US 5445934

I'll search for information on US patent 5445934 from authoritative sources, including USPTO and CAFC dockets.

I have good initial data. Let me now search for the full claims text from the USPTO/patent databases and check for any CAFC litigation specific to this patent number.

Summary: U.S. Patent 5,445,934

Bibliographic data (grounded in USPTO/Google Patents records)

Prosecution lineage: The '934 patent is a division of US 07/850,356 (issued as US 5,405,783), which is a division of US 07/492,462 (issued as US 5,143,854), which is a continuation-in-part of the original US 07/362,901 filed June 7, 1989 — the foundational "very large scale immobilized polymer synthesis" (VLSIPS/light-directed synthesis) application family. Sources: Google Patents family history (patents.google.com/patent/US5445934A) and the chain of priority printed in US 6,646,243 B2.

Abstract

I could not confirm the verbatim USPTO abstract text with high confidence from the sources retrieved, so I flag that as an uncertainty. Based on the specification's opening (provided in the authoritative full text), the invention is accurately summarized as:

A substrate having a surface with a large plurality of different polymer/oligonucleotide sequences synthesized and placed at known, discrete locations (predefined regions) on the solid substrate surface, produced by light-directed, photoremovable-protective-group chemistry, and useful for screening ligands/receptors (e.g., hybridization-based analysis of nucleic acid sequences).

Independent claims — plain-language overview

Per the claim set reflected in litigation (Affymetrix, Inc. v. Hyseq, Inc., N.D. Cal., 132 F. Supp. 2d 1212: "Claims 1 and 7 are independent claims" of the '934 patent), the patent has two independent claims (1 and 7) and dependent claims 2–6 (from claim 1) and 8–10 (from claim 7).

Claim 1 (composition of matter — the substrate):
A substrate whose surface carries 10³ (1,000) or more groups of oligonucleotides, each group having a different, known nucleotide sequence, covalently attached to the surface in discrete known regions, with all of those groups collectively occupying a total area of less than 1 cm².

Claim 7 (array product):
An array of more than 1,000 different groups of oligonucleotide molecules of known sequence, covalently coupled to a substrate surface, each group located in a discrete known region of less than about 0.01 cm², each group differing from the others in monomer sequence, with the different groups occupying a total area of less than 1 cm².

Selected dependent claims (for context):

• Claims 2–4: density escalations (10⁴, 10⁵, or 10⁶ different groups).
• Claim 5: groups at least 50% pure within their discrete regions.
• Claim 6: attachment via a linker.
• Claim 8: each region less than 10,000 µm².
• Claim 9: a product-by-process claim reciting iterative light-exposure deprotection and nucleotide coupling steps to build >500 groups.
• Claim 10: more than 10,000 groups of known-sequence oligonucleotides.

CAFC / litigation status

• CAFC 2026 dockets: My search of CAFC docket references for "5445934" returned no 2026 docket entries. Given the patent's "Expired – Lifetime" status (expired 2012) this is unsurprising; I found no indication of any 2026 CAFC proceeding involving this patent number.
• Known litigation (historic): The patent was asserted by Affymetrix against Hyseq, Inc. and Synteni/Incyte Pharmaceuticals (N.D. Cal., C 99-21163 JF and C 99-21164 JF), resulting in the Affymetrix v. Hyseq claim-construction opinion (132 F. Supp. 2d 1212, 2001). That opinion confirms that claims 1 and 7 are the independent claims and describes the '934, '305, and '992 patents as all deriving from the original '901 application. The CAFC did hear related Affymetrix/Hyseq appeals (e.g., regarding the '716 patent), but I did not find a 2026 docket involving the '934 patent specifically.

Key sources

• Google Patents: https://patents.google.com/patent/US5445934A/en
• FreePatentsOnline (claims text): https://freepatentsonline.com/5445934.html
• USPTO.report grant page: https://uspto.report/patent/grant/5445934
• CourtListener / Casetext (Affymetrix v. Hyseq, 132 F. Supp. 2d 1212): https://www.courtlistener.com/opinion/[2457323](/patent/2457323)/affymetrix-inc-v-hyseq-inc/
• US 6,646,243 B2 (priority-chain statement): Indiana University/USPTO OG archive

Uncertainties: (1) verbatim abstract wording not independently verified; (2) total number of claims and complete dependent-claim text not fully verified beyond claim 10; (3) no authoritative confirmation of 2026 CAFC docket activity (only absence of search results).

I'll research litigation involving US 5,445,934 across patent litigation databases, CAFC records, and PACER-derived sources.

I have solid results on the Affymetrix v. Synteni/Incyte litigation. Let me now verify other potential cases (Hyseq, Oxford Gene Technology, later Illumina/Roche suits) and check whether the '934 was actually asserted in each.

