ABCT5103 · Individual Written Assignment · Topic #92
The Role of IP in Diagnostics and Medical Devices Development
First full draft, 8 paragraphs, 1,065 words (limit 1,100). The scaffold labels above each box are the brief — the cream boxes hold the prose. Rewrite in your own words before submission.
0Frame: title, thesis, budget
Every paragraph below must be traceable back to the thesis sentence. If a paragraph cannot be linked to it, cut it.
Working title
From Service Bench to Platform Owner: Why Claims, Not Read Length, Decide Who Captures Value in NGS-Based Diagnostics
Core thesis
The commercial value of a sequencing-based diagnostic is set not by its analytical performance but by how claims are distributed across three layers — instrument → reagent/chemistry → algorithm & interpretation — and whether the layer you operate in can be converted into an excludable asset depends on the regulatory route (RUO → IVD) and on how each jurisdiction treats diagnostic method claims.
Scope line
This essay is about next-generation sequencing (NGS) — the platform, its consumables, and the analysis that turns reads into a clinical statement — and compares the United States, Europe, mainland China and Hong Kong.
| ¶ | Section | Function | Words (actual) |
|---|
| 1 | Introduction | Hook — the machine is a claims stack | 127 |
| 2 | Introduction | Thesis + jurisdiction scope | 125 |
| 3 | 1. Instrument & consumables | Who earns, and why (market evidence) | 127 |
| 4 | 1. Instrument & consumables | What that means for a service laboratory | 136 |
| 5 | 2. Hong Kong test case | Services can be the infringing act | 145 |
| 6 | 3. Jurisdiction | HK re-registration: no PCT, no repair | 129 |
| 7 | 3. Jurisdiction | RUO→IVD + claim eligibility divergence | 137 |
| 8 | Conclusion | Synthesis + forward statement | 139 |
| Total · actual draft | 1,065 |
Rules of the exercise: acceptable range 900–1,100 words (±10%); references do not count toward the limit.
Deliverable: file upload, pdf or docx, Calibri 12, double-line spacing, one consistent citation style.
Deadline: 9 Nov, 23:59 — the page also prints “2027”, which conflicts with its own availability window and should be treated as a typo (confirm in writing).
1Introduction
Two paragraphs only. Job: make the reader accept one decomposition and one sentence of argument before any law appears.
¶1 — Decompose the object (hook)
A sequencer arrives as a single box, but it is sold, and owned, as three separable layers. Layer A is the instrument itself: the flow cell, the optics, the fluidics, the stage that holds temperature to a fraction of a degree. Layer B is the consumable chemistry that makes the box do anything at all — reversible terminators and polymerase for sequencing by synthesis, antibody-based recognition on an array, the SMRTbell template or the motor protein that threads a strand through a nanopore. Layer C is the software and the interpretation: basecalling, alignment, variant calling, and the logic that turns a variant into a clinical statement. Each layer can be claimed separately, each has a different competitive lifetime, and each belongs, in practice, to a different owner.
¶2 — State the thesis and scope
This essay argues that the commercial value of a sequencing-based diagnostic is set not by its analytical performance but by how claims are distributed across those three layers, and that whether the layer you occupy can be converted into an excludable asset depends on two conditions: the regulatory route the test travels, and the way each jurisdiction treats diagnostic method claims. The scope is deliberately narrow — next-generation sequencing, its consumables, and the analysis that converts reads into a clinical statement — and the comparison runs across the United States, Europe, mainland China and Hong Kong. The claim is tested in that order. First, who earns at the top of the stack; then, what that arrangement leaves for a laboratory that owns none of it.
Why here: the introduction does two jobs in fixed order — first make the reader accept the object's anatomy (no argument to resist), then hand them the single sentence they will spend the rest of the essay testing. Putting the jurisdictions up front also signals to the marker that every later comparison is inside the declared scope, not drifting.
2Layers A and B: the instrument and its consumables
Turn anatomy into economics before turning economics into law. The Hong Kong case lands harder once the reader already knows who earns what.
¶3 — Where the money sits (evidence paragraph)
The structure is not subtle. In 2019 the global NGS upstream market was worth about US$4.138 billion, and Illumina took roughly 74.1% of it — some US$3.068 billion. Thermo Fisher held 13.6%, about US$563 million, and every other supplier together, MGI included, accounted for about 12.3%; MGI's own instrument revenue that year, ¥1.001 billion, converts to roughly US$145 million, or about 3.5% of the market (Grand View Research, 2020, as cited in MGI's IPO documents). The revenue mix inside each vendor explains the concentration. Instruments are sold once, at a margin thinned by competition; consumables are bought every week for the life of the machine. Patents at Layers A and B are therefore not a reward for invention but the mechanism by which a customer is kept.
¶4 — Consequence for a service laboratory
A laboratory that sells sequencing as a service sits underneath that arrangement and cannot negotiate with it. It buys instruments from several vendors, so its capital plan is somebody else's roadmap; it buys reagents at list price, so its gross margin is set upstream; and its ability to launch a test depends on whether the chemistry it needs is licensed for that particular use. Supply interruption is not a commercial risk it can hedge — it is a decision taken elsewhere. The only layer the laboratory genuinely produces is Layer C: the workflow it has validated, the thresholds it has chosen, the report that a clinician actually reads. Whether that layer is a real asset or merely labour is the question the next section answers, and in Hong Kong the answer has already been given once.
Why here: sequence is deliberately money → law. If the case came first, the same facts read as a dispute between two vendors; after ¶3–4 they read as the operating condition of everyone downstream. The hanging question at the end of ¶4 is the hinge — it converts description into suspense and buys the transition into Section 3 without a bridging sentence.
