scATAC-seq: Comparing chromatin accessibility in hematopoietic stem cells (HSCs) of young vs aged mice

I wrote this R/Bioconductor script analysis pipeline for single-cell ATAC-seq (scATAC-seq) data looking at differential accessibility in hematopoietic stem cells (HSCs) from 10-week-old 'young' mice vs 20-month-old 'aged' mice. The data is from the NCBI Gene Expression Omnibus (GEO) repository and is obtained by running the Cell Ranger ATAC pipeline which carries out the following steps:

• demultiplexing of raw BCL base call files into FASTQ files
• read alignment
• barcode counting
• peak calling with reference “refdata-cellranger-arc-mm10-2020-A-2.0.0”,
• outputs: BED file of peaks, CSV file with cell barcodes metadata, TSV/BED file of each unique fragment and associated cell barcode, etc

Overall Functionality:

1. Data Loading and Preprocessing: It loads necessary R packages (Signac, Seurat, tidyverse, etc.) and reads in BED peak files, CSV cell barcodes metadata, and TSV/BED fragment files for young and aged single-cell ATAC-seq datasets.
2. Peak Filtering and Common Set Creation: It identifies a common set of peaks by reducing peaks from individual datasets and filters out peaks based on width criteria.
3. Initial Quality Control: Filters out low-quality cells based on specific cutoffs for various quality control metrics.
4. Count Matrix Generation: Generates count matrices for both young and aged datasets based on the common peak set.
5. Seurat Object Creation: Constructs Seurat objects for each dataset.
6. Integration and Dimensionality Reduction: Integrates datasets, performs TF-IDF normalization followed by SVD, and conducts UMAP-based visualization and clustering.
7. Gene Annotation and Analysis: Extracts gene annotations, adds them to the Seurat object, and conducts gene activity analysis.
8. Normalization and Visualization of "RNA" Data: Normalizes gene activity "RNA" data, visualizes canonical marker genes, and identifies differential peaks between young and aged datasets.
9. Visualization of Peaks: Visualizes coverage plots for selected peaks and their closest genomic features.

Input Files Required:

• Individual peak files for young and aged datasets (youngPeaks.txt, agedPeaks.txt)
• Metadata files for young and aged datasets (GSM5723631_Young_HSC_singlecell.csv, GSM5723632_Aged_HSC_singlecell.csv)
• Fragment files for young and aged datasets (GSM5723631_Young_HSC_fragments.tsv.gz, GSM5723632_Aged_HSC_fragments.tsv.gz)
• Annotation file (EnsDb.Mmusculus.v79)

R/Bioconductor Packages and Bioinformatics Tools Required:

• Signac
• Seurat
• tidyverse
• patchwork
• GenomicRanges
• future
• EnsDb.Mmusculus.v79 (annotation)

Outputs:

• Seurat objects (combined, young, aged) containing integrated and processed single-cell ATAC-seq data.
• Visualizations: Various plots for quality control, gene activity, differential peaks, UMAP visualization, coverage plots for peaks, and more.