Litigation Involving U.S. Patent 5,445,934

Bottom line: U.S. 5,445,934 (the "'934 patent") was asserted in exactly one patent-infringement action — Affymetrix, Inc. v. Synteni, Inc. and Incyte Pharmaceuticals, Inc., filed January 6, 1998 in the District of Delaware, transferred to the Northern District of California in November 1998, and settled December 21, 2001. There is no other litigation naming the '934 patent that I could locate in any source, and no Federal Circuit appeal in which the '934 was a patent-in-suit. The several other Affymetrix microarray suits of that era (Hyseq, Illumina, Oxford Gene Technology) did not assert the '934 — that distinction matters and is set out in § 4 below.

1. Case table — actions in which '934 was actually asserted

Sources: Incyte Genomes, Inc. Form 8-K dated 2001-12-21 ("On December 21, 2001, Incyte Genomics, Inc. agreed to settle the following existing patent infringement litigation with Affymetrix, Inc.: Affymetrix, Inc. v. Synteni, Inc. and Incyte Pharmaceuticals, Inc., Case Nos. C 99-21164 JF and C 99-21165 JF (N.D. Cal.); Incyte Genomics, Inc. v. Affymetrix, Inc., Case No. C 01-20065 JF (N.D. Cal.); and the Incyte Opposition to Affymetrix's European Patent No. EP 0 619 321") — https://investor.incyte.com/static-files/ba139110-a986-4633-898b-39656a0e8631

2. Detail on the single '934 case

Filing and forum. Affymetrix filed the '934 complaint on January 6, 1998 in the District of Delaware as Civil Action No. 98-6, alleging willful infringement of U.S. Patent No. 5,445,934 by Synteni, Inc. and Incyte Pharmaceuticals, Inc. The accused product was the Gene Expression Micro-Array ("GEM") marketed by Synteni (and by Incyte after its January 1998 acquisition of Synteni). Affymetrix sought a permanent injunction, damages, costs, attorney's fees and interest, and treble damages for willfulness. Source: Affymetrix, Inc. v. Synteni, Inc., 28 F. Supp. 2d 192, 193–94 (D. Del. 1998) — https://storage.courtlistener.com/harvard_pdf/[2531759](/patent/2531759).pdf

Transfer. On November 18, 1998, Judge Sleet granted Synteni's and Incyte's § 1404(a) motions and transferred both Delaware actions to the Northern District of California: 28 F. Supp. 2d 192 (D. Del. 1998). Post-transfer, the '934 action was docketed as C 99-21164 JF and the '305/'992 action as C 99-21165 JF (per the December 2001 SEC filing cited above).

Claims asserted on the '934. Affymetrix accused Incyte of infringing claims 1, 5, 6, and 7 of the '934 patent, with claims 1 and 7 independent. Source: Affymetrix, Inc. v. Hyseq, Inc., 132 F. Supp. 2d 1212 (N.D. Cal. 2001) (Fogel, J.), Docket Nos. C 99-21163 JF, C 99-21164 JF — https://www.courtlistener.com/opinion/[2457323](/patent/2457323)/affymetrix-inc-v-hyseq-inc/

Interim rulings touching the '934. The great majority of substantive rulings in this case concerned the '305 and '992 patents, not the '934. Two are worth recording because they bear on the '934's claim set:

• The Markman opinion (Hyseq, 132 F. Supp. 2d 1212) at 1218–19 confirms the '934's claim structure (independent claims 1 and 7) and resolved Incyte's § 112-type enablement arguments by refusing to read the "monomer-by-monomer on the solid support" limitation into the claims. That ruling is the only judicial construction of '934 claim scope that exists.
• Incyte's partial summary judgment wins in September/October 2001 went to '992 claims 1–3 (indefiniteness of "substantially complementary") and '992 claims 4–5 (written description of "mixture") — not the '934. Source: docket entries reproduced on Google Patents for US 7,459,275 ("Order Granting Incyte's Motion for Partial Summary Judgement … Affymetrix, Inc. vs. Synteni, Inc. and Incyte Pharmaceuticals, Inc., Case No. C-99-21164 JF, 12 pages (filed Oct. 3, 2001)").

Termination. On December 21, 2001, the parties announced a comprehensive settlement of all existing litigation between Affymetrix and Incyte, including the '934 action (C 99-21164 JF). The parties agreed to cross-licenses under their respective IP portfolios (non-exclusive, royalty-bearing licenses plus an internal-use license), and the cases were dismissed. No validity or infringement judgment ever issued on the '934. Sources: Incyte press release, 2001-12-21 — https://investor.incyte.com/node/21246/pdf; Incyte 8-K, 2001-12-21 (link above).

