Sequencing

Technology Snapshot

1.1 Sequencing-by-synthesis with reversible terminators

Illumina’s SBS chemistry uses reversible terminator deoxynucleotides. Each sequencing cycle incorporates one modified nucleotide, detects a fluorescent signal, then removes the terminator and label to allow the next cycle. This is the chemistry described in the founding Solexa/Illumina patent family.

US7057026B2, “Labelled nucleotides,” discloses nucleotide analogues with a base linked to a detectable label through a cleavable linker and/or a sugar linked to a cleavable blocking group. The patent states that these nucleotides are useful in sequencing by sequential addition to a growing polynucleotide chain (Google Patents: US7057026B2).

Illumina’s current SBS description provides the same technical logic: reversible terminators, cluster generation, imaging, and cleavage enable massively parallel base calling (Illumina sequencing technology page).

1.2 Long-read alternative: PacBio HiFi

PacBio’s SMRT/HiFi approach uses circular consensus sequencing of long DNA molecules. PacBio describes HiFi reads as having accuracy above Q30, or >99.9% single-molecule consensus accuracy, while preserving long read length (PacBio HiFi technology).

The complementary use case matters: short-read SBS excels at high-throughput single-nucleotide and small-variant detection at low cost; long-read HiFi improves structural variant detection, repeat resolution, and de novo assembly.

Patent Deep-Dive and Company Filing Cross-Reference

Patent: US7057026B2

US7057026B2 is listed by Google Patents with a 2001 filing date and a 2006 grant date, assigned in the Solexa/Illumina SBS lineage. The core technical contribution is reversible terminator nucleotide chemistry: a blocking group on the sugar and/or a cleavable label on the base allows synchronous, stepwise base addition.

This matters commercially because it is the chemistry that enabled massively parallel SBS. It also matters legally: the original application has a 2001 priority date. A US patent’s term is generally 20 years from earliest non-provisional filing, but exact expiration requires checking USPTO PAIR or a current legal-status primary source. This is an IP-moat risk point.

Cross-reference to company filing: Illumina’s FY2023 Form 10-K describes its sequencing instruments and consumables and identifies its proprietary SBS technology as central to its platform. The 10-K also includes risk-factor language on intellectual property and competition, which is relevant because core SBS patents have finite terms (SEC EDGAR: Illumina 10-K filings).

Exact expiration of any single patent should be confirmed against USPTO records; do not rely on this note for freedom-to-operate analysis. This is not a financial data point.

arXiv Paper Deep-Dive and Cross-Reference

arXiv:1708.01492 — “Minimap2: pairwise alignment for nucleotide sequences”

Minimap2 is a widely used long-read aligner. It indexes minimizers from sequences and chains them through dynamic programming to align long reads to references or to each other. The arXiv version explicitly discusses alignment of Pacific Biosciences and Oxford Nanopore reads, which historically have high error rates and long lengths compared with short Illumina reads (arXiv:1708.01492).

This is not a company-controlled publication, but it is directly relevant to the long-read thesis because PacBio HiFi and ONT data require long-read-aware alignment for assembly, structural variant calling, and diploid phasing.

Cross-reference to company filing: PacBio’s FY2023 Form 10-K describes its SMRT/HiFi platform as generating high-accuracy long reads for applications including human WGS, structural variant detection, and de novo assembly (SEC EDGAR: PacBio 10-K filings). The Minimap2 paper is the algorithmic counterpart: it provides the long-read alignment capability that makes those reads usable at scale.

Parameter Table with Source-Confidence Tags

Recent developments

• 2026-09-29 — Oxford Nanopore joins RISE global rare disease genomics programme. Oxford Nanopore Technologies plc (LSE: ONT) joined RISE, Genetic Alliance's global rare-disease genomics programme (partner laboratories and clinical sites in 19 countries; ~4,000 people with rare diseases sequenced through the network to date). ONT will make its sequencing technology available through RISE partner labs and clinical sites; partners are expected to begin incorporating nanopore sequencing in late 2026 / early 2027. Framed by ONT as expanding clinical access while building real-world evidence for nanopore sequencing. Partnership signed in Singapore on 28 Sep 2026. Material as a commercial/clinical go-to-market win for long-read sequencing. Source: nanoporetech.com

Core technology and source tier

The dominant commercial sequencing technology is short-read sequencing by synthesis (SBS) using reversible terminator chemistry. The core chemistry is documented in a granted US patent, US7057026B2 (“Labelled nucleotides”), granted 2006-06-06 (Google Patents, accessed 2026-04-26). The patent is evidence of recorded legal claims, not proof of commercial performance or yield. The peer-reviewed demonstration of whole-human-genome sequencing using this chemistry is Bentley et al., Nature 2008, DOI 10.1038/nature07517 (Nature, accessed 2026-04-26).

