US 5202231

I'll search for the specific patent number as instructed.

Let me check for litigation/docket activity and confirm the bibliographic details.

Search method and limits

I ran literal-string searches on identifiers 5202231, US5202231, and "5,202,231" against patent databases (Google Patents/USPTO data, Justia, EPO Global Patent Index/register, PlainSite) and against Federal Circuit / district-court docket sources (govinfo, archive.org court-records collections, fedcircuitblog, E.D. Tex. docket records). I did not find any CAFC docket in 2026 (or any year) citing U.S. Patent No. 5,202,231. I am flagging that as an explicit negative result, not an assertion that no such case exists. I also did not auto-correct the number.

---

US 5,202,231 — bibliographic data

Prosecution relationship: The application is a continuation of Ser. No. 07/175,088, filed 1988-03-30, abandoned. Continuation-in-family members sharing the 1987-04-01 priority include US 5,492,806, US 5,667,972, US 6,018,041, and US 6,451,996 (all now expired).

Uncertainty flag: the Google Patents "Related Parent Applications" table lists an entry US17508898A — Continuation — 1998-03-16 / 1998-03-30. The number as rendered (17508898 with a 1998 filing date) is internally inconsistent and I cannot reconcile it to a real application number; treat it as a source-data anomaly rather than a genuine record. It is not the parent of the '231 patent.

---

Abstract (as issued)

The stated conditions under which oligonucleotides hybridize only with entirely homologous sequences are recognized; the sequence of a DNA fragment is read by hybridization and assembly of positively hybridizing probes through overlapping portions. Simultaneous hybridization of DNA molecules applied as dots and bound to a filter — representing single-stranded phage vector with cloned insert — with about 50,000–100,000 groups of probes of the main type (A,T,C,G)(A,T,C,G)N8(A,T,C,G) yields data for computer determination of a DNA sequence of mammalian-genome complexity in one step. Three M13 libraries are used to obtain a maximally completed sequence: 0.5 kb inserts, 7 kbp inserts consisting of two sequences averaging 100 kbp apart in genomic DNA, and "jumping" subclones. Per 1,000,000 bp: 25,000 subclones of 0.5 kb, 700 subclones of 7 kb, and 170 jumping subclones; the 0.5 kb subclones are applied in pools of 20, giving 2,120 samples per million bp. The process is described as readily robotizable for factory-scale reading.

---

The single independent claim, in plain language

Claim 1 — A three-step method of sequencing a nucleic acid fragment:

• (a) Hybridize. Contact the fragment with a set of oligonucleotide probes of predetermined sequence and length, under hybridization conditions stringent enough to distinguish (i) probes that are exactly complementary to a part of the fragment from (ii) probes that are not exactly complementary. The "predetermined" wording means the probe sequences are known in advance (a defined library/array), and the conditions must resolve perfect matches from single-base mismatches.
• (b) Detect. Identify which probes in the set actually bound as exact complements.
• (c) Assemble. Determine the fragment's sequence from that detected subset by compiling their overlapping sequences — i.e., infer the sequence computationally by aligning probes that overlap one another, rather than by gel-based readout.

Practical significance: Claim 1 is a broad, function-level claim to what the field now calls sequencing by hybridization (SBH). It recites no particular probe length, label, substrate, filter, or algorithm — the only structural limitations are "a set of oligonucleotide probes of predetermined sequence and length" and "hybridization conditions that allow differentiation between exact and non-exact complements." That breadth is why the patent is treated as a foundational document for hybridization-based sequencing and array technology.

Dependent claims (not independent, summarized for completeness):

• Claim 2 — the probes are 8-mers to 20-mers in length.
• Claim 3 — the probes are 11-mers. (The specification explains that 11-mers were regarded as the shortest oligonucleotides that hybridize reliably, and that full 11-mer coverage would require 4¹¹ = 4,194,304 probes, hence the grouped-probe scheme.)
• Claim 4 — depends on claim 2 and recites that each probe is an 11-mer containing a unique 8-mer, characterized as (N2)N8(N1) or (N1)N8(N2), where N2 is two unspecified nucleotides at either terminus, N1 is one unspecified nucleotide at the other terminus, and N8 is the unique 8-mer. This is the "ordered ONP group" scheme from the specification: 64 11-mers share one 8-mer, collapsing ~4 million probes into ~65,000 informative groups.

Analytical note: claim 4's dependency on claim 2 is textually awkward, since claim 2 covers 8-mers through 20-mers while claim 4 requires an 11-mer with an internal 8-mer. This is an internal-consistency quirk of the issued claim set, not a correction I have made.

---

Litigation / CAFC status

• Google Patents flags this family (family ID 27390498) as "Family has litigation — First worldwide family litigation filed," linking to a Darts-IP family record. The fetched page supplies no case name, court, or docket number, so I cannot characterize the litigation from that source. PlainSite shows Hyseq Inc. as a recorded assignee of 5,202,231 as of 1993-11-29, with attorney Thomas C. Meyers of Marshall, O'Toole, Gerstein, Murray & Borun listed as correspondent counsel.
• No 2026 CAFC docket referencing 5,202,231 was found in my searches. Given the recorded legal status is "Expired – Lifetime" with an anticipated expiration of 2010-04-13, a 2026 Federal Circuit appeal on this patent's claims is unlikely absent a collateral or historical-citation context. I do not have authoritative information confirming any such 2026 matter, and I am not asserting one exists — my search returned nothing, which is a null result, not proof of absence.
• The patent is heavily cited as a prior-art/reference document in later genomic-sequencing patents (471 "Cited By" entries on the Google Patents page), which is why it surfaces in registers such as EPO's even where no live case is pending.

---

Confidence assessment

• High confidence: title, inventors, application number, filing date, issue date, priority date, claim count and text, abstract, assignee-at-issue and the 1993 Hyseq assignment, expired status. These are grounded in the full patent text and corroborated by Google Patents, Justia, EPO GPI, and secondary literature (e.g., Encyclopedia of Analytical Chemistry citing "US Patent 5202231, 1993" for Drmanac & Crkvenjakov).
• Medium confidence: post-2004 ownership. The reassignment table mixes a Hyseq change-of-name chain, a 2004 security agreement naming NUVELO/AFFYMETRIX, and Callida Genomics as assignor, so "current assignee" is genuinely ambiguous from the available record.
• Low / explicit uncertainty: the US17508898A parent-application entry (apparent source-data error), and the existence of any 2026 CAFC proceeding on this patent (no results found).

I'll search patent litigation sources for this specific patent number.

Let me search more targeted litigation sources.

I found strong leads. Let me dig into case numbers, dates, and outcomes.

Litigation involving US 5,202,231

I searched litigation-oriented sources (Google Patents litigation metadata, the Affymetrix 8‑K settlement disclosure, SEC filings, CourtListener, and trade press). I found one known litigation campaign involving US 5,202,231 — the Hyseq/Affymetrix "sequencing by hybridization" (SBH) dispute, plus a related follow‑on case. No other litigation asserting or challenging US 5,202,231 surfaced.

The known cases

Patents asserted:

• Hyseq asserted US 5,202,231, US 5,525,464, US 5,695,940, US 6,018,041 and US 5,972,619.
• Affymetrix asserted US 5,795,716, US 5,744,305 and US 5,800,992.

Settlement (outcome): On October 24, 2001, Affymetrix and Hyseq entered a Settlement Agreement providing for comprehensive settlement of "all existing litigation between the two companies that began in March 1997." Key terms disclosed in Affymetrix's Form 8‑K:

• Dismissal with prejudice of all three pending suits listed above;
• The parties acknowledged the validity and enforceability of the patents asserted, expressly including Hyseq's US 5,202,231;
• Hyseq agreed to enter an abandonment of contest in the pending interference ("Chee v. …");
• Cross-license/contract relationship followed — Hyseq's 2001 10‑K reports $0.8 million in contract revenue from Affymetrix in 2001.

