| dc.contributor.advisor | Stark, Holger Prof. Dr. | |
| dc.contributor.author | Bunzel, Georg | |
| dc.date.accessioned | 2022-09-22T11:19:29Z | |
| dc.date.available | 2022-09-29T00:50:14Z | |
| dc.date.issued | 2022-09-22 | |
| dc.identifier.uri | http://resolver.sub.uni-goettingen.de/purl?ediss-11858/14254 | |
| dc.identifier.uri | http://dx.doi.org/10.53846/goediss-9452 | |
| dc.language.iso | eng | de |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | |
| dc.subject.ddc | 510 | de |
| dc.title | Neural Network Based Methods for the Conformational Landscape Determination in High Resolution Cryo-Electron Microscopy | de |
| dc.type | doctoralThesis | de |
| dc.contributor.referee | Stark, Holger Prof. Dr. | |
| dc.date.examination | 2021-11-22 | de |
| dc.description.abstracteng | During the last decades 3D transmission electron cryo-microscopy (cryoEM) has emerged to be the method of choice for the study of motion in larger protein complexes. The aim of Single Particle Analysis (SPA) in cryoEM is to combine a large amount of noisy projection images, usually obtained by a transmission electron microscope (TEM), of the same macromolecular complex into one noise reduced structure (or a finite set of them to study the motion of the respective complex over time). This process is computationally very demanding as the amount of images needed increases tremendously for higher resolution levels. The advances in cryoEM, hence, were backed up by the advances in both, instrumentation (e.g. TEMs, sensors, correctors) but also in computational image processing. Various software tools have been proposed over the years to tackle specific subtasks of the image processing cycle. However, they often rely on certain assumptions (e.g. starting models as reference) or require human input when it comes to the sorting of finite conformational states to describe the dynamics of macromolecular complexes.
In this work a novel software tool based on Artificial Neural Networks (ANNs) is proposed. It does not rely on prior information on the studied dataset. Instead, it processes cryoEM particle images fully autonomously without human intervention. Furthermore, it aims to estimate a continuous conformational space of the studied complex, which can be sampled in order to generate smooth trajectories, instead of a small finite set of conformations.
The software tool is then evaluated with respect to its performance on synthetic test data and also on existing cryoEM datasets featuring two different macromolecular complexes. | de |
| dc.contributor.coReferee | Wörgötter, Florentin Prof. Dr. | |
| dc.subject.eng | GAUSS | de |
| dc.subject.eng | Dissertation | de |
| dc.subject.eng | cryoEM | de |
| dc.subject.eng | machine learning | de |
| dc.subject.eng | variational autoencoder | de |
| dc.subject.eng | VAE | de |
| dc.subject.eng | ANN | de |
| dc.identifier.urn | urn:nbn:de:gbv:7-ediss-14254-7 | |
| dc.affiliation.institute | Fakultät für Mathematik und Informatik | de |
| dc.subject.gokfull | Informatik (PPN619939052) | de |
| dc.description.embargoed | 2022-09-29 | de |
| dc.identifier.ppn | 1817360760 | |
| dc.notes.confirmationsent | Confirmation sent 2022-09-22T11:45:01 | de |