From One at a Time to Millions at Once
For decades, DNA sequencing meant reading one fragment at a time using the Sanger method — a painstaking process that was nonetheless powerful enough to sequence the first human genome over 13 years. Then came next-generation sequencing (NGS), which turned everything on its head by reading millions of DNA fragments simultaneously in a single instrument run.
The result was a dramatic drop in cost, a dramatic increase in speed, and a complete transformation of what was possible in genomics research and clinical medicine.
The Core Principle: Massively Parallel Sequencing
All NGS platforms share a key concept: instead of sequencing one DNA molecule at a time, they sequence enormous numbers of molecules in parallel. This is why the technology is sometimes called massively parallel sequencing.
The general workflow looks like this:
- Library preparation: DNA is fragmented into smaller pieces, and short adapter sequences are attached to each end.
- Cluster generation (or flow cell loading): DNA fragments are attached to a surface (a flow cell) and amplified into clusters of identical copies.
- Sequencing by synthesis: Fluorescently labeled nucleotides are incorporated one at a time into a growing DNA strand. A camera captures which nucleotide is added at each position in each cluster, reading the sequence.
- Data analysis: The billions of short sequence reads are computationally assembled and aligned to a reference genome.
Short-Read vs. Long-Read Sequencing
Most mainstream NGS platforms produce short reads — typically 100–300 base pairs long. While short reads are highly accurate and cost-effective, they can struggle with repetitive regions of the genome.
Long-read sequencing technologies, such as those developed by Pacific Biosciences (PacBio) and Oxford Nanopore Technologies, can generate reads spanning thousands to tens of thousands of base pairs, making it easier to assemble complex genomic regions and detect structural variants.
| Feature | Short-Read NGS | Long-Read Sequencing |
|---|---|---|
| Read length | 100–300 bp | 1,000–100,000+ bp |
| Accuracy | Very high (>99.9%) | High (improving rapidly) |
| Cost per base | Lower | Higher (decreasing) |
| Best for | Variant calling, RNA-seq | Structural variants, de novo assembly |
What Can NGS Be Used For?
The flexibility of NGS means it underpins a huge range of applications:
- Whole genome sequencing (WGS): Reading an organism's entire genome.
- Whole exome sequencing (WES): Sequencing only protein-coding regions (the exome), which represents about 1–2% of the genome but contains a large proportion of disease-causing variants.
- RNA sequencing (RNA-seq): Measuring which genes are being expressed in a cell or tissue at a given moment.
- ChIP-seq: Identifying where proteins bind to the genome.
- Metagenomics: Sequencing all the DNA in an environmental sample — like a soil or gut microbiome sample — without isolating individual organisms.
- Liquid biopsy: Detecting tumor-derived DNA fragments in blood, enabling non-invasive cancer screening.
The Data Challenge
NGS generates enormous volumes of data. A single whole genome sequencing run can produce hundreds of gigabytes of raw data. Processing, storing, and interpreting this data requires specialized bioinformatics pipelines — a field that has grown enormously alongside sequencing technology itself.
How Much Does It Cost Today?
The cost of sequencing a human genome has dropped from roughly $3 billion in 2003 to well under $1,000 today — one of the most dramatic cost reductions in the history of technology. This trajectory continues, with several companies working toward truly affordable genomic sequencing at scale.
Key Takeaways
- NGS reads millions of DNA fragments simultaneously, making sequencing fast and affordable.
- Short-read platforms are highly accurate and widely used; long-read platforms excel at complex genomic regions.
- NGS powers everything from cancer diagnostics to environmental microbiome studies.
- Managing the resulting data is a major challenge — and a major opportunity for bioinformatics.