Transcriptional regulation of glycosyltransferase genes in MCF-7 human breast cancer cell line following drug treatment

Bioinformatics is a subfield in computational science that is principally focused on developing methods and performing data analytics in the areas of proteomics and genomics. In this thesis I draw a link between proteomics and genomics by focusing on the regulation patterns of glycosyltransferase (GT) genes in breast cancer cell line following the treatment with a large set of Food and Drug Administration (FDA) approved drugs. This is based on the understanding that aberrant glycosylation in breast cancer tumours stem from altered GT gene expression. A major goal of genomic research is the identification of genes that have been differentially expressed under abnormal conditions. A gene expression profile provides a snapshot of the transcriptional level of a cell. A comparative gene expression profile between a diseased and normal-state can be used to map out the regulatory mechanisms of disease. In this thesis, the results of Microarray experiments on MCF-7 human breast cancer cell-lines are analysed using statistical and computational tools to identify differentially expressed genes. Here a bioinformatics analysis of the regulation of GT gene expressions was performed to identify a set of glycosylation related genes with the aim of making an inference about their biological functions. A set of raw gene expression profiles from MCF-7 human breast cancer cell-line treated with different therapeutic drugs were obtained from the Connectivity Map (CMap) database. Initially 7,000 gene expression profiles were used and these were treated by 1,309 different FDA-approved drugs. The number of genes initially was counted up to 22,000. Using the Bioconductor open source software in R statistical programming environment a statistical differential expression analysis followed by several data filtering and pre-processing steps were performed to identify up and down regulated GT genes using. Using non-parametric rank sum meta-analysis three cancer drugs and two non-cancer drugs were identified as effective agents able to control the transcriptional regulatory state of GT genes. The study concluded by employing co-expression gene module analysis using the Weighted Gene Co-Expression Network Analysis (WGCNA) package on each of the cancer and non-cancer drug treatments. The gene modules discovered from the analysis were used to perform gene ontology enrichment analysis to identify the biological functions where they were significantly enriched in. The co-expression modules where GT genes have been down regulated by the drugs, were involved in processes such as Wnt signalling and cell surface pattern recognition receptor signalling important for cancer development. Immune response and apoptotic processes in the cell were identified from co-expression modules where GT genes were up regulated. This key finding that the GT gene expressions are markers for treatment analysis points to their use in drug development studies. The second more direct finding is that non-breast cancer specific FDA-approved drugs may have a role in treating breast cancer and may be the subject of future drug repurposing strategies