run_BM_bed_profile

run_BM_bed_profile

This folder contains the bone marrow CoCnT fragment QC and pseudo-bulk BED/BigWig generation workflow.

Overview

The workflow is split into two R scripts:

• run_qc.R: loads fragment BED files, builds fragment-per-cell summaries, merges them with sample metadata, and produces QC plots.
• run_pseudo_bulk_bed.R: filters fragments to annotated cells, writes pseudo-bulk BED files by cluster and histone mark combination, and converts those BED files into bedGraph and BigWig tracks.

Inputs

Common metadata inputs:

• ../data/sample_sheet_PR001798.tsv
• ../data/sample_sheet_PR001799.tsv
• ../data/sample_sheet_PR001855.tsv
• ../data/sample_sheet_PR001856.tsv

QC-specific raw inputs:

• BED files under ../data/Janssens_Lab_Data/BMMC_Beds/PR001798_sciCoCnT_JAND_DNA_ALIGNMENT_WELLID_BARCODES/BED/
• BED files under ../data/Janssens_Lab_Data/BMMC_Beds/PR001799_sciCoCnT_JAND_DNA_ALIGNMENT_WELLID_BARCODES/BED/
• BED files under ../data/Janssens_Lab_Data/BMMC_Beds/PR001855_sciCoCnT_JAND_DNA_ALIGNMENT_WELLID_BARCODES/BED/
• BED files under ../data/Janssens_Lab_Data/BMMC_Beds/PR001856_sciCoCnT_JAND_DNA_ALIGNMENT_WELLID_BARCODES/BED/

Pseudo-bulk-specific inputs:

• ../run_BM_archR_analysis/tmp/table_cell_cluster_annotation_final_round2.tsv This file should be prepared in advance by running the ArchR analysis workflow in ../run_BM_archR_analysis/run_117_joint_umpa.R.
• ./tmp/all_fragments.rds
• ../data/ref_genome/hg38.sizes, please prepare this file in advance by downloading the reference genome FASTA and running samtools faidx on it.

Outputs

Files written by the workflow include:

• ./tmp/all_fragments.rds
• ./tmp/cell_fragments.tsv
• ./tmp/table_good_cells_fragments.tsv
• ./tmp/pseudo_bulk_bed/*_pseudo_bulk.bed
• ./tmp/pseudo_bulk_bed/*.sorted.bed
• ./tmp/pseudo_bulk_bed/*.bedGraph
• ./tmp/pseudo_bulk_bed/*.bw
• ./tmp/pseudo_bulk_bed/*.scale_co.bedGraph
• ./tmp/pseudo_bulk_bed/*.scale_co.bw
• ./tmp/pseudo_bulk_bed/*.scale_cell_co.bedGraph
• ./tmp/pseudo_bulk_bed/*.scale_cell_co.bw

run_qc.R also renders QC plots to the active graphics device via httpgd.

Typical Order

1. Run run_qc.R
2. Run run_pseudo_bulk_bed.R to generate pseudo-bulk BED and BigWig tracks.