Patent vs paper distinction: US7057026B2 is a granted patent, not an application; Bentley et al. is a peer-reviewed journal article, not a preprint. These are separate source classes. However, the patent and the Bentley paper share an intellectual lineage through Solexa/Illumina invention and author overlap, so independent validation should also come from a separate institution. NIST’s Genome in a Bottle program provides standards-body validation of sequencing performance using reference human genomes (NIST GIAB, accessed 2026-04-26), and Zook et al. 2016, DOI 10.1038/sdata.2016.25, documents that reference material (Scientific Data, accessed 2026-04-26).

I did not locate an independent non-Illumina replication of the exact reversible-terminator chemistry claim in a standards-body document; NIST GIAB validates sequencing output quality but does not independently prove the patent’s legal scope or the company’s current per-genome cost claim.

Quantitative worked calculation

Question: What is the implied sequencing cost per gigabase if Illumina’s claimed $200 human genome at 30x coverage is accurate?

Formula:

Total yield per genome = coverage × haploid genome size

Cost per gigabase = cost per genome / total yield

Inputs:

• Claimed cost per genome = $200 (Illumina NovaSeq X product page, self-reported, accessed 2026-04-26; source)
• Coverage = 30x (common whole-genome diagnostic convention; an assumption, not independently benchmarked here)
• Haploid human genome size = 3.1 × 10⁹ bases (NIST GIAB reference context; source, accessed 2026-04-26)

Calculation:

Total yield = 30 × 3.1 × 10⁹ = 93 × 10⁹ bases = 93 Gb

Cost per gigabase = $200 / 93 Gb = $2.15/Gb

Sensitivity:

• If the actual cost per genome is $500, implied cost = $500 / 93 = $5.38/Gb.
• If coverage is reduced to 20x, required yield = 62 Gb; implied cost at $200 = $3.23/Gb.

These are derived figures, not externally quoted prices. The $200 input is a company claim, not SEC-verified.

Parameter table

Unit economics / business model

Sequencing is monetized primarily as a razor-and-blade: instrument makers sell sequencers and then recurring consumables (flow cells, reagents). Illumina’s business is reported in its SEC 10-K as instrument and consumable revenue; I did not independently re-derive a gross-margin figure, so an exact margin must be labeled needs primary source validation. The company’s own NovaSeq X page claims a $200 human genome at scale, which is a Tier 2 self-reported marketing claim and is not independently confirmed by a regulator filing.

Value capture occurs across layers: instrument producers capture instrument/consumable margin; diagnostic companies and service providers buy or integrate sequencing capacity and monetize per-test revenue; standards bodies consume sequencing output to certify reference materials. The $200 claim, if accurate, implies a derived cost of roughly $2.15/Gb at 30x coverage and 3.1 Gb haploid genome size, as calculated above.

Disconfirming Evidence

The thesis that short-read sequencing retains long-term pricing power and durable intellectual property would be weakened if: (1) US7057026B2 or related claims were invalidated or narrowed in inter partes review or opposition proceedings; (2) Illumina’s SEC 10-K showed declining consumable revenue or gross margin, contradicting the company’s $200 genome claim; (3) independent benchmark results showed new entrants achieving comparable accuracy and yield at materially lower cost — a scenario that is currently unsupported by the retrieved sources but would need primary-source validation; and (4) long-read platforms such as those from Pacific Biosciences gained share in the clinical and diagnostic workflows now dominated by short-read SBS.

Original scope and definitions

DNA/RNA sequencing — the foundational Read-layer technology that converts biological samples into digital information. Illumina's Sequencing by Synthesis (SBS) dominates short-read with ~90% of global sequencing data output. The NovaSeq X platform drives the economics: ~$200 genomes at scale, with clinical consumables growing 20%+ outside China. Roche Axelios 1 (SBX chemistry) launched June 2026 as the first credible high-throughput threat to Illumina in years.