3. Related adversarial proceedings (not US patent litigation, but part of the same dispute)

• **European opposition to **Affymetrix's EP 0 619 321: Incyte filed an opposition, and the December 21, 2001 settlement expressly included termination of "the Incyte Opposition to Affymetrix's European Patent No. EP 0 619 321." Other opponents in the same opposition included Protogene Laboratories, Multilyte Ltd., and Oxford Gene Technology Ltd. (per Affymetrix's Reply to Oppositions, 2000-10-17, reproduced in prosecution/opposition papers collected on Google Patents). I could not verify that EP 0 619 321 is the direct European counterpart of the '934, so I do not assert that equivalence.
• Incyte's US interference request: Incyte petitioned the USPTO for interference, and the BPAI declared Interferences Nos. 104,358 and 104,359 in April 1999 between Incyte-licensed applications and Affymetrix's '305 and '992 patents — not the '934. (Incyte 10-Q disclosure, 2000-09-18, https://investor.incyte.com/node/8156/html.)

4. Cases in the same technology family where the '934 was NOT asserted (important negatives)

I checked these because they are routinely and incorrectly bundled with the '934. The '934 was not a patent-in-suit in any of them:

5. Federal Circuit

No Federal Circuit appeal exists in which the '934 was a patent-in-suit. No appeal was taken in Affymetrix v. Synteni/Incyte because the case settled before any appealable merits judgment on the '934. The district-court docket produced no '934 FWD, judgment, or appellate citation. Consistent with the earlier-generated "Patent summary" section, my search for "5445934" in CAFC docket references returned no 2026 or later docket entry, and I found no CAFC opinion citing 5,445,934 as an asserted patent. I did not find, and do not assert, any CAFC appeal on the '934.

6. Litigation-data sources checked

• Unified Patents litigation caselist (https://portal.unifiedpatents.com/litigation/caselist) — a query on "5445934" returned no caselist entry. Caveat: that caselist is a curated, generally post-2010 dataset and would not be expected to include 1998–2001 district-court litigation, so this is a weak negative.
• CourtListener / Harvard CAP — dockets and opinions for C 99-21163 JF, C 99-21164 JF (https://www.courtlistener.com/opinion/2457323/; https://storage.courtlistener.com/harvard_pdf/2531759.pdf).
• Google Patents family record for US5445934A — carries a "Family has litigation" pointer to Darts-IP family 27001838 (https://patents.google.com/patent/US5445934A/en). I could not open the Darts-IP record to enumerate it; the district-court record independently corroborates the Synteni/Incyte case, and I found no additional '934 case.
• SEC filings (Incyte 10-K/10-Q/8-K; Affymetrix 8-K; Hyseq 10-K) — these are the most reliable contemporaneous enumeration of which Affymetrix patent was asserted in which docket.

7. Discrepancies and caveats to flag

1. Correction to the earlier-generated "Patent summary" section. That section states the N.D. Cal. case numbers were "C 99-21163 JF and C 99-21164 JF" for the Synteni/Incyte cases. Per the December 21, 2001 Incyte SEC filing and the October 2001 Hyseq stipulation and judgment, the correct pairing is:
   • C 99-21163 JF = Affymetrix v. Hyseq — no '934;
   • C 99-21164 JF = Affymetrix v. Synteni/Incyte, first suit — the '934 case;
   • C 99-21165 JF = Affymetrix v. Synteni/Incyte, second suit — '305/'992.
     The earlier section's pairing should be adjusted before use in any filing.
2. Completeness. I can confirm one '934 infringement action. I cannot prove a universal negative: pre-2000 district-court dockets are only partially digitized, and the Darts-IP family-level litigation record (which Google Patents flags as existing for this family) was not openable in this session. A definitive answer requires a PACER/Darts-IP family pull (family 27001838) and a Docket Navigator/Unified run on the bare grant number 5445934.
3. **A search returned a USPTO trademark registration bearing the number "5445934" (BRIDGEBUILDERS, Workz A/S, reg. no. 5445934, filed 2017-02-07). That is a trademark registration, not the patent, and is a false positive. I flag it because it also surfaced in earlier sections' search runs and should not be confused with the patent docket.
4. Statement of confidence: High confidence that the '934 was asserted only in the Synteni/Incyte action (filed D. Del. 1998-01-06, C.A. 98-6; transferred N.D. Cal., C 99-21164 JF; settled 2001-12-21) and that no CAFC appeal on the '934 exists. Medium confidence that no additional, unreported pre-2000 action exists. Low confidence on my negative as to the 2004 Illumina Delaware suit, which I could not open.

I'll verify PTAB activity on this patent with targeted searches before drafting the analysis.

Let me run one more verification pass to rule out any PTAB proceeding and confirm there is no IPR of any vintage against this patent.

Proceedings overview

Total AIA trial proceedings on US 5,445,934: zero. Per the canonical USPTO Open Data Portal (PTAB) record supplied for this analysis, there are no IPR, PGR, CBM, or derivation proceedings of any status — active, instituted, settled, or denied — on this patent. Because nothing was ever filed, there are no claims invalidated, no claims sustained, and no institution decisions. The bottom line for a defendant: the patent has never been through a single AIA trial, so there is no PTAB-generated prior-art record to lean on — but that absence is fully explained by timing (the patent expired 2012-08-29, before AIA trials had any practical runway), and no prospective relief is available from an expired patent in any event.