Related decision: Affymetrix, Inc. v. Hyseq, Inc., 132 F. Supp. 2d 1212 (N.D. Cal. 2001) — a Markman/claim-construction ruling. Note this ruling construed claim terms of the Affymetrix patents ('716, '934, '305, '992) at issue in the counterclaims; I did not find a claim-construction ruling specifically construing the claims of the '231 patent. A later order clarifying three Markman terms was reported in the trade press.

Corroboration of a litigation family

Google Patents' page for US 5,202,231 carries the Darts‑ip indicator "Family has litigation — First worldwide family litigation filed" (https://patents.google.com/patent/US5202231/en), consistent with the Hyseq/Affymetrix campaign.

Related but distinct proceeding (not US litigation over the '231 patent)

Hyseq and Affymetrix (with other companies) jointly filed an opposition at the European Patent Office against Ed Southern's sequencing-by-hybridization patent (Oxford University/Oxford Gene Technology). That is an EPO administrative proceeding directed at Southern's patent, not a US court case involving US 5,202,231 — I mention it only because it is often conflated with the Hyseq/Affymetrix dispute.

Important caveats

• Strict-identifier note: I searched specifically for the literal number 5202231. I did not find any litigation tied to a different/similar number, and I am not substituting any other patent.
• Source limitations: My searches did not return an exhaustive docket-level record (e.g., Darts‑ip full litigation table, PACER docket sheets, or the Unified Patents litigation caselist for this number). Consequently:
  • The precise filing date of case C 97-20188 RMW is not confirmed from a primary docket source; the March 4, 1997 date comes from a Nature Biotechnology news report (May 1997). Trade press also referred to "two 1997 suits," which I could not reconcile to the three case numbers in the 2001 8‑K — so there may be an additional 1997 filing that is not in my results. I flag this as unverified rather than asserting it.
  • I cannot rule out later, non-Hyseq assertions (e.g., by assignee-plaintiffs after the Hyseq→Nuvelo/Callida lineage) that may not surface in the sources searched.
• Why no current litigation is expected: Google Patents lists the patent as Expired – Lifetime, with anticipated expiration April 13, 2010. As a matter of the patent's term, US 5,202,231 cannot be the subject of new infringement litigation today; the cases above are historical.

Primary sources

• Affymetrix Form 8‑K (Nov. 1, 2001) disclosing the Oct. 24, 2001 settlement and the three case numbers: https://capedge.com/filing/[913077](/patent/913077)/0000912057-01-[537335](/patent/537335)/AFFX-8K
• Hyseq 10‑K (FY2001), Item 3 Legal Proceedings: https://www.sec.gov/Archives/edgar/data/[907654](/patent/907654)/000089161802001598/f79510e10-k.htm
• Nature Biotechnology 15, 406 (May 1997), "Litigation escalates for patents on a chip": https://www.nature.com/articles/nbt0597-406a.pdf
• BioCentury reports on the suits and Markman clarification: https://www.biocentury.com/article/[141953](/patent/141953)/ ; https://www.biocentury.com/article/56155/
• Affymetrix, Inc. v. Hyseq, Inc., 132 F. Supp. 2d 1212 (N.D. Cal. 2001): https://www.courtlistener.com/opinion/[2457323](/patent/2457323)/affymetrix-inc-v-hyseq-inc/
• Google Patents patent page (litigation indicator): https://patents.google.com/patent/US5202231/en

Bottom line: The known litigation over US 5,202,231 is the Hyseq v. Affymetrix SBH dispute (N.D. Cal., cases C 97‑20188 RMW, C 99‑21163 JF, C 00‑20050 RMW), which began in March 1997 and ended in a comprehensive October 24, 2001 settlement with dismissal with prejudice and an acknowledgment of the '231 patent's validity and enforceability. I found no other litigation on this patent number.

I'll start by checking the structured ODP block against live sources, then probe for any PTAB or related proceedings.

No PTAB hits yet. Let me probe more specifically and check for other post-issuance proceedings (reexam) and the litigation history.

Confirming no AIA trials and checking the other post-issuance/litigation trail.

Proceedings overview

Total AIA trial proceedings on file: 0. The canonical ODP-derived block reports no IPR, PGR, or CBM for US 5,202,231, and my independent searches (PTAB E2E / PTACTS, Google Patents PTAB tab, CourtListener, and general web) surfaced no AIA trial petition, no institution decision, no Final Written Decision, and no settlement or termination of any PTAB proceeding involving this patent or its family members. Breakdown is therefore 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denied.

Bottom-line defensive posture: this is not "the patent survived IPRs and is hardened," and it is not "claims 1–5 were canceled." It is a third state: the '231 patent was never tested at the PTAB at all, and by the time AIA trials existed (2012-09-16) the patent had already expired (2010-04-13). No § 315(e)(2) estoppel attaches to anyone, so every prior-art ground remains formally available — but the patent's 20-year term ran out sixteen years ago, and the only public assertion campaign (Hyseq v. Affymetrix, N.D. Cal.) was comprehensively settled in October 2001. A demand letter citing the '231 today is asserting an expired patent whose last reported infringement claim was resolved by settlement a quarter-century ago.

---

Per-proceeding detail

Not applicable — there are no proceedings to tabulate. In place of the per-proceeding template, here is the verified-absence record and the substitute defensive evidence a defendant should actually rely on.

Verified absence of AIA trial activity

• Sources checked: the structured PTAB block from USPTO ODP (authoritative, reports zero proceedings); web search for "US5202231 IPR," "'5202231' IPR petition Patent Trial and Appeal Board," and AIA-trial-indexed sources. No petitioner name, proceeding number, or Board paper referencing US 5,202,231 appeared.
• Family check: Google Patents lists the five-member family — US 5,202,231; US 5,492,806; US 5,667,972; US 6,018,041; US 6,451,996 (family ID 27390498) — all with the same 1987-04-01 priority and all marked Expired / Expired - Fee Related. No PTAB proceeding surfaced on any family member either, though I could not exhaustively confirm this for each sibling from the sources searched; treat sibling status as unverified rather than confirmed-negative.
• Timing reality: IPR/PGR became available 2012-09-16, CBM 2012-09-16 (sunset 2020-09-16). The '231 anticipated expiration is 2010-04-13 and Google Patents records legal status as Expired - Lifetime. The patent was already off-term when the first AIA petition could have been filed, which explains the absence better than any inference about patent strength.
• Reexamination caveat: I found no ex parte or inter partes reexamination of the '231. I cannot rule one out entirely from the searches run; if you need a negative on the post-issuance review history, pull the '231's full transaction history in Patent Center / the file wrapper (https://patents.google.com/patent/US5202231/en → "USPTO PatentCenter"). Flagging this as unverified rather than asserted.

The proceeding that does exist — and it isn't a PTAB one: Hyseq, Inc. v. Affymetrix, Inc. (N.D. Cal.)

This is the litigation Darts-IP flags on the Google Patents record ("Family has litigation," family=27390498). It is district court, not an AIA trial.

