NSCLC-ScRNASeq

Personal sc-RNASeq Project that uncovers Tumour Micro-Environment (TME) of a Non-Small Cell Tumour Sample via cell-subtype clustering. This project is still under development. To this date, cell-type clustering was performed to uncover all the different cell types within this tumour sample. The idea is to hopefully perform further analysis via trajectory inference, and cell-to-cell interactions to identify potential therapeutics.

Dependencies

R tidyverse Seurat ggplot2 DoubletFinder hdf5r

Input Data

The raw input data file 20k_NSCLC_DTC_3p_nextgem_Multiplex_count_raw_feature_bc_matrix.h5 was derived from 10X Genomics website. It is not stored in a repository due to large file size. For more information, feel free to contact me.

Output

The processed Seurat object with cell type clustering should be saved in the results/ directory as nsclc_seurat_obj_dblt_adj.rds.

Detailed Steps

1. Data Loading The raw scRNA-seq data is loaded from an HDF5 file using the Read10X_h5 function.

2. Quality Control Cells are filtered based on the number of detected features and the percentage of mitochondrial genes to ensure high-quality data for analysis.

3. Normalization The data is normalized using the NormalizeData function to make gene expression levels comparable across cells.

4. Feature Selection Highly variable features are identified using the FindVariableFeatures function, which are used in downstream analyses.

5. Scaling The data is scaled using the ScaleData function to standardize gene expression values.

6. Dimensionality Reduction Principal Component Analysis (PCA) is performed using the RunPCA function to reduce the dimensionality of the data and highlight the most important features.

7. Doublet Detection and Filtration Doublets are detected using the DoubletFinder package, and the data is adjusted to remove these artifacts.

8. Clustering Cells are clustered using a graph-based clustering approach with the FindNeighbors and FindClusters functions.

9. Visualization The clusters are visualized using UMAP with the RunUMAP function, allowing for the identification of distinct cell populations.

10. Save Results The final Seurat object, with doublets removed, is saved as an RDS file for future use.