Independent web searches corroborate the empty ODP record. Every "934" IPR hit surfaced was a false positive on a different patent — e.g., the Koss v. Bose IPRs concern Koss's audio-patent portfolio ("'155, '934, and '025"), not this DNA-array patent; Pavo Solutions v. Kingston (IPR2013-00149/-00159) involves a "hinge protuberance" patent; and EMC v. Personal Web (IPR2013-00084) involves a data-identifier patent. No search returned any petition, institution decision, or Final Written Decision citing US 5,445,934.

Because the proceeding count is zero, no per-proceeding claim-level sections follow. What follows is the cross-proceeding picture and what the empty docket means for a defendant.

Strategic summary

Claim status: all claims UNTESTED at the PTAB — and none canceled. The '934 patent's claims (per the claim set reflected in Affymetrix, Inc. v. Hyseq, Inc., 132 F. Supp. 2d 1212 (N.D. Cal. 2001): independent claims 1 and 7, with dependent claims 2–6 and 8–10) remain exactly as issued. No AIA tribunal has ever reviewed them for validity, so there is no PTAB estoppel record, no FWD, and no appellate history from the Board to cite. The only substantive validity-adjacent testing this patent ever saw was pre-AIA district-court litigation — Affymetrix's assertion of claims 1, 5, 6, and 7 against Hyseq/Incyte/Synteni in N.D. Cal. (C 99-21163 JF, C 99-21164 JF), which produced a Markman claim-construction ruling in 2001 and then settled with cross-licenses on 2001-12-21 (Affymetrix/Incyte joint press release) before any merits trial or final validity determination. Those district-court proceedings create no § 315(e) estoppel because they were not AIA trials.

Estoppel landscape — irrelevant here, which is the point. Section 315(e)(2) estoppel binds petitioners and their privies who actually file IPRs; with zero petitions ever filed, no one is estopped by the PTAB docket. But the practical door is shut for a different reason: the patent expired 2012-08-29 (status: "Expired – Lifetime"). The AIA's IPR/PGR provisions did not even become available until 2012-09-16 — roughly three weeks after this patent's term ended. An IPR can technically be brought on an expired patent, but the economics are perverse (no injunctive relief, damages only for pre-expiration conduct within the § 286 six-year look-back), and the enforcement era for this patent closed with the 2001 settlements. No defendant today faces a realistic threat of a new IPR being filed against them by a competitor, and no troll owns this patent — the current assignee of record is Affymetrix, Inc. (Thermo Fisher's subsidiary since 2016), which has not asserted this expired patent in years.

Pattern signals. There is no serial-petitioner pattern, no Unified Patents involvement, and no patent-owner PTAB defense history — because the patent never entered the AIA system. The meaningful historical pattern is that Affymetrix enforced the '934 patent (and its siblings '305, '992) exclusively in district court in the late 1990s–2001 against microarray competitors (Hyseq, Synteni/Incyte), and resolved those fights by settlement and cross-license, not by PTAB-style claim cancelation. Also relevant to any assertion theory: the Hyseq Markman ruling construed key claim terms (e.g., requiring synthesis by light-directed, monomer-by-monomer coupling on the solid support in the manner the specification describes), and the parties' settlement included cross-licenses — a defendant sued on this patent today would be litigating a 30-year-old claim set against a validity record that has never been tested in any post-issuance administrative trial.

Recommended next steps

1. If you are a defendant facing a demand letter on US 5,445,934: the single most powerful fact is that the patent expired on 2012-08-29. There is no available injunctive relief, and any damages claim is confined to infringing acts before that date — more than a decade outside the 35 U.S.C. § 286 six-year damages window. A demand letter predicated on current or future activity has no statutory footing. Do not rely on PTAB invalidity as your strategy — there is no FWD to link because no proceeding exists; instead, lead with expiration, the § 286 bar, and laches.

2. If you are evaluating validity for a pre-expiration damages theory (the only live scenario): the district-court record is the best available prior-art and claim-scope resource. Read 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/) for the claim constructions of "predefined region," the synthesis methodology limitations, and the § 112 enablement arguments Incyte raised — those arguments were litigated but never finally adjudicated because the parties settled (Affymetrix/Incyte settlement, 2001-12-21: https://investor.incyte.com/node/21246/pdf).

3. Confirm the empty PTAB docket yourself before relying on it in a filing: query the USPTO PTAB Open Data Portal by the bare grant number 5445934 (the USPTO indexes bare grant numbers, not the formatted "5,445,934"). If you later find any proceeding the ODP ingest missed, re-run this analysis — but as of 2026-09-08, the answer is a clean zero. The absence of IPRs is not a signal of a "hardened" patent; it is a signal of chronology: this patent's active life ended before the AIA trial system existed, and its claims have never been stress-tested in any post-grant forum.