• Court: U.S. District Court for the Northern District of California, Judge Ronald M. Whyte.
• Filed: March 1997 (reported as 1997-03-03 by one secondary source and 1997-03-04 by Nature Biotechnology; the exact day differs by source — verify against the docket before citing a specific date).
• Patents asserted by Hyseq: US 5,202,231, US 5,525,464, and US 5,695,940 (the '231 plus its Lipshutz-era SBH siblings).
• Accused product: Affymetrix GeneChip system; Affymetrix countered that its use was mutation/sequence checking, not de novo sequencing. Affymetrix also filed its own suits against Hyseq and Incyte on the '934/'305/'992/'716 patents (see Affymetrix, Inc. v. Hyseq, Inc., N.D. Cal. claim-construction opinion).
• Claim construction — the high-water mark for a defendant: Judge Whyte issued the first Markman ruling 1999-10-26 and a revised Markman ruling in August 2000 that Affymetrix publicly characterized as narrowing the '231. The revised construction stated that determining that a fragment contains a known sequence, or determining differences from a known sequence, "is not sequencing as defined by the '231 patent" (Silicon Investor/PRNewswire release of 2000-08-02, reporting Affymetrix's announcement). That ruling is the single most useful non-PTAB claim-scope document for anyone reading claims 1–4 narrowly. I did not locate the Markman order itself in full text; treat the quoted language as a party's press release reporting the ruling, not the order.
• Disposition: comprehensive settlement, October 2001, not an invalidity judgment. Hyseq's FY2001 Form 10-K discloses that "In October 2001, the Company entered into a settlement agreement with Affymetrix providing for the comprehensive settlement of a [dispute]," and its exhibit index lists both an Interference Settlement Agreement dated 2001-10-24 and a Settlement Agreement dated 2001-10-24 between Hyseq and Affymetrix. Public reporting describes the outcome as cross-licensing arrangements among Hyseq, Affymetrix, Callida Genomics, and N-Mer. The financial terms are not public — say so; do not characterize them.
• Appeal: no Federal Circuit appeal on the '231 appears in the record. The case settled rather than going to judgment, so there was no validity or infringement judgment to appeal. The only CAFC-adjacent event in this space is the Markman/substantial complementary line of cases on the Affymetrix patents, which does not involve the '231.
• Defensive value: the '231 never obtained a merits validity ruling in either direction. Its practical scope was constrained by the Whyte revised Markman construction, and the litigation ended in confidential settlement, so there is no stare decisis protection and no estoppel of any kind — but also no adjudicated validity that a plaintiff could point to.

---

Strategic summary

Claim status of US 5,202,231. Claims 1–4 are neither CANCELED nor SUSTAINED — they are UNTESTED by any PTAB panel and EXPIRED as of the anticipated expiration date of 2010-04-13. Claim 1 (the sequencing-by-hybridization method), claim 2 (8-mers to 20-mers), claim 3 (11-mers), and claim 4 (the (N2)N8(N1)/(N1)N8(N2) grouped-probe limitation) all lapsed with the term. There is no IPR certificate, no reexamination certificate, and no final judgment of invalidity narrowing any of them. Any statement that a claim of the '231 "was invalidated" would be a fabrication.

Estoppel landscape. Because no IPR or PGR was ever instituted on this patent, § 315(e)(2) estoppel is empty. No petitioner and no privy is barred from raising any § 102/§ 103 ground — before the Board, in district court, in the ITC, or in an ex parte reexamination. The full universe of prior art remains available to a party that needs to attack validity in any forum that will still hear the issue. That said, this is a freedom-to-operate curiosity rather than a defensive asset: with the term expired since 2010-04-13, an IPR would be an academic exercise (the Board can review expired patents, but there is no amendment-possible claim scope and no injunction to defeat), and any hypothetical infringement claim would run into the § 286 six-year damages bar, which reaches only acts after 2020 — i.e., after expiration.

Pattern signals. (1) No repeat-petitioner pattern — there is no petitioner at all. (2) No defensive aggregator — nothing surfaced linking Unified Patents, RPX, or any similar entity to the '231. (3) Patent owner litigation posture was historically aggressive but confined to the 1997–2001 window: Hyseq asserted the SBH family against Affymetrix, opposed Southern's competing European SBH patent alongside six other companies, and litigated with Incyte and others; the Hyseq estate later became Nuvelo, then ARCA Biopharma (via Variagenics and Nuvelo mergers). Current assignee of record per Google Patents is Hyseq Inc., with a security-interest reassignment to NUVELO, INC. and AFFYMETRIX, INC. dated 2004-12-14 (assignor Callida Genomics). (4) No PTAB appeal pattern — there is nothing to appeal.

---

Recommended next steps

1. State the negative plainly to your client. There is no PTAB proceeding to distinguish on US 5,202,231. Do not let anyone argue "the patent survived IPRs" — nobody ever filed one. The defensible framing is: never PTAB-tested, expired 2010-04-13, last assertion settled in 2001.
2. Anchor a scope argument in the Whyte revised Markman ruling (N.D. Cal., revised August 2000; first ruling 1999-10-26). If a demand letter reads claim 1 broadly onto sequence checking, mutation detection, or comparison against a known sequence, the reported construction language — that determining known-sequence content or differences from a known sequence "is not sequencing as defined by the '231 patent" — is directly on point. Pull the actual Markman order from the N.D. Cal. docket before quoting it; the quote above comes from a party press release, not the order.
3. Obtain the file wrapper / transaction history for the '231 in Patent Center and confirm (a) no ex parte or inter partes reexamination, (b) no terminal disclaimer or certificate of correction affecting claims 1–4, and (c) the exact expiration computation. Link: https://patents.google.com/patent/US5202231/en → USPTO PatentCenter.
4. Check the siblings if you actually need a defensive posture. US 5,492,806, US 5,667,972, US 6,018,041, and US 6,451,996 share the 1987-04-01 priority and are all recorded Expired / Expired - Fee Related. All post-date the '231's expiration if they expired later (the '996 patent, filed 2000-01-21, would run to roughly 2017–2018), so if a plaintiff is asserting the family rather than the '231, that is the patent to worry about — and it is also the one where an IPR history would matter. Confirm each sibling's PTAB status individually; I could not verify it for the siblings.
5. Correct the record if anyone cites a PTAB proceeding number. Any proceeding number offered for US 5,202,231 should be treated as an error or as belonging to an unrelated patent. Verify anything offered against PTAB E2E (https://ptacts.uspto.gov/ptacts/) and the USPTO PTAB decisions library (https://www.uspto.gov/patents/patent-trial-and-appeal-board/ptab-decisions) before repeating it.

Sources consulted: the structured PTAB-proceedings block (USPTO ODP ingest, authoritative); https://patents.google.com/patent/US5202231/en (family, priority, expiration, Darts-IP litigation flag); Affymetrix, Inc. v. Hyseq, Inc. claim construction opinion, https://www.courtlistener.com/opinion/[2457323](/patent/2457323)/affymetrix-inc-v-hyseq-inc/; Hyseq FY2001 Form 10-K (settlement and October 2001 agreements), http://getfilings.com/o0000891618-02-001598.html; Affymetrix/Hyseq 2000-08-02 revised-Markman announcement as republished at https://www.siliconinvestor.com/readmsg.aspx?msgid=14148993; W. Bains, "Litigation escalates for patents on a chip," Nature Biotechnology (1997), https://www.nature.com/articles/nbt0597-406a.pdf.

I'll start by pulling the assignment and litigation records from the web, since the Google Patents legal-events list gives conveyance types and dates but no reel/frame numbers.

Useful lead: a correspondent attorney (Thomas C. Meyers) is indexed for the 1993 Hyseq recording. Let me dig into the assignment correspondents, the security agreement, and the litigation record.

Assignment & Ownership Reconstruction — US 5,202,231 (Drmanac & Crkvenjakov, "Method of sequencing of genomes by hybridization of oligonucleotide probes")

Source-access caveat (read first). My only retrieval channel for this task was open web search; I could not execute a live query against assignmentcenter.uspto.gov / assignment.uspto.gov during this session. The ownership events below are taken from the authoritative patent text you supplied (Google Patents "Legal Events" for US5202231) plus independent corroboration found in search (SEC filings, trade press, PlainSite assignment-correspondent index). I therefore cannot supply the reel/frame numbers your template asks for, and I will not invent them. Every entry below is marked with how the date is known. Reel/frame, execution date vs. recording date, and correspondent-of-record address should be verified directly at https://assignmentcenter.uspto.gov/ (search 5202231) and at https://patents.google.com/patent/US5202231/en.

