MSc_Thesis_Transcriptomics

Repository for all R scripts and results associated with my research dissertation for the MSc Bioinformatics and Computational Biology degree in University College Cork, 2022.

Project title:

"Transcriptomic analysis of the substantia nigra reveals alterations in key biological systems involved in the pathogenesis of Parkinson’s Disease"

Main objectives:

1. Merge SN gene expression data generated from independent microarray studies for a statistically high-powered transcriptomic analysis of the SN in PD cases and healthy controls.
2. Use the merged microarray dataset to perform differential gene expression and functional enrichment analyses of SN tissue in PD cases and healthy controls.
3. Perform differential gene expression analysis of individual laser-captured dopaminergic neurons of the substantia nigra pars compacta and compare to the results obtained from the whole SN.
4. Use the merged microarray dataset to perform weighted gene co-expression network analysis of SN tissue in PD cases and healthy controls, for the detection of biologically relevant differentially co-expressed gene sets.
5. Achieve a comprehensive tissue-wide insight into the molecular alterations and mechanisms contributing to PD at the transcriptome-level of the SN.

Datasets:

Gene expression data used in this study was downloaded from the Gene Expression Omnibus (GEO).

Merging - GSE8397, GSE20292, GSE49036, GSE7621, GSE20186.

Laser Captured - GSE20141

Scripts folder:

DE_GO_GSEA_workflow.R - Workflow used for data pre-processing, merging, differential expression analysis, GO over-representation analysis and GSEA.

process_meta.R - Manual assembly and processing of the merged sample metadata.

Laser_dissected.R - Differential expression nalysis of Laser Captured Microdissection (LCM) Dataset - GSE20141

WGCNA_workflow.R - Workflow used to perform WGCNA and module profiling.

Helper_functions.R - All functions utlized in this study.

Results folder:

Batch Analysis:

PCA plot, sample correlation matrix sample clustering dendrogram from before and after outlier removal and ComBat adjustment.

Dopaminergic_markers:

Results from the correlation and simple linear regression analysis between RET and dopaminergic markers, including a correlation heatmap, expression heatmap and scatter plots.

GO_over_rep:

Results generated from GO over-representation analysis, including the associated plots.

GSEA:

Results generated from gene set enrichment analysis, including the associated plots. GSEA_GO contains a breakdown of each GO category enrichment.

Laser_dissected:

Results associated with the LCM data series, including PCA plot, correlation matrix, sample clustering, dopaminergic marker expression and limma differential expression results.

limma_results:

Results generated from differential gene expression analysis of the merged microarray dataset, including a plot of the top 30 most signifcant differentially expressed genes.

Processed:

Results generated from pre-processing procedures, including the combined_array, the pre-processed merged dataset and the pre-processed sample metadata. Note that certain objects, such as the 'full_processed_list.rds', could not be uploaded to its file size.

WGCNA:

Results generated from the WGCNA_workflow.R script.

Module_info - Contains module eigengenes (MEs), module genes, signedKME for all genes with all modules (kMEtable), signedKME of each gene for its associated module (moduleKME) and hub genes.

Over-rep - Contains results from GO over-representation analysis of specific module gene sets.

Plots - Contains all plots associated with the WGCNA workflow.

RDS_objects - Contains all .rds objects associated with the WGCNA workflow. Certain files were too large for uplaod.

sessionInfo is located in the main/home directory.