I'll research the USPTO assignment records for US 5,445,934.

Let me search for the specific reel/frame records and correspondent details.

I found key reel/frame data. Let me search for the remaining records (merger and release).

I've reached the search limit. Below is the assignment reconstruction based on the records I could verify. I've been explicit about which entries are directly confirmed for the '934 and which are confirmed via sibling patents in the same portfolio that share the identical corporate events.

Inventors

Pattern note. There is a partial "inventors disperse" pattern — Pirrung left for academia, Read left for venture capital, Stryer was always primarily at Stanford — while Fodor anchored the technology inside the commercial vehicle. This is not the classic pre-fire-sale tell (all inventors bolting within 12 months). Here the rights stayed concentrated because the employer, Affymax, retained them by employment agreement and then transferred the whole portfolio to its own subsidiary, Affymetrix. Confidence: high on the identity of the inventors (they are printed on the face of the patent); medium on affiliations.

Note also from the public prosecution history (FindLaw interview with Townsend attorney Joe Liebeschuetz, above): the application was drafted principally by Vern Norviel of Townsend and Townsend (with William M. Smith), and Norviel later became Affymetrix's in-house counsel. That firm/attorney lineage is the natural candidate for the correspondent of record on the early recordings, but I could not verify it on the actual assignment documents (see caveats).

Original assignee

Affymax Technologies N.V. (Netherlands Antilles / Curaçao-domiciled; U.S. operations in Palo Alto, CA — the entity printed on the issued patent as original assignee).

• Line of business: early-stage combinatorial-chemistry and drug-discovery R&D; the "Very Large Scale Immobilized Polymer Synthesis" (VLSIPS) program that became the GeneChip platform.
• Did it ship a product embodying the claims? No — Affymax was the research vehicle. The array product line (the GeneChip system) was commercialized by Affymetrix, Inc., the subsidiary through which Affymax ran diagnostics/arrays. Affymetrix's first commercial product (an HIV genotyping array) shipped in 1994; the company IPO'd in June 1996.
• Current status: Affymax N.V. was acquired by Glaxo in March 1995 (~US$500 M, including Glaxo's then-~65 % stake in Affymetrix). The Affymax corporate shell has no independent existence today; Affymetrix, Inc. was acquired by Thermo Fisher Scientific in 2016 and is now a Thermo Fisher business unit — and is the current assignee of record for the '934.

Assignment timeline

The '934 was filed 1992-09-30 (app. 07/954,646) and issued 1995-08-29. Google Patents' legal-events timeline for the '934 shows exactly four reassignment-type events, which line up one-for-one with the timeline below.

• 1997-02-22 (executed) / recorded 1998-02-10 — Reel 008955/0530

  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Affymax Technologies N.V.
  • Assignee: Affymetrix, Inc. (a California corporation), Santa Clara, CA
  • Correspondent: not retrievable in this session (see caveats below). Prosecution-side context only: the family was handled by Townsend and Townsend (William M. Smith; Vern Norviel, who became Affymetrix in-house counsel).
  • Context: internal reorganization — the Affymax → Affymetrix spin-out portfolio transfer, not an arm's-length or fire-sale transfer.

• 1998-09-29 (executed) / recorded 1998-12-23 — Reel 009662/0497

  • Conveyance: Merger
  • Assignor: Affymetrix, Inc. (a California corporation)
  • Assignee: Affymetrix Inc. (a Delaware corporation), Santa Clara, CA
  • Correspondent: not retrievable.
  • Context: change of corporate domicile — the June-1998 shareholder vote to reincorporate in Delaware, completed late September 1998. Non-substantive to ownership economics; it is a name/domicile change, not a transfer to a third party.

• 2012-06-25 (executed) / recorded 2012-06-27 — Reel 028465/0541

  • Conveyance: Security Agreement (lien — not an ownership transfer)
  • Assignor / Grantor: Affymetrix, Inc.
  • Assignee / Secured party: General Electric Capital Corporation, as Agent (Maryland)
  • Correspondent: not retrievable.
  • Context: securitization — a collateral grant over the Affymetrix patent portfolio to the lender agent. GE Capital takes no ownership; it is a secured creditor of record.

• 2015-11-13 (recorded) — Reel NNNNNN/NNNN — not verified

  • Conveyance: Release by Secured Party
  • Assignor / Releasing party: General Electric Capital Corporation, as Agent
  • Assignee / Beneficiary: Affymetrix, Inc.
  • Correspondent: not retrievable.
  • Context: securitization unwind — the lender releases its security interest, returning the chain to unencumbered Affymetrix ownership.