---

Inventors

Pattern note — departures: Both inventors left the originating institute before the patent issued (1993-04-13); Drmanac was at Argonne by 1991 and co-founded Hyseq in 1994. This is the classic academic-institute → US-startup technology-transfer pattern, not the "all inventors flee the assignee within 12 months, then portfolio fire-sale" pattern. The institute's interest was formally assigned to Hyseq on 1993-11-29, i.e., the technology followed the inventors to a company Drmanac co-founded. There is no evidence of inventor estrangement, abandonment, or a distressed sale.

---

Original assignee

• As issued (1993-04-13): no corporate assignee named on the face of the patent. Google Patents records the "Original Assignee" field as "Individual" — i.e., the named inventors/applicants held it, with IMGGI asserting its institutional interest shortly thereafter.
• First recorded corporate owner: Hyseq, Inc. (Illinois corporation, then Nevada corporation after the 1993-11-29 merger), recorded 1993-11-29.
• Business: Hyseq was a Sunnyvale, CA genomics company built on the inventors' sequencing-by-hybridization (SBH) chemistry. It was a public company (Nasdaq: HYSQ), not a pure holding vehicle — it performed gene discovery, held SBH patents, and formed the subsidiary Callida Genomics (2001) to commercialise SBH arrays/tools. It also carried a drug-discovery portfolio (the later Nuvelo–Amgen alfimeprase collaboration).
• Product embodying the claims? Cannot be confirmed. Hyseq's revenue in the relevant period was driven by licences, collaborations and settlements (e.g., the Affymetrix settlement of 2001-10 and an Applied Biosystems agreement), and it explicitly described competing on SBH technology through its subsidiary Callida. I found no evidence of a Hyseq-branded sequencing instrument or kit practising claim 1 (11-mer hybridisation + overlap assembly) in commerce.
• Current status: Hyseq → renamed Nuvelo, Inc. (announced 2003-01-28; effective 2003-02-03, after merger with Variagenics) → Nuvelo merged into ARCA biopharma, Inc. (2009-01-27). The SEC registrant with CIK 907654 is the same registrant for Hyseq, Nuvelo and ARCA; filings under that CIK are today made by Oruka Therapeutics, Inc. (consistent with the ARCA–Oruka combination). Hedge: the Oruka attribution comes from third-party filing aggregators, not from a primary document I read directly. Google Patents' "Current Assignee: Hyseq Inc" is a stale indexing artifact — Hyseq ceased to exist under that name in 2003.

---

Assignment timeline

Important: Google Patents' legal-events list gives conveyance type, assignor, assignee and a single date but does not expose reel/frame numbers, execution dates, or correspondents. I have substituted the best available corroboration and flagged the gaps rather than filling them in. Patent expired 2010-04-13 (17-year term from grant; pre-1998 term rule), so no assignment can postdate that as a live asset.

1. 1993-11-29 (recorded/effective as indexed) — Reel/frame: NOT RETRIEVABLE

• Conveyance: Assignment of assignors' interest
• Assignor: INSTITUTE OF MOLECULAR GENTICS AND GENETIC ENGINEERING (spelling per record)
• Assignee: HYSEQ, INC.
• Correspondent: Thomas C. Meyers — PlainSite indexes him as correspondent-of-record against the 1993-11-29 Hyseq recording for patent 5202231, and indexes him under the firm slug Marshall, O'Toole, Gerstein, Murray & Borun (a Chicago IP firm). Address as recorded: not retrievable. Recurrence flag: Meyers also appears as correspondent for Hyseq recordings on 6/22/1995 for US 6,270,961 / application 08/353,554 and in the 1993 Hyseq block for application 07/723,712 — i.e., he is a recurring Hyseq-family recording attorney, but I found no evidence he appears on any Unified Patents / RPX / Patent Progress NPE assertion list.
• Context: Institutional technology transfer — the Belgrade institute assigns its interest to the startup co-founded by the inventor.

2. 1993-11-29 — Reel/frame: NOT RETRIEVABLE

• Conveyance: Merger (see document for details)
• Assignor: HYSEQ, INC., AN ILLINOIS CORPORATION
• Assignee: HYSEQ, INC., A NEVADA CORPORATION
• Correspondent: Thomas C. Meyers (per the same 1993-11-29 Hyseq recording block); address not retrievable. Recurrence flag: same attorney as entry 1 — second appearance in this chain.
• Context: Internal corporate reorganisation — Illinois parent merged into Nevada entity on the same day as the incoming institutional assignment.

3. 1993-11-29 — Reel/frame: NOT RETRIEVABLE

• Conveyance: Merger (a second, separately indexed recording of the same event — duplicate indexing or a second property; the Google Patents event list shows the merger twice with identical parties)
• Assignor: HYSEQ, INC., AN ILLINOIS CORPORATION
• Assignee: HYSEQ INC., A NEVADA CORPORATION
• Correspondent: not retrievable; likely Meyers as above.
• Context: Internal reorganisation (duplicate of entry 2).

4. 1996-10-21 — Reel/frame: NOT RETRIEVABLE

• Conveyance: Change of name
• Assignor: SBHG CORPORATION
• Assignee: HYSEQ, INC.
• Correspondent: not retrievable in my sources.
• Context: Change of name only — SBHG Corporation renamed itself Hyseq, Inc., absorbing the surviving corporate identity.

5. 1996-11-20 — Reel/frame: NOT RETRIEVABLE

• Conveyance: Change of name (second, separately indexed recording of the same name-change event, ~30 days after entry 4)
• Assignor: SBHG CORPORATION
• Assignee: HYSEQ, INC.
• Correspondent: not retrievable.
• Context: Change of name only.

6. 2004-12-14 — Reel/frame: NOT RETRIEVABLE

• Conveyance: Security Agreement (not a title transfer)
• Assignor: CALLIDA GENOMICS, INC.
• Assignee: NUVELO, INC. and AFFYMETRIX, INC. (both named on the recorded event)
• Correspondent: not retrievable in my sources.
• Context: Securitisation — Callida (Hyseq/Nuvelo's SBH subsidiary, co-founded by Drmanac in 2001) pledges its patent family as collateral. This is consistent with the public record of the October 2001 Affymetrix settlement under which Affymetrix obtained a minority equity stake and options in Callida. Callida is the assignor here, which indicates the '231 family passed into or was held within Callida by 2004 — a point worth verifying at the Assignment Center, because the chain of title is not fully legible from the public event list alone.

No later recordings found. The patent expired 2010-04-13.

Chain-of-title gap to flag: the events show IMGGI → Hyseq (1993), Hyseq-internal reorganisations (1993, 1996), and Callida → Nuvelo/Affymetrix security (2004), but not the intermediate instrument by which Hyseq/Nuvelo's title in the '231 reached Callida. There is no recorded transfer out of the Hyseq/Nuvelo family and no NPE vehicle. Treat the Hyseq → Callida leg as unverified.

---

Timeline diagram

timeline
    title Ownership of US 5202231
    1987 : Priority filing in Yugoslavia
    1988 : US parent application filed
    1993 : Patent issued Apr 13
         : IMGGI assigns its interest to Hyseq
         : Hyseq Illinois merges into Hyseq Nevada
    1996 : SBHG Corporation renamed Hyseq Inc
    1997 : Hyseq sues Affymetrix on the 231 patent
    2001 : Affymetrix settlement and cross license
    2004 : Callida grants security interest
    2009 : Nuvelo merged into ARCA biopharma
    2010 : Patent expires Apr 13

---

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. Every recorded counterparty is an operating or institutional entity: IMGGI (a Yugoslav research institute), Hyseq, Inc., SBHG Corporation (rename vehicle), Callida Genomics, Nuvelo, Inc., Affymetrix, Inc. There is no "IP / Holdings / Ventures / Licensing" LLC, no Delaware or Texas single-member vehicle, and no recorded registered-agent address in any event. The 1996-10-21 and 1996-11-20 SBHG → Hyseq recordings are opaque as to what SBHG was, but they are change-of-name conveyances, which by definition do not move title to a new party — that is the opposite of a shell transfer. Note the service-address test could not be applied because correspondent addresses were not retrievable.

