UBS-seq_260308_chenyou

UBS-seq (Uracil Bisulfite Sequencing) analysis pipeline for samples YSL-5 and YSL-6, mapping to the human reference genome.

Overview

This project converts an existing Jupyter notebook-based UBS-seq pipeline into a standalone, reproducible bash script. The pipeline processes paired-end sequencing data through trimming, alignment, duplicate handling, and C-to-T conversion table generation.

Pipeline Steps

1. trim -- Adapter trimming and quality filtering (cutadapt)
2. map -- Bisulfite-aware alignment to human genome (hisat-3n)
3. mark_duplicates -- PCR duplicate marking (GATK)
4. dedup -- Duplicate removal (samtools)
5. conv_unconv3n -- C-to-T conversion/unconversion tables (hisat-3n-table)

Files

File	Description
pipeline.sh	Main pipeline script
test_results.sh	Verification script to compare outputs against original notebook results
PIPELINE.md	Original task specification and requirements
PLAN.md	Detailed pipeline documentation, prerequisites, and usage

Quick Start

./pipeline.sh                        # Run full pipeline
./pipeline.sh --start-from map       # Resume from a specific step
./pipeline.sh --sample YSL-5         # Run a single sample

See PLAN.md for full usage details, prerequisites, and reference file setup.

Output

Final output for downstream analysis is in conv_unconv3n/:

• {sample}.tsv.gz -- Raw hisat-3n-table output
• {sample}.tmp.tsv.gz -- Processed table with sample name, depth, and conversion ratios