3The Hong Kong test case
One paragraph, but it carries the essay's most original claim: running the service is itself an act within the reach of the claim.
¶5 — Evidence that services are inside the claim
In August 2020 Illumina Cambridge Limited sued MGI and its BGI affiliates in the Hong Kong Court of First Instance on Hong Kong standard patent HK1253509, whose application was published in June 2019 and whose grant was published on 7 August 2020. The accused subject matter was not confined to hardware. It covered the MGISEQ-2000, the DNBSEQ-G400, G50, T7, Tx and E series, the BGISEQ-500, their reagents, and related services — and the defendants included an entity whose business was sequencing services rather than instrument manufacture. The relief sought was a declaration of infringement, damages, an account of profits and an interim injunction (Shenzhen Stock Exchange disclosure, 18 August 2020). The inference is the one that matters to anyone in this room: not building the machine does not put a service provider in a safe harbour. Running the workflow can itself be the infringing act.
Why here: this is the only paragraph most classmates could not write from general reading, so it sits at the centre of the essay rather than in an appendix. It is also the load-bearing bridge: it answers Section 2's question and simultaneously raises Section 3's (why Hong Kong at all?). One venue, argued once, doing both jobs — which is why Section 3 can then stay short.
4Why jurisdiction decides whether Layer C becomes an asset
Two moves: can you obtain anything here (¶6), and is it worth having / how must it be written (¶7).
¶6 — Hong Kong: re-registration, not PCT
The venue also explains the instrument. A Hong Kong standard patent is not examined locally in the ordinary sense: it is a re-registration, available only by extending a grant that has not yet entered into force elsewhere — from CNIPA, or from a designated European application through the United Kingdom route. Hong Kong is not a member of the Patent Cooperation Treaty, so an international application cannot take effect in the territory at all; there is no local file to open, amend or rescue. The consequence is temporal rather than doctrinal. The substantive life of the Hong Kong right is decided upstream, by the quality and the timing of the designated application. Once that filing is mistimed, or lost to an earlier publication, no local proceeding can repair it.
¶7 — Regulatory route and claim eligibility
Filing is one condition; what may be sold under the filing is the other. A test offered for research use only can be described in claims that would never survive a regulatory submission, whereas a test that must be cleared as an in vitro diagnostic has to be written around a fixed intended use — and that intended use is what gives the claim its value. Eligibility then diverges. In the United States the Federal Circuit held in Ariosa Diagnostics v Sequenom (788 F.3d 1371, 2015) that claims to detecting paternally inherited cell-free fetal DNA were ineligible under §101, because amplifying and detecting were routine and conventional. In Europe, Article 53(c) EPC excludes diagnostic methods practised on the human body, but under G 1/04 a single step performed off the body takes the claim outside that exclusion.
Why here: ¶6 answers availability and ¶7 answers value-and-drafting; reversing them would leave the reader asking "why does drafting matter if I cannot file?". Keeping both after the Hong Kong case means the abstract divergence (US §101 vs EPC Art. 53(c)) is anchored to a venue already introduced, so the essay never opens a second example set.
5Conclusion
One paragraph. Restate in different words, do not repeat ¶2 verbatim, and do not introduce new evidence.
¶8 — Synthesis and one forward statement
The pattern holds across all three layers. A service organisation's ceiling is set by claims it does not own: the instrument and the chemistry belong to the vendor, and access is rented rather than held. Layer C is the only asset such an organisation can build for itself, and it becomes an asset only when two things are true at once — the test crosses from research use into a cleared diagnostic, and the claim is drafted for the jurisdiction that will examine it. Recognition without a claim is expertise, and expertise leaves with the staff who hold it. That is the structural reason the industry consolidates upward, and the reason a career in diagnostics arrives at the same question with which this essay opened: which layer do you want to be standing on when the claims are counted?
Why here: the conclusion is written last but designed first — it must be answerable to every earlier paragraph, which is why it carries the three moves in fixed order (fact → condition → consequence). Restricting the first-person line to one sentence protects the academic register while keeping the argument yours; option (b) is the safe swap if you learn the markers are doctrinal.
6Reference slots & self-check
Fact slots that need a citation — fill the source line before submitting
| Fact used | Citation slot |
|---|
| Market shares (US$4.138 bn; 74.1% / 13.6% / 12.3%) | Grand View Research (2020), cited in MGI STAR Market IPO documents — add your own locator |
| Illumina Cambridge v MGI/BGI affiliates, HK CFI, Aug 2020; HK1253509 | Shenzhen Stock Exchange disclosure filing, 18 Aug 2020 — add URL / accession |
| Outcome status (MGI bound by undertakings/settlements incl. Hong Kong) | CITIC Securities 2022 continuous supervision report — add page |
| US: Ariosa v. Sequenom, 788 F.3d 1371 (Fed. Cir. 2015) | Reporter citation — verify against official judgment text |
| EU: EPC Art. 53(c); G 1/04 | EPO Official Journal reference for G 1/04 — add |
| China: anything you assert about CNIPA practice | Only write what you can source — otherwise leave it out |
Before you submit — four checks.
- Elimination test: if a paragraph would still be true for "any medical device", it is a wasted paragraph. Every block must be demonstrably about NGS.
- HK is not a PCT member. Writing that a PCT application covers Hong Kong is a fatal error.
- Count in the final layout (Calibri 12, double spaced) before polishing prose; references do not count.
- One citation style throughout, and every fact above has a live locator.