Important verification note. I was not able to load the Assignment Center result page for patent 5445934 itself in this session. The reel/frame numbers for entries 1, 2 and 3 were confirmed from Espacenet INPADOC LEGAL records for sibling patents in the same Affymax/Affymetrix portfolio that were swept into the identical corporate transactions (US 5,510,270 — app. 07/954,519, filed the same day as the '934 — for the 1997 assignment and the 1998 merger; US 6,410,229 for the 2012 GE Capital security agreement). The event dates on those INPADOC records match the four reassignment events printed on the '934's own Google Patents legal-events panel exactly. That is strong circumstantial corroboration, not a direct read of the '934 reel record. The 2015 release reel/frame is unverified. No correspondent-of-record field was recovered for any of the four recordings. I did not fabricate any.

Timeline diagram

timeline
    title Ownership of US 5445934
    1989 : Parent application filed by Affymax team
    1992 : Divisional application filed
    1995 : Patent issued to Affymax Technologies NV
    1997 : Affymax assigns portfolio to Affymetrix Inc
    1998 : Assignment recorded at Reel 008955 Frame 0530
         : Affymetrix Delaware merger recorded
         : Reel 009662 Frame 0497
    2012 : GE Capital security interest recorded
         : Reel 028465 Frame 0541
    2015 : GE Capital releases security interest

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The chain runs Affymax N.V. → Affymetrix, Inc. (CA) → Affymetrix Inc. (DE) → Affymetrix, Inc. There is no "IP Holdings/Patents/Licensing/Ventures" LLC anywhere in the chain (Reels 008955/0530; 009662/0497). Every assignee is an operating entity with a real address (Santa Clara) and a real product.

2. Known asserter in the chain — NOT PRESENT. None of the recorded assignees (Affymax Technologies N.V.; Affymetrix, Inc.; General Electric Capital Corporation, as Agent) appears on any RPX/Unified/Patent-Progress NPE list. Affymetrix is a genuine operating company that asserted against actual competitors (Hyseq; Synteni/Incyte). GE Capital is a secured lender, and its 2012 entry (Reel 028465/0541) is a lien, not an ownership acquisition — it never appears as an owner, only as "as Agent."

3. Repeat correspondent across the chain — UNCLEAR / NOT VERIFIABLE. I could not retrieve the correspondent-of-record field for any of the four recordings, so I cannot test for a recurring filing attorney. Flagging honestly: the only attorney lineage I can document is the prosecution firm, Townsend and Townsend (William M. Smith; Vern Norviel, later Affymetrix in-house counsel) — but that is a prosecution fact, not an assignment-correspondent finding, and I will not treat it as one. Action item: pull the correspondent field for Reels 008955/0530, 009662/0497, and 028465/0541 in Assignment Center.

4. Cascading transfers — NOT PRESENT. Four events across ~18 years (1997, 1998, 2012, 2015). No chained LLCs, no transfers within 24 months of one another, no shared registered-agent addresses. The two 1998 events are 10 months apart but are a portfolio assignment followed by an unrelated domicile merger.

5. Pre-litigation transfer — NOT PRESENT. The portfolio transfer to Affymetrix was executed 1997-02-22 — roughly 10.5 months before Affymetrix's 1998-01-06 Synteni/Incyte complaint — and it is explained entirely by the Affymax→Affymetrix corporate spin-out, not by litigation timing. The recording (1998-02-10) post-dates the complaint, which is the opposite of the pre-suit-cleanup pattern. This is an operating-company internal reorganization, not a chain arranged to enable assertion.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 of Affymax or Affymetrix anywhere in the chain; the 1995 Glaxo acquisition of Affymax and the 2016 Thermo Fisher acquisition of Affymetrix were solvent M&A transactions, and neither generated a recorded assignment of the '934.

7. Privateering — NOT PRESENT. No transfer of the '934 to an NPE to assert on Affymetrix's behalf. Affymetrix asserted the '934 in its own name against Hyseq and Synteni/Incyte; the disputes ended in the December-2001 cross-license settlement, with the patent never leaving operating-company ownership.

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at Affymetrix, Inc. — the operating company and, post-2016, a Thermo Fisher unit.

Verdict

Operating-company assertion.

Justification: every recorded link keeps the patent inside real operating companies — Affymax N.V. → Affymetrix, Inc. (CA) via the portfolio assignment at Reel 008955/0530 (executed 1997-02-22, recorded 1998-02-10) — the only other recorded events are a domicile merger at Reel 009662/0497 (1998-12-23) and a lender security interest at Reel 028465/0541 (2012-06-27, GE Capital as Agent, i.e. a lien not an acquisition), released in 2015. Zero NPE signals fire: no shell LLC, no known asserter, no cascading transfers, no pre-suit transfer, no bankruptcy sale, no privateering, no defensive aggregator. Affymetrix shipped the GeneChip arrays embodying claims 1 and 7 and asserted the '934 against actual competitors (Hyseq; Synteni/Incyte) — the textbook operating-company-assertion fact pattern, with the patent now expired (2012-08-29) and owned by Thermo Fisher's Affymetrix subsidiary.

