- Title
- Comparison of protein binding microarray derived and ChIP-seq derived transcription factor binding DNA motifs
- Creator
- Hlatshwayo, Nkosikhona Rejoyce
- Subject
- Protein binding
- Subject
- DNA
- Subject
- DNA microarrays
- Subject
- Transcription factors
- Subject
- DNA-protein interactions
- Subject
- Gene regulatory networks
- Date Issued
- 2015
- Date
- 2015
- Type
- Thesis
- Type
- Masters
- Type
- MSc
- Identifier
- vital:4146
- Identifier
- http://hdl.handle.net/10962/d1017907
- Description
- Transcription factors (TFs) are biologically important proteins that interact with transcription machinery and bind DNA regulatory sequences to regulate gene expression by modulating the synthesis of the messenger RNA. The regulatory sequences comprise of short conserved regions of a specific length called motifs . TFs have very diverse roles in different cells and play a very significant role in development. TFs have been associated with carcinogenesis in various tissue types, as well as developmental and hormone response disorders. They may be responsible for the regulation of oncogenes and can be oncogenic. Consequently, understanding TF binding and knowing the motifs to which they bind is worthy of attention and research focus. Various projects have made the study of TF binding their main focus; nevertheless, much about TF binding remains confounding. Chromatin immunoprecipitation in conjunction with deep sequencing (ChIP-seq) techniques are a popular method used to investigate DNA-TF interactions in vivo. This procedure is followed by motif discovery and motif enrichment analysis using relevant tools. Protein Binding Microarrays (PBMs) are an in vitro method for investigating DNA-TF interactions. We use a motif enrichment analysis tools (CentriMo and AME) and an empirical quality assessment tool (Area under the ROC curve) to investigate which method yields motifs that are a true representation of in vivo binding. Motif enrichment analysis: On average, ChIP-seq derived motifs from the JASPAR Core database outperformed PBM derived ones from the UniPROBE mouse database. However, the performance of motifs derived using these two methods is not much different from each other when using CentriMo and AME. The E-values from Motif enrichment analysis were not too different from each other or 0. CentriMo showed that in 35 cases JASPAR Core ChIP-seq derived motifs outperformed UniPROBE mouse PBM derived motifs, while it was only in 11 cases that PBM derived motifs outperformed ChIP-seq derived motifs. AME showed that in 18 cases JASPAR Core ChIP-seq derived motifs did better, while only it was only in 3 cases that UniPROBE motifs outperformed ChIP-seq derived motifs. We could not distinguish the performance in 25 cases. Empirical quality assessment: Area under the ROC curve values computations followed by a two-sided t-test showed that there is no significant difference in the average performances of the motifs from the two databases (with 95% confidence, mean of differences=0.0088125 p-value= 0.4874, DF=47) .
- Format
- 53 p.
- Format
- Publisher
- Rhodes University
- Publisher
- Faculty of Science, Biochemistry and Microbiology
- Language
- English
- Rights
- Hlatshwayo, Nkosikhona Rejoyce
- Hits: 2035
- Visitors: 2170
- Downloads: 162
Collections
RU Department of Biochemistry and Microbiology (2015-)
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | SOURCEPDF | 3 MB | Adobe Acrobat PDF | View Details Download |