Browsing by Author "Karabekmez, Muhammed Erkan."

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Comparative analysis of teafs and NP analysis to integrate interactome and transcriptome data to reveal response to C-pulse in Saccharomyces Cerevisiae
    (Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2010., 2010.) Karabekmez, Muhammed Erkan.; Kırdar, Betül.
    Sudden glucose introduction to S. cerevisiae cells grown in carbon limited medium triggers a complex response mechanism. Time series expression data can be used to identify differentially active genes after the C-pulse. However, in order to reveal dynamics of the response and formulate more meaningful hypotheses, integration of interactome data and transcriptome data can be useful. In this study, two previously proposed integrative methods, Topological Enrichment Analysis of Functional Subnetworks (TEAFS) and NP analyses were applied using the time series expression data. In TEAFS, condition specific networks corresponding to each time point were constructed based on presence/absence criteria. In order to analyze local dynamic topological changes in the interaction network biological process modules as group of connected proteins with the same GO biological process term were identified. Then standard deviations of topological parameters of the modules were scored to identify significantly active biological processes after the C-pulse. 216 distinct modules were identified as active. The most active modules during overall time span were found to be related to ribosome biogenesis, transcription, rRNA processing, regulation of transcription and translation. Module related to regulation of transcription from RNA polymerase II promoter was identified in all distinct time spans. Nine, six and 96 active modules were identified as seconds, minutes and hours specific modules respectively. In NP analysis, protein-protein interaction network were reduced to a global active subnetwork (NP network) by classifying interactions as correlated, anti-correlated or not correlated based on expression profiles. Four modules of similar expression profiles were determined by hierarchical clustering of expression profiles. Gene Ontology annotations of these modules revealed down-regulated or up-regulated biological processes and their possible interactions after a glucose pulse. The expression level of module M decreased and that of module P increased significantly with time until the 8th minute. The expression levels of these two modules changed in opposite directions after 8th minute. The expression level of D module displayed similar trends with M module but in a more smooth way. The expression level of module S was almost constant over entire time. The genes of P, M, D and S modules were found to be enriched in 147, 97, 169 and 137 distinct GO biological process terms, respectively. The module P was enriched in GO annotations related to ribosome biogenesis, transcription, and rRNA processing. Metabolic process, oxidation reduction, cellular response to heat and transport were found to be enriched terms in module M. The module S was enriched in translation, DNA repair, transport, cell cycle, chromatin modification, and glycolysis. The module D was enriched in the terms transport, protein amino acid phosphorylation, and ubiquitin-dependent protein catabolic process. Enrichment in transcription in both S and D modules were significant. PSA1 which is related to cell wall biosynthesis was determined to be contributing to the communication between P and D modules and HHF1 which is related to chromatin assembly was identified to be contributing to the communication between P and M modules. The active biological processes which were identified by two distinct approaches using the same transcriptome data set were compared in order to interpret the advantages /disadvantages of these two approaches.
  • Thumbnail Image
    Item
    Network topology and dynamic data analysis in saccharomyces cerevisiae
    (Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2016., 2016.) Karabekmez, Muhammed Erkan.; Kırdar, Betül.
    Biological systems which can be represented as networks and graphs are highly dynamic and responsive to environmental and genetic perturbations in a time dependent manner. These networks are hierarchically organized and consist of tightly clustered groups of proteins that work together as part of a biological process or a complex to achieve a specific function in a cell. With the emergence of high-troughput dynamic datasets, dynamic data analysis became a challenge in systems biology with the other challenges such as representation of biological systems as networks and elucidation of graph properties of these networks biologically and integration of multi –omics datasets in order to extract biologically meaningful results. The aim of this thesis is to develop a novel metric of centrality to identify biologically important nodes and to develop novel approaches to investigate dynamic datasets. In the first part, a novel global metric of centrality, weighted sum of loads eigenvector centrality (WSL-EC), counting all eigenvectors was proposed to identify essential and biologically central nodes. WSL-EC was found to outperform in capturing biologically central nodes, such as pathogen-interacting, HIV-1, cancer, ageing, and disease-related genes and genes involved in immune system process and related to autoimmune diseases in the human interactome compared with other metrics of centrality. In the second part dynamic transcriptional response of S. cerevisiae cells to doxorubicin, which is used as chemotherapeutic reagent in the treatment of different types of cancer, was monitored by quantification of RNA transcripts in cells which were grown in a chemostat fermenter, through microarray technology. Resulting dynamic transcriptome data were investigated by using different approaches and integrating interactome and regulome. The clustering and analysis of the transcriptomic response of S. cerevisiae cells to doxorubicin indicated that the genes involved in DNA replication, mismatched repair, cell cycle and base excision repair pathways were affected and several transcriptional factors were identified. In the third part the data collected from literature related to the transcriptional response of yeast cells to DNA damage was similarly investigated and compared with the response to doxorubicin.