GC Content + Sliding Window (Beginner → Useful)

A simple, reproducible Jupyter workflow to compute overall GC% and sliding-window GC profiles from any DNA FASTA file. GC% analysis is a foundational genomics quality-control step: it helps spot compositional biases, sequencing artefacts, and GC-rich/poor regions that often map to functional elements.

• Notebooks: toy demo + real FASTA analyses
• Outputs: publication-ready figures (PNG) and tables (CSV)
• Inputs: any .fasta / .fa / .fna sequence

---

Introduction / Goal

This project shows, end-to-end, how to:

• Compute overall GC% for a sequence
• Generate a sliding-window GC% profile across the genome
• Save results as CSV tables and figures for reports and QC

It starts with a toy sequence (for clarity), then extends the pipeline to real genomes:

• Human mitochondrial DNA (rCRS, NC_012920.1)
• Escherichia coli K-12 MG1655 (NC_000913.3)

The notebooks are beginner-friendly and designed to be adapted to any FASTA input.

---

Methods

Environment: Python (Jupyter), standard library + Matplotlib.

1. FASTA parsing

   • Lightweight reader concatenates sequence lines, extracts header.

2. GC% computation

   • Overall GC% = (G + C) / (A + C + G + T) × 100
   • Sliding windows: configurable window & step to compute GC% per window.

3. Outputs

   • CSV of all windows; top/bottom 10 windows by GC%
   • Figures: line plot (GC% vs position) and single-bar overall GC%

4. Real genomes

   • FASTA retrieved from NCBI (RefSeq accessions below).
   • Parameters tuned to genome size (e.g., mtDNA: smaller window; E. coli: larger).

---

Results

Toy demo (synthetic DNA)

• GC% profile:
  
• Overall GC:
  

Human mtDNA (NC_012920.1)

• GC% profile:
  
• Overall GC:
  
• Tables:
  • All windows: results/mtDNA/gc_windows.csv
  • Extremes (top/bottom 10): results/mtDNA/gc_extremes.csv

E. coli K-12 MG1655 (NC_000913.3)

• GC% profile:
  
• Overall GC:
  
• Tables:
  • All windows: results/ecoli/gc_windows.csv
  • Extremes (top/bottom 10): results/ecoli/gc_extremes.csv

---

Reproduce / Use on your own FASTA

1. Open the notebooks in Jupyter (Anaconda → Jupyter).
2. In the “analyse_real_FASTA” notebook, set fasta_path to your .fasta/.fa/.fna.
3. Choose window/step (smaller for short genomes; larger for long chromosomes).
4. Run all cells. Outputs (PNGs/CSVs) will be written under results/.

Tip: If the single bar looks too “thick,” set xlim (e.g., plt.xlim(-1, 1)) so the width is visible.

---

Next Steps

• Add multi-FASTA support to batch-run many sequences.
• Overlay GC% with annotation tracks (genes, CDS, CpG islands) for interpretation.
• Package the notebook logic as a small Python module/CLI for easier reuse.

---

Data Sources

• Human mitochondrion (rCRS) — NCBI RefSeq NC_012920.1
  FASTA viewer: https://www.ncbi.nlm.nih.gov/sviewer/viewer.fcgi?id=NC_012920.1&db=nuccore&report=fasta
• Escherichia coli K-12 substr. MG1655 — NCBI RefSeq NC_000913.3
  FASTA viewer: https://www.ncbi.nlm.nih.gov/sviewer/viewer.fcgi?id=NC_000913.3&db=nuccore&report=fasta

---

License

This repository is open-sourced under the MIT License. See LICENSE.