| dc.contributor.advisor | Beißbarth, Tim Prof. Dr. | |
| dc.contributor.author | Wachter, Astrid | |
| dc.date.accessioned | 2017-05-17T09:38:42Z | |
| dc.date.available | 2017-05-17T09:38:42Z | |
| dc.date.issued | 2017-05-17 | |
| dc.identifier.uri | http://hdl.handle.net/11858/00-1735-0000-0023-3E4B-9 | |
| dc.identifier.uri | http://dx.doi.org/10.53846/goediss-6296 | |
| dc.language.iso | eng | de |
| dc.rights.uri | http://creativecommons.org/licenses/by-nc-nd/4.0/ | |
| dc.subject.ddc | 570 | de |
| dc.title | Data Integration of High-Throughput Proteomic and Transcriptomic Data based on Public Database Knowledge | de |
| dc.type | doctoralThesis | de |
| dc.contributor.referee | Beißbarth, Tim Prof. Dr. | |
| dc.date.examination | 2017-03-22 | |
| dc.description.abstracteng | With the advance of high-throughput methods enabling deep characterization of the cell
on different cellular layers, ideas to combine different data types for inference of regulatory
processes have emerged. Such integration promises an improved molecular understanding of
physiological and pathophysiological mechanisms, which aids in the identification of drug
targets and in the design of therapies. Current integration approaches are based on the
idea of reducing false negatives by reinforcing concordant information between datasets. In
most cases optimized for a specific integration setting and data structure, these approaches
are rarely accompanied by bioinformatic tools enabling researchers to work on their own
datasets.
In this thesis I present the public knowledge guided integration of phosphoproteomic,
transcriptomic and proteomic time series datasets on the basis of signaling pathways. This
integration allows to follow signaling cascades, to identify feedback regulation mechanisms and
to observe the coordination of molecular processes in the cell by monitoring temporal variation
upon external perturbation. To extract these cellular characteristics the cellular layers on
which the individual datasets have been generated are taken into consideration. Separate
downstream and upstream analyses of phosphoproteome and transcriptome data, respectively,
and subsequent intersection analysis are coupled with a combination of network reconstruction
and inference methods. Graphical consensus networks and co-regulation patterns can be
extracted by this cross-platform analysis. Moreover, it provides high flexibility in terms of
high-throughput platforms used for data generation as analysis is based on preprocessed
datasets.
On the examples of epidermal growth factor signaling and B cell receptor signaling
we were able to show that the results gained by this integration method confirm known
regulatory patterns but also point to interactions that were not described previously in these
contexts. This is demonstrated by performing a response-specific analysis instead of the
typical layer-specific analysis.
Limitations of the approach described here are linked to database-bias and -dependency,
to the low temporal resolution of high-throughput measurements and to data standardization.
While overcoming these issues constitutes a challenge for the whole systems biology community,
the integration approach itself can be optimized in future by working with refined disease-specific and tissue-specific signaling pathway models and database entries. The presented
integration method was implemented as R software package 'pwOmics' and made available
to other researchers. | de |
| dc.contributor.coReferee | Wingender, Edgar Prof. Dr. | |
| dc.subject.eng | cellular layer | de |
| dc.subject.eng | data integration | de |
| dc.subject.eng | high-throughput data | de |
| dc.subject.eng | proteomics | de |
| dc.subject.eng | transcriptomics | de |
| dc.identifier.urn | urn:nbn:de:gbv:7-11858/00-1735-0000-0023-3E4B-9-7 | |
| dc.affiliation.institute | Göttinger Graduiertenschule für Neurowissenschaften, Biophysik und molekulare Biowissenschaften (GGNB) | de |
| dc.subject.gokfull | Biologie (PPN619462639) | de |
| dc.identifier.ppn | 887658547 | |