2. Known asserter in the chain — NOT PRESENT. None of the listed entities (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities) appears anywhere in the chain. The one security-interest holder that could look like an aggressor, Affymetrix, Inc., is a public operating company (now part of Thermo Fisher) and appears as collateral holder under the 2004-12-14 Security Agreement, not as an assignee asserting the '231 — in the 1997–2001 litigation Affymetrix was the defendant, not the plaintiff. The Unified Patents portal entry for US-5202231-A lists Hyseq Inc as assignee and the patent as prior art; that is an index entry, not evidence of ownership by an NPE.

3. Repeat correspondent across the chain — PRESENT as recurrence, but with no NPE nexus. Thomas C. Meyers (indexed under the firm slug Marshall, O'Toole, Gerstein, Murray & Borun) is the correspondent of record on the 1993-11-29 Hyseq recording for patent 5202231, and reappears on Hyseq recordings dated 6/22/1995 (US 6,270,961 / application 08/353,554) and in the 1993 Hyseq block for application 07/723,712. So the recurrence test is met — but the recurrence is entirely within one operating company's corporate housekeeping, not across chained LLCs, and I found no evidence of Meyers on any NPE assertion list. Under your own precision rule ("the signal is recurrence" plus NPE context), this is a recurrence finding, not an NPE finding.

4. Cascading transfers — NOT PRESENT. The closest thing is three recordings clustered on a single day (1993-11-29: one institutional assignment + two merger recordings) and two name-change recordings 30 days apart (1996-10-21, 1996-11-20). These are corporate reorganisation artifacts, not a chain of distinct LLCs passing the patent in under 24 months. The subsequent gap to 2004-12-14 is eight years, and the final event is a security interest, not a sale.

5. Pre-litigation transfer — UNCLEAR, timing met but substance does not match. The first infringement suit naming this patent was filed 1997-03-03 (Hyseq v. Affymetrix, N.D. Cal.; see below). The last recorded ownership event before it is the 1996-11-20 change of name (SBHG → Hyseq, Inc.) — approximately 3.5 months pre-suit, which satisfies the six-month timing test. But the conveyance is a name change within the same corporate family, not a transfer to an assertion vehicle, and title was already with Hyseq from 1993. I therefore do not score this as a genuine pre-litigation transfer.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 proceeding for Hyseq, Nuvelo, Callida or IMGGI appears in the record I reviewed. Nuvelo's 2009 end was a merger into ARCA biopharma, and Callida's SBH intellectual property was licensed onward (Drmanac states he "exclusively licensed his technologies to Complete Genomics"; Complete Genomics was acquired by BGI in 2013) rather than sold in insolvency. The 2004-12-14 Security Agreement is a secured-financing collateral pledge — securitisation, not a fire-sale.

7. Privateering — NOT PRESENT. Hyseq asserted the '231 in its own name as an operating SBH company against a direct competitor (Affymetrix's GeneChip system). It did not push the patent to an NPE to sue on its behalf. Affymetrix's later relationship with Hyseq/Callida (minority equity stake and options in Callida, plus the 2004 security interest) ran the other direction — a settlement-driven strategic investment, not privateering.

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT Network, Unified Patents or OIN. The patent was never neutralised by a defensive aggregator; it simply expired on 2010-04-13 while still nominally inside the Hyseq/Nuvelo/ARCA lineage.

---

Verdict

Operating-company assertion.

The only assertion of US 5,202,231 in its 23-year life was made by its operating owner, Hyseq, Inc. — a public genomics company (Nasdaq: HYSQ) — against Affymetrix, Inc., a direct competitor whose GeneChip system allegedly used Hyseq's claimed SBH method. The first suit naming the '231 was filed 1997-03-03 in the Northern District of California (with '5,525,464); a second suit followed 1997-12-09 on '5,695,940; the dispute ended in a 2001-10-01 settlement and cross-licence, with Affymetrix taking a minority stake and options in Hyseq's Callida Genomics subsidiary — the same relationship later reflected in the 2004-12-14 Security Agreement recorded against this patent (Callida → Nuvelo/Affymetrix). The chain of title contains no shell LLC, no known NPE asserter, and no defensive aggregator: it is IMGGI → Hyseq (1993-11-29 assignment), a 1993-11-29 Hyseq Illinois→Nevada merger, 1996 SBHG→Hyseq name changes, and the 2004 security interest, after which the patent expired (2010-04-13). The one honest qualification to the label's strict definition is that the current nominal owner (ARCA biopharma / registrant CIK 907654, today Oruka Therapeutics) does not ship an SBH product — the operating-company assertion is historical, not ongoing.

Verification links:

• USPTO Assignment Center (search 5202231): https://assignmentcenter.uspto.gov/
• Legacy assignment search: https://assignment.uspto.gov/patent/index.html
• Google Patents legal events and full text: https://patents.google.com/patent/US5202231/en
• Low-cost PlainSite correspondent index entry (Thomas C. Meyers, 11/29/1993 Hyseq / 5202231): https://www.plainsite.org/attorneys/marshall-otoole-gerstein-murray-and-borun/thomas-c-meyers/

What still needs a manual docket pull at the Assignment Center: (i) reel/frame for all four events (1993-11-29 ×3, 1996 ×2, 2004-12-14 ×1 — note the duplicate merger and name-change indexings); (ii) execution dates vs. recording dates; (iii) the correspondent of record and its address for the 1996 and 2004 events; and (iv) the missing instrument by which Hyseq/Nuvelo's title in the '231 reached Callida Genomics, Inc. before it pledged the patent in December 2004.

I'll search for the specific patent and its cited prior art.

The Google Patents rendering I have doesn't include a front-page "References Cited" list. Let me hunt for the issued front page and other registers.

EveryPatent and uspto.report often reproduce the front-page reference lists. Let me target those.

Result of the search step (with an explicit limitation)

I searched for 5202231, US5202231, 5,202,231, and US 5,202,231 against Google Patents, Espacenet/EPODOC cited-document views, FreePatentsOnline, Justia, EveryPatent and uspto.report. Two findings must be stated up front, before any prior-art analysis:

1. I could not retrieve a front-page "References Cited" (examiner citation) list for US 5,202,231. The authoritative full text you supplied contains the abstract, description, and claims — but no References Cited / U.S. Patent Documents block and no Patent Citations table. Google Patents renders "Cited By (471)" for this patent (documents that cite it), which is a forward citation list, not a list of art cited against it. My search budget was exhausted before I could pull an EveryPatent / uspto.report / patentimages front-page scan. I therefore cannot enumerate "each patent citation for 5202231" from a verified examiner-citation set, and I will not generate one. Anything resembling a fabricated front-page patent list would be worse than an admitted gap.

Note also an internal conflict in the instructions: the task header says current date April 26, 2026, while the fetched patent page is timestamped 2026-09-30. I have used neither date as a substantive input.

2. What is verifiable on the face of the patent is the specification's own "STATE OF THE ART / Prior Art" discussion and its in-text references. That is the only citation set I can ground in the authoritative text. I analyze it below, and separately address the collateral documents I did retrieve, with their citation direction clearly labeled.

---

A. References cited within the '231 specification (verifiable)

These are the only citations appearing on the face of the document I was given. All are non-patent literature; there is one self-citation to a co-pending application.

Bottom line on the facial citations: no reference cited on the face of US 5,202,231 discloses the full three-step combination of claim 1. The reference set is essentially the enabling literature (mismatch discrimination: Wallace 1979; length-independent stringency: Wood 1985; detection sensitivity: Al-Hakim & Hull 1986) plus background on the state of the art. Wallace 1979 + Wood 1985 is the strongest § 103 pair, and even then the missing element — step (c), sequence inference by compiling overlapping probes — would have to come from somewhere else.

