| Datum der Promotion: | 26.05.2020 |