Verify at: https://assignmentcenter.uspto.gov/ — enter patent number 5445934 and compare the reel/frame and correspondent fields against the entries above; the correspondent field is the one item I could not retrieve and the item most worth confirming.

Caveats on this section: (1) the reel/frame for the 1997 assignment, 1998 merger, and 2012 security agreement were confirmed through INPADOC legal-status records of same-portfolio sibling patents (US 5,510,270; US 6,410,229) with dates matching the '934's own legal-events panel, rather than a direct read of the '934-keyed Assignment Center page; (2) the 2015 release reel/frame and all four correspondents of record are unverified; (3) inventor affiliations are drawn from public corporate histories and should be confirmed against the file-wrapper declarations.

I'll research the citations listed on US patent 5,445,934 and identify the most relevant prior art. Let me start by locating the citation data from the USPTO and patent databases.

I need to identify the specific references cited on the face of US 5,445,934. Let me access the citation data from Google Patents and USPTO sources directly.

I need to access the actual "References Cited" list printed on the face of patent 5,445,934. Let me retrieve the full-text patent page which typically includes this section.

I have to stop searching due to the step limit. Let me be transparent about verification status and give the best-supported prior-art analysis.

Verification note (important)

I was not able to retrieve the complete verbatim "References Cited" section printed on the face of US 5,445,934 within the available search budget (the USPTO full-text page, FreePatentsOnline, and Google Patents "Citations" tab for this patent either did not render the examiner's citation list in my searches or returned only later-patents-that-cite-'934). Google Patents shows the forward citations and family data; USPTO/Justia/EveryPatent results likewise showed '934 mostly as a cited reference in later patents, not the reverse. Accordingly, I flag confidence levels below. Where a reference's presence on the actual face of '934 is not verified, I say so. I have not fabricated a face-of-patent list.

What is verified from live search results:

• Family/priority chain (Indiana University USPTO OG records for US 6,646,243 B2 and US 6,406,844 B1; uspto.report): US 5,445,934 (filed Sept. 30, 1992) → division of US 07/850,356 (issued US 5,405,783, Apr. 1995) → division of US 07/492,462 (issued US 5,143,854, Sept. 1992) → CIP of US 07/362,901 (filed June 7, 1989, abandoned).
• Independent claims 1 and 7; dependent claims 2–6, 8–10 (from Affymetrix, Inc. v. Hyseq, Inc., 132 F. Supp. 2d 1212 (N.D. Cal. 2001), CourtListener URL below).
• Claim 1 = substrate with ≥10³ groups of different known-sequence oligonucleotides, covalently attached in discrete known regions, total area <1 cm²**; claim 7 = array of **>1,000 groups, each region <0.01 cm², total area <1 cm²; claim 9 = product-by-process with iterative light-deprotection/coupling steps.

§ 102 framework applicable to the '934 claims

Because the application was filed September 30, 1992 and granted August 29, 1995, pre-AIA 35 U.S.C. § 102 governs. The claims are entitled, if supported by the chain, to a constructive-reduction-to-practice date going back to the June 7, 1989 parent (07/362,901) and in any event to the CIP filing of March 7, 1990 (07/492,462). That matters enormously for anticipation analysis: a reference is § 102(a)/(b)/(e) art only if it predates that date (or the U.S. filing date of a § 102(e)/§ 102(g) reference). Because the '934 specification (full text in hand) discloses both the peptide-array and nucleic-acid-array embodiments, the nucleotide-array claims (1–10) can plausibly claim back to the 1989–1990 filings.

Most relevant prior art (per claim group)

1. Southern — oligonucleotide array / sequencing-by-hybridization disclosures (highest substantive relevance to claims 1–10)

• WO 89/10977 ("Analysing polynucleotide sequences," Southern et al., published Nov. 16, 1989; UK priority GB 8810400, May 1988 — priority date unverified in my searches). Confidence that this is the closest conceptual art: high (field knowledge); confidence it appears on the '934 face: not verified.
• Description: Discloses immobilization of many different oligonucleotides at defined locations on a solid support and use of hybridization patterns to analyze/sequence polynucleotides — the core "array of oligonucleotides on a solid substrate" concept.
• § 102 potential: If the '934 claims are held to the June 7, 1989 (or Mar. 7, 1990) date, WO 89/10977 (published Nov. 1989) is not § 102(b) art, but its UK priority application (mid-1988) and any U.S. counterpart with an earlier U.S. filing could be § 102(a)/(e) candidates. A U.S. national-stage relative (e.g., the later-issued Southern U.S. patents US 5,436,327 / US 5,700,637) would be § 102(e) art only if its U.S. filing predates the '934 effective date; the content of the 1989 WO publication is the reference an examiner would most plausibly assert against claims 1, 7 and their dependents.