One caveat on claim lengths. Whether Wallace 1979 or Wood 1985 disclose probe lengths falling inside claim 2's "8-mers to 20-mers" or claim 3's "11-mers" is something I would need the full texts to confirm. I am not asserting that from memory. Treat table entries 3 and 4 as "length recitation unverified."

---

B. Documents that cite '231 (forward citations — not § 102 art)

These surfaced in my searches and must not be confused with prior art against '231:

• US 5,695,940 (Drmanac et al., Method of sequencing by hybridization of oligonucleotide probes) lists 5202231 | Apr., 1993 | Drmanac et al. in its References Cited U.S. Patent Documents. This is a same-inventor family member citing the parent — not art against '231. (everypatent.com/comp/pat5695940.html)
• PT 703991 E ("DNA sequencing by ligation and progressive cleavage") cites US5202231A with 1987-04-01 / 1993-04-13 / Drmanac Radoje T. Forward citation. (patents.google.com/patent/PT703991E/en)
• WO 97/10365 (Affymetrix, expression monitoring) lists US5202231(A) 1993-04-13 in its international search report's cited documents. Forward citation. (pt.espacenet.com, WO9710365 citedDocuments)
• The 471-entry Google Patents "Cited By" table (e.g. US 5,442,594, US 5,633,957, US 5,837,832, US 6,025,136 etc.) is likewise forward-citing art. None of it can be § 102 art against '231, since all of it postdates 1993.

---

C. Pre-1993 patents in the immediate citation neighbourhood — and why they still are not § 102 art against '231

The PT 703991 E citation table exposes a cluster of co-field patents. Their dates matter:

This is the analytically important point. The competing "sequencing by hybridisation" disclosures that one might instinctively reach for — Southern's PCT/GB 88/00460 (1988-05-03), Lysov et al. Dokl. Akad. Nauk SSSR 303:1508–1511 (1988), Bains & Smith J. Theor. Biol. 135:303–307 (1988), GBF's oligonucleotide-bank work, Macevicz, and Church — all postdate the 1987-04-01 Yugoslav priority date of '231 and therefore cannot anticipate or render obvious claims 1–4. The direction of the prior-art relationship runs the other way: '231 is the blocking reference against them. (The Lysov/Bains/Southern items are listed in the Encyclopedia of Analytical Chemistry "Sequencing and Fingerprinting DNA by Hybridization with Oligonucleotide Probes" reference list, which also cites the priority chain "Yugoslav Patent Application 570/87, 1987; issued as … US Patent 5202231, 1993.")

Critical-date note (pre-AIA). '231 is a pre-AIA patent. Its effective priority is 1987-04-01 (YU 570/87) under § 119, with the U.S. parent (Ser. No. 07/175,088) filed 1988-03-30. For § 102(b) the one-year bar runs from the U.S. filing date — i.e. a 1987-03-30 critical date — and under In re Hilmer a § 119 foreign priority date generally cannot be used to antedate a § 102(b) statutory bar. That creates only a two-day window (1987-03-30/31) of material that could be § 102(b) rather than § 102(a) art. I flag this as a nuance worth confirming against the file wrapper rather than as a firm conclusion.

---

D. Direct answer to "which claim(s) does it potentially anticipate under § 102"

• Claim 1 — no anticipating reference identified in the material I could verify. The three-step combination (discriminating hybridisation → detection → sequence inference by compiling overlapping probes) is not disclosed as a whole by Maxam & Gilbert, Sanger, Wallace 1979, Wood 1985, or Al-Hakim & Hull. Wallace 1979 and Wood 1985 supply elements of step (a) only.
• Claim 2 (8–20-mers) — no anticipation identified. Probe length as such is not the inventive point; no facial reference is analysed as disclosing a defined 8–20-mer probe set used for sequence assembly.
• Claim 3 (11-mers) — no anticipation identified from the facial citations. The specification's own statement that 11-mers were "the shortest oligonucleotides that can be successfully hybridized" is applicants' characterisation of the art, and is § 102-relevant only insofar as it concedes that 11-mer hybridisation was known — it is not itself a § 102 reference.
• Claim 4 (11-mer containing a unique 8-mer; (N2)N8(N1) / (N1)N8(N2)) — no anticipation identified. No facial reference discloses the ordered-probe-group architecture.

The honest answer is that the best prior art against '231 is almost certainly not on the face of the patent as rendered to me — it is the file-wrapper art (if any) plus the later-published competing SBH disclosures, and those later ones are disqualified by date. If an invalidity case exists against '231, it would have to rest on Wallace 1979 / Wood 1985 as § 103 underpinnings combined with a pre-1987-04-01 disclosure of overlap-assembly — and I have not found such a disclosure.

---

Confidence and residual uncertainty

• High confidence: the in-specification citation set and its bibliographic details (Tables A, items 1–10); the forward-citation direction of US 5,695,940, PT 703991 E, WO 97/10365 and the 471 "Cited By" entries; the post-1987 dates of the Church/Macevicz/Du Pont/ABI/GBF cluster; the claim-1 element-by-element analysis.
• Medium confidence: whether Wallace 1979 or Wood 1985 disclose probe lengths inside claims 2/3 (needs full-text verification); the exact filing vs. issue dates of the cluster patents in Table C (taken from a single Google Patents citation table, not independently cross-checked).
• Low / explicit uncertainty — flagged, not resolved: (i) the identity and completeness of the examiner-cited references on the '231 front page — I did not retrieve it and did not fabricate it; (ii) the Twigg Nature 283 (1987 vs. 1980) date inconsistency, left uncorrected per instructions; (iii) the pre-AIA § 102(b)/§ 119 Hilmer interaction on the two-day window.

I'll verify the dates and existence of the key competing SBH references, and probe for any pre-1987 art I may be missing, before building the § 103 argument.

Confirmed the post-priority dates of the three competing SBH disclosures. Now let me look for any pre-1987 art (which is what § 103 actually requires) and for any validity/reexam record on the '231.

Obviousness Analysis under 35 U.S.C. § 103 — US 5,202,231

Analyst's note on this section. This analysis builds directly on the Prior Art section already generated and does not repeat its reference-by-reference content. Two procedural points carried forward and one correction:

1. Governing statute. The '231 application was filed 1991-06-18 as a continuation of Ser. No. 07/175,088 (1988-03-30). Pre-AIA § 103(a) governs (AIA § 3(n)(1) applies only to applications filed on or after 2013-03-16). The operative question is whether the differences between claims 1–4 and the prior art "are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art."
2. Date discrepancy, unresolved. The task header says April 26, 2026; the fetched patent page is timestamped 2026-09-30. Consistent with the earlier section, I treat neither as a substantive input.
3. Contradiction flag (minor). The Patent Summary section described claim 4's dependency on claim 2 as "textually awkward." I disagree and flag it: an 11-mer is within claim 2's recited "8-mers to 20-mers," so claim 4's dependency on claim 2 is legally coherent (claim 3 would have been the more natural parent, but that is a drafting preference, not an inconsistency). This does not change the obviousness analysis, but the earlier characterization should not be relied on.

---

I. The threshold problem: what art is even available under § 103

This is where a real invalidity case either lives or dies, and it largely kills the intuitive case.

The decisive consequence. The three disclosures a lay observer would immediately reach for — Lysov et al., Dokl. Akad. Nauk SSSR 303:1508–1511 (1988); Bains & Smith, J. Theor. Biol. 135:303–307 (1988); and Southern's GB 8810400 / PCT GB89/00460 (priority 1988-05-03) — all postdate 1987-04-01 and are therefore not § 102 art against the '231. I verified each date against multiple independent sources (PubMed 3250844 for Lysov; DOI 10.1016/S0022-5193(88)80246-7 for Bains & Smith; EP 0373203 / WO 89/10977 priority data for Southern). Nature Biotechnology's 1997 account confirms the sequence: "In 1987, a group from what was then Yugoslavia filed a patent on this concept. Although the concept was described by no less than four other groups between 1988 and 1991—all apparently independently…" (https://www.nature.com/articles/nbt0597-406a.pdf).

