ChIPSeq_GPS2_CTCF_PerissiLab

BU-BMSIP / Perissi Lab (Boston University Chobanian & Avedisian School of Medicine)
Reproducible multi-omics analysis framework for ChIP-seq and RNA–chromatin interaction (iMARGI) data, focusing on GPS2-mediated mitochondrial retrograde signaling and its interaction with ATF4, CTCF, and related transcription factors.

---

1. Project Overview

Mitochondria depend on nuclear-encoded proteins, requiring tight coordination of gene expression via anterograde (nucleus→mitochondria) and retrograde (mitochondria→nucleus) signaling.
GPS2 (G-protein Pathway Suppressor 2) is a key retrograde mediator that translocates to chromatin under mitochondrial stress, potentially cooperating with stress-response TFs such as ATF4/ATF5 and the chromatin organizer CTCF.

This repository contains:

• Snakemake-based workflows for scalable, reproducible processing of:
  • GPS2, ATF4, CTCF ChIP-seq datasets (in-house + public)
  • iMARGI RNA–DNA interaction data (including mtRNA–DNA contacts)
• Integrated analyses:
  • Peak calling, intersection, and annotation
  • Motif discovery and promoter topology analysis
  • Signal profiling and heatmaps (deepTools)
  • Functional enrichment (GO/KEGG, GSEA)
  • Mapping of mtRNA–DNA interactions and overlap with TF peaks

---

2. Repository Layout

├── docs/                # Documentation, diagrams, protocol notes
├── envs/                # Conda environment YAMLs (bedtools, deeptools, macs3, homer, etc.)
├── notebooks/           # Jupyter/R notebooks (QC, annotation, enrichment, motif)
├── profile/             # Snakemake profiles (threads, paths, runtime configs)
├── scripts/             # Python/R/Bash utilities & Snakemake workflow files
│   ├── ChIP_CTCF_mouse_cleaned.smk
│   ├── iMargi.smk
│   └── ...
├── adapters_and_annotations/  # Reference FASTA/BED, genome annotations, blacklists
├── Peak_Calls_with_Control.csv
├── Peak_Calls_without_Control.csv
├── Sample_Data_for_CTCF.csv
├── Updated_Sample_Sheet.csv
└── README.md

---

3. Prerequisites

• Access to BU Shared Computing Cluster (SCC) or equivalent HPC
• Conda (≥4.10)
• Snakemake (≥7.0)
• Required tools: bedtools, deeptools, macs3, homer, samtools, pairtools, bwa

Activate environment before running:

conda activate chip_seq

---

4. Running the Pipeline

Dry run first (recommended):

snakemake -s scripts/ChIP_CTCF_mouse_cleaned.smk --profile profile -np

Actual run:

snakemake -s scripts/ChIP_CTCF_mouse_cleaned.smk --profile profile

For iMARGI workflow:

snakemake -s scripts/iMargi.smk --profile iMargi

---

5. Adding New Data

ChIP-seq

1. Place FASTQ files:
   • Raw (unmerged): CTCF_3T3L1/raw_samples/
   • Pre-merged: CTCF_3T3L1/samples/
2. Or add FTP links + SRR IDs to Sample_Data_for_CTCF.csv for auto-download & merge
3. Register sample in Updated_Sample_Sheet.csv
4. Update peak calling configs:
   • With control → Peak_Calls_with_Control.csv
   • Without control → Peak_Calls_without_Control.csv

iMARGI

• Edit scripts/iMargi.smk → update "Global config" with new SRA IDs
• Follow iMARGI preprocessing steps (cleaning, mapping, parsing)

---

6. Key Outputs

• Alignment & QC
  *.bam, multiqc_report.html
• Coverage tracks
  *.bw
• Peak calls
  *_peaks.narrowPeak
• Annotation
  results/annotation/*.bed & gene lists
• Motifs
  motifs/**/knownResults.txt
• Signal profiling
  matrix/*.gz, plots/*.png
• Enrichment analysis
  results/Enrichment/*.tsv
• iMARGI outputs
  output/final_*.pairs.gz + promoter overlap stats

---

7. Example Analyses

• GPS2–ATF4 co-binding during adipocyte differentiation
• CTCF motif enrichment flanking GPS2 peaks
• Condition-specific GPS2 occupancy shifts (day 0 vs day 6)
• mtRNA-binding sites overlapping GPS2/NCOR peaks (T263 endothelial cells)

---

8. References

1. Cardamone et al., Mol Cell, 2018 — GPS2 retrograde signaling mechanism
2. Chen et al., Cell Biol Toxicol, 2022 — ATF4–CTCF cooperation in adipogenesis
3. iMARGI Pipeline — Yan Lab Documentation