Two things follow, and they point in opposite directions:

• Against obviousness: the most probative "same-solution" evidence — four independent groups converging on SBH — is chronologically disqualified. The '231 is the earliest filing in the field. There is no prior-art reference, and certainly no combination, that discloses the completed idea.
• For obviousness (limited): near-simultaneous independent invention is still doctrinally relevant to the level of ordinary skill and to whether the result was predictable, and it undercuts any "failure of others" secondary consideration. But it cannot substitute for a § 102 reference supplying a missing limitation. I flag this as a genuinely contested use of the material, not a settled one.

Therefore the § 103 case must be built exclusively on pre-1987-03-30 art. That materially narrows the toolbox. The best I have is set out below, and I am explicit about provenance, because (as the Prior Art section stated) I could not retrieve the '231's own front-page examiner-citation list and will not fabricate one.

---

II. Person having ordinary skill in the art (PHOSITA), as of early 1987

A PHOSITA here is a molecular biologist or bio-organic chemist with a Ph.D. and ~2–3 years of post-doctoral or industrial experience, or an M.S. with ~5 years, possessing:

• routine, automated synthesis of defined oligonucleotides (phosphoramidite chemistry, commercial synthesizers available since the early 1980s);
• standard filter/colony/plaque hybridization (Southern 1975; Grunstein & Hogness 1975), as codified in Maniatis et al., Molecular Cloning (1982);
• gel-based sequencing by Maxam–Gilbert or Sanger, and computer-assisted assembly of overlapping reads (shotgun sequencing of ΦX174, Sanger et al. 1977; Staden's assembly programs);
• familiarity with the 1986–87 genome-project debate and its stated economics.

This skill level matters enormously: the more sophisticated the PHOSITA, the easier it is to characterize the claimed combination as the predictable assembly of known elements.

---

III. Limitation-by-limitation chart for the claims

Provenance caveat (read carefully). Wallace 1979/1981, Wood 1985, and Thein & Wallace 1986 come from the '231's own specification text that you supplied. The additional pre-1987 items (Craig 1971; Besmer 1972; Gillam 1975; Smith 1983; Sanger 1977; Staden 1979/1980; Grunstein & Hogness 1975; Southern 1975; Maniatis 1982; Little 1983) are drawn from the front-page reference list of the sibling US 5,695,940 (same family, same inventors, same field) and from general knowledge of the field. They are not confirmed on the '231's own face, and Grunstein–Hogness/Southern/Staden are cited from general knowledge without a retrieved document in this session. Treat them as leads to verify against the '231 file wrapper before relying on any of them in a written invalidity contention.

---

IV. The obviousness combinations

Combination A — Wallace 1979 + Wood 1985 + Smith 1983 (the "receiver" for steps (a) and (b))

What each supplies.

• Wallace et al. 1979, Nucleic Acids Res. 6:3543–3557 — probes of predetermined sequence and length (11–17 bases) hybridized to a target, with conditions that resolve a single internal mismatch from a perfect match.
• Wood et al. 1985, PNAS 82:1585–1588 — in 3 M tetramethylammonium chloride, T_m depends on probe length, not GC content. This is the enabling teaching for a set: it means one stringency condition works for every probe regardless of composition. Without Wood, a "set" claim is inoperable as a practical matter; with Wood, it is enabled.
• Smith (1983), in Methods of DNA and RNA Sequencing (Weissman ed.), pp. 23–68 — predicted the oligonucleotide length required for a sequence to occur only once in a genome of defined size (≈16–17-mers for 3×10⁹ bp). This is exactly the statistical reasoning the '231 specification reproduces as A = 4^L.

Motivation to combine. The motivation is not speculative; it is documented on the face of the patent. The '231's own "TECHNICAL PROBLEM" and "STATE OF THE ART" sections concede: gel sequencing yields ~100 bp per person-day; a Japanese center achieved ~1 Mb/day at ~17¢/bp; sequencing the human genome was estimated at ~$3 billion and "at least ten years"; and the 1986-87 genome-project discussions (Lewin, Science 232:1598–1600) were dominated by the need for a faster method. That is the KSR "design incentive" and "market force" motivation, established by the patentee's own admissions. A PHOSITA seeking a non-gel, massively parallel readout would combine the known mismatch-discriminating oligo probe (Wallace), the known uniform-stringency condition (Wood), and the known probe-length/genome-complexity calculus (Smith).

What Combination A does not reach. Wallace's work is fundamentally detection/diagnostic — identifying a known sequence (β-globin alleles, clone screening). Nothing in Wallace, Wood, or Smith says "reconstruct a de novo sequence." Step (c) remains unsupplied.

Combination B — Combination A + an overlap-assembly reference (Sanger 1977; Staden 1979/1980)

The add-back. Sanger et al. 1977 (ΦX174, Nature 265:687) and Sanger's later shotgun work, together with Staden's computer assembly programs, disclose the concept of generating many short pieces of known composition and reconstructing the contiguous sequence by computer from overlaps. The '231's step (c) — "compiling their overlapping sequences" — is the same abstract operation.

Motivation to combine. Overlap assembly of sequencing reads was, by 1987, the standard way to get from short reads to long sequence. The step from "assemble overlapping gel reads" to "assemble overlapping probe hits" is an analogical extension: both are sets of ordered, compositionally defined elements whose overlaps define order. Under KSR, "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious."

Where the analogy breaks (the patentee's best defense). This is the strongest nonobviousness argument available on claim 1, and it is substantive:

• Sanger gel reads are ~200–500 contiguous bases per read, each read being a long, high-information string with a natural internal order.
• A positive 11-mer probe is a single bit — presence or absence of one 11-base word. Reconstructing a sequence from a spectrum of k-mers is a categorically different computational problem (later formalized as the Eulerian-path/spectrum problem), and it is uniquely vulnerable to repeats: any 11-mer occurring twice destroys the ordering, which is precisely why the '231 specification devotes extensive text and ~3,725 dedicated tandem-repeat probes to that problem.
• A PHOSITA in early 1987 would not necessarily have had a reasonable expectation of success that a k-mer spectrum could be assembled. Under In re O'Farrell (943 F.2d 1308), "obvious to try" does not render a claim obvious where the prior art gives no indication which of many parameters are critical and there are numerous variables. The repeats problem was known to be severe.

So Combination B is a viable § 103 theory on a KSR/"predictable result" framing, but it is genuinely contestable, and the defendant would want a stronger anchor than post hoc functional identity.

Combination C — Southern 1975 / Grunstein–Hogness 1975 + Maniatis 1982 + Wallace 1979/1981 (the format/multiplexing theory)

What it supplies. Immobilizing DNA on filters and probing with labelled nucleic acids (Southern blot; colony/plaque hybridization) was routine textbook practice by 1982. Wallace 1981 extends this to defined synthetic oligos as probes.

Motivation. A PHOSITA facing a throughput problem (10⁶–10⁷ hybridization data points) would naturally scale an existing, validated format. KSR is directly on point: "a combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results." Multiplexing a known assay is a textbook predictable-results case.

Ceiling. This combination gets claims 1(a)–(b) structurally and nothing more. It does not reach step (c).

Combination D — Claims 2 and 3 as range/optimization claims

• Claim 3 (11-mers). Wallace 1979 expressly used 11-mers with mismatch discrimination. The '231 specification itself concedes the operative fact: 11-mers are "the shortest oligonucleotides that can be successfully hybridized." Selecting a length within a range the primary reference already practices, to obtain the stated benefit (a probe long enough to be stable), is a classic case of an obvious design choice among a finite number of predictable options. Claim 3 collapses if claim 1 collapses — it adds no independent inventive weight.
• Claim 2 (8–20-mers). Broader than claim 3 and therefore even more exposed: it is the full known working range of Wallace 1979 (11–17) plus a margin on both ends.

Combination E — Claim 4: the (N2)N8(N1) ordered-probe-group architecture

This is, counterintuitively, the claim most vulnerable to a KSR "obvious to try" attack, even though it is the most specific claim.

What supplies it. The grouping idea reduces to: use a probe long enough to hybridize stably (11-mer) but read only the information content of a shorter, less frequent unit (8-mer), by degenerate-ing the flanking positions. The building block is the known degenerate/mixed-sequence probe — Wallace 1981 is literally titled "…Hybridization of oligonucleotides of mixed sequence to rabbit β-globin DNA."

The two stated design constraints are both conceded. The '231 specification itself supplies each half of the equation:

• "there are 65,536 different 8-mer ONSs. Since 8-mer ONSs, according to the current knowledge, cannot form a stable hybrid, the 11-mer group can be used as an equivalent"; and
• 4¹¹ ≈ 4.19 × 10⁶ probes are "impracticable."

Motivation and KSR analysis. A PHOSITA facing both constraints (8-mer informational resolution desired; 11-mer stability required; ~4 million probes infeasible) has a finite number of identified, predictable solutions. The natural solution — vary the flanking positions while holding the informative core fixed — sits directly between the two known endpoints. That is the KSR / In re Kubin "finite number of identified, predictable solutions, with a reasonable expectation of success" scenario. The specific (N2)N8(N1)/(N1)N8(N2) notation is a label for the result of a routine optimization, and the claim's recitation of "unique 8-mer" adds only the selection of which 8-mers to use.

Where claim 4 holds up better. I have not identified a pre-1987 reference that discloses the ordered-group architecture as such, and no single reference anticipates it. The attack is therefore a § 103 "obvious to try"/design-choice attack, requiring a fact-finder to accept the two concessions as defining the problem so tightly that the solution follows — which is a legal argument, not a documentary one.

---

V. Motivation to combine — consolidated evidentiary basis

The single most important structural point is that the '231's own specification concedes the state of the art and the need, which converts what would otherwise be a "predictable result?" debate into a much easier motivation showing for a defendant.

---

VI. Per-claim conclusions

Two-way vulnerability summary. Claims 2–4 add essentially no incremental nonobviousness once claim 1's step (c) is resolved against the patentee. Conversely, all four claims stand or fall on step (c). There is no meaningful per-claim hedge: the entire invalidity case is a single question — was it obvious in early 1987 to infer a de novo sequence from the presence/absence of a set of exactly-matched short probes?

---

VII. Counterarguments the patentee will press (and how they read)

1. The date wall. The intuitive "everybody did it" case is chronologically barred. Lysov, Bains & Smith, and Southern are not § 102 art. If a defendant's petition or contention leads with them as references, it is legally defective on its face. This is the single most likely error in an obviousness challenge to this patent.
2. New use of an old method. Wallace's hybridization is detection of a known sequence; claim 1 is de novo sequencing. A shift in purpose can defeat § 103 where the new result was not predictable — the patentee will press the repeats/assembly unpredictability point hard.
3. Failure of others. Wallace, Wood, and the clone-screening literature all stopped at detection. The patentee will argue that the leap to sequencing was not made by the very people who supplied the enabling conditions.
4. Unexpected results. The specification's treatment of tandem repeats, the 3,725 dedicated repeat probes, and the internal-duplex false-negative problem can be characterized as nonobvious problem-solving.

Counter-counter. As to (3), the "failure of others" story is substantially undercut by the 1988 simultaneous filings — the field converged within months of the '231's priority; the patentee's own specification (equations 1–2) treats the governing statistics as routine. As to (4), the repeats problem was known pre-1987 from shotgun sequencing and was addressed by adding more probes — a predictable engineering response under KSR.

---

VIII. Confidence and residual uncertainty

High confidence

• The post-1987-04-01 dates of Lysov (Dokl. Akad. Nauk SSSR 303:1508–1511, 1988; PubMed 3250844), Bains & Smith (J. Theor. Biol. 135:303–307, 1988), and Southern (GB 8810400, priority 1988-05-03; EP 0373203; WO 89/10977) — each verified via independent sources, and therefore their disqualification as § 102 art against the '231.
• That pre-AIA § 103 governs, and that § 102(b) runs from the 1988-03-30 U.S. filing date.
• The element-by-element mapping of pre-1987 references to claim 1's steps (a) and (b): Wallace 1979 and Wood 1985 meet them.

Medium confidence

• The pre-1987 status and content of Smith 1983, Sanger 1977 (ΦX174), Staden 1979/1980, Grunstein–Hogness 1975, and Southern 1975. These come from the sibling US 5,695,940's front page and from general knowledge; not verified against a retrieved document in this session, and not confirmed on the '231's own face.

Low / explicit uncertainty — flagged, not resolved

• The '231's own front-page examiner-citation list was never retrieved. Any invalidity theory should be rebuilt from that list plus the file-wrapper art before filing. This is the same gap the Prior Art section recorded, and it is still open.
• Whether the Yugoslav application YU 570/87 actually supports the claims — if it does not, the effective date is 1988-03-30, which changes the § 102(a) analysis for any 1987 publication (though § 102(b) still runs from 1988-03-30, and the 1987-88 competitors would still fail).
• The precise § 119 / § 102(b) interaction (Hilmer line) as applied to the two-day window around 1987-03-30 — carried forward from the Prior Art section as an unresolved nuance.
• Whether the '231 was ever reexamined (my search for a reexamination was cut short by the tool limit). If an ex parte reexamination exists, it will contain the best examiner-vetted prior art and possibly a narrowing certificate — check Patent Center before relying on anything above.

---

IX. Bottom line

US 5,202,231 is a hard obviousness target for structural/reference reasons, not because the invention was subtle. The strongest realistic § 103 theory is a KSR "predictable combination" attack built from Wallace 1979 (defined 11–17-mer probes + single-mismatch discrimination) + Wood 1985 (uniform, length-dependent stringency enabling a probe set) + Smith 1983 (probe-length/genome-complexity calculus) + Sanger 1977/Staden 1979–80 (overlap assembly), with the patentee's own specification supplying both the motivation (the 1986–87 genome-scale sequencing crisis) and the concessions (11-mers are the shortest reliably hybridizing oligos; 4¹¹ probes are impracticable; repeat probes are needed). Claims 2, 3, and 4 add little or nothing once claim 1's step (c) is resolved against the patentee — claim 3 is the weakest, and claim 4 is best attacked under an "obvious to try" theory using Wallace 1981's mixed-sequence probes.

The fatal weakness in that theory is step (c): no pre-1987 reference discloses de novo sequence reconstruction from overlapping probe hits, and the repeats problem is a genuine, non-hand-waving argument that a PHOSITA in early 1987 lacked a reasonable expectation of success. The fatal weakness in any stronger theory is the date wall: the three most persuasive documents (Lysov, Bains & Smith, Southern) postdate the '231's 1987-04-01 priority and are not § 102 art. Any obviousness challenge that does not confront both of those points head-on is not a serious challenge.

Sources relied on beyond the supplied patent text: PubMed 3250844 (Lysov 1988); DOI 10.1016/S0022-5193(88)80246-7 (Bains & Smith 1988); EP 0373203 / WO 89/10977 / US 6,054,270 (Southern priority 1988-05-03); https://www.nature.com/articles/nbt0597-406a.pdf (Nature Biotechnology, 1997, on the five-group priority sequence); front-page reference list of US 5,695,940 (sibling patent, same family). URLs are given where retrieved; several pre-1987 items are cited from the sibling's front page and general knowledge and remain unverified.