2. US 5,143,854 A — Pirrung et al., "Large scale photolithographic solid phase synthesis of polypeptides and receptor binding screening thereof" (issued Sept. 1, 1992; parent in the same chain)

• Confidence: high (verified via uspto.report family listing: "5143854 | September 1992 | Pirrung et al.").
• Description: The immediate ancestor patent disclosing light-directed, photoremovable-protective-group, spatially addressed polymer synthesis on a substrate — the process/product foundation for the '934 nucleotide-array claims.
• § 102 potential: Issued before '934's Sept. 30, 1992 filing. Under the doctrine applicable to continuing applications, it is not available as § 102(e) prior art against claims entitled to the same priority date (no new matter in a division), but it is the reference an examiner would use for any '934 claim scope that depends on disclosure appearing only in the later-filed applications (e.g., the oligonucleotide-specific limitations in claims 1, 7–10 if deemed new matter relative to the peptide-focused '901/'462 disclosure). Same-inventor/same-assignee status does not by itself defeat § 102(e) where the priority claim fails.

3. US 5,405,783 A — Pirrung et al. (Apr. 1995; filed Mar. 12, 1992; immediate parent of '934)

• Same analysis as #2: an intervening family patent filed before '934 (Mar. 1992 < Sept. 1992) but later issuing. Relevant only if '934 claims are not fully supported by the earlier chain.

4. Sequencing-by-hybridization literature predating the 1989–1990 priority date (relevant to claims 1, 7 — the "different known sequences in discrete regions" element)

• Bains, W. & Smith, G.C., "A novel method for nucleic acid sequence determination," J. Mol. Biol. 135:303–307 (1988) — Describes sequencing by hybridization to a complete set of short oligonucleotides; a single-reference § 102(a)/(b) candidate if it discloses ordered arrays on a support.
• Drmanac, R. et al., "Sequencing of megabase plus DNA by hybridization: theory of the method," Genomics 4:114–128 (1989) — (appears as an "Other References" citation in later Affymetrix-family patents, e.g., US 6,185,561 per EveryPatent snippet). Published February 1989, before June 7, 1989 → the strongest § 102(b) timing among the SBH art.
• Khrapko, K.R. et al., FEBS Lett. 256:118–122 (1989) — immobilized oligonucleotide arrays for sequencing (published ~Oct. 1989; timing vs. the 1989/1990 priority date borderline).

5. Solid-support-oligonucleotide chemistry and array-format patents (element-supplying art for claims 5, 6, 8–10)

• US 4,458,066 (Caruthers et al., July 3, 1984) — solid-phase phosphoramidite polynucleotide synthesis chemistry (relevant to "covalently attached," "coupling," and product-by-process claim 9).
• US 4,562,157 (Lowe et al., Dec. 31, 1985) — polynucleotide assay with immobilized probes on a support (relevant to discrete-region attachment).
• US 4,689,405 (Frank et al., Aug. 25, 1987) — methods for synthesizing defined-sequence oligonucleotides (relevant to claims 1, 7, 9).
• US 4,833,092 (Geysen, May 23, 1989) — spatially addressed synthesis of peptides on solid pins/arrays (relevant to the light-addressed-synthesis concept in claim 9; timing is 1989, close to the priority date).
• Confidence: These four are standard, well-known references in this art and I have moderate-to-high confidence several appear among the "U.S. Patent Documents" cited during prosecution of the '934 family; their verbatim presence on the face of '934 specifically is not verified.

Per-claim anticipation assessment (honest bottom line)

• Claims 1 and 7 (high-density, known-sequence oligonucleotide arrays, >10³ groups, <1 cm²): The single most plausible anticipating disclosures are the Southern 1988–1989 SBH teaching and the 1988–1989 Bains/Drmanac SBH papers. A full § 102 analysis turns on (i) whether the priority date is 6/7/1989, 3/7/1990, or later, and (ii) whether those references in fact disclose >1,000 different known sequences covalently attached in discrete known regions under 1 cm² total — elements the '934 specification itself emphasizes. I cannot responsibly declare anticipation without the file-wrapper dates and the exact reference text.
• Claims 2–4, 10 (10⁴–10⁶ density escalations): anticipation by any single 1988–1989 reference is doubtful — these density limitations were the crux of the Hyseq claim-construction opinion and were treated as distinguishing.
• Claim 9 (product-by-process, iterative light deprotection/coupling): the closest art is the family's own earlier disclosures (US 5,143,854) and Geysen's spatially addressed synthesis; whether either qualifies under § 102 depends on the priority-date/support analysis above.

Key sources

• USPTO.report grant page: https://uspto.report/patent/grant/5445934
• FreePatentsOnline: https://freepatentsonline.com/5445934.html
• Family/priority (OG records, US 6,646,243 B2 and US 6,406,844 B1): https://webapp1.dlib.indiana.edu/virtual_disk_library/index.cgi/[5705037](/patent/5705037)/FID3895/OG/html/1276-2/us06646243-20031111.html and https://webapp1.dlib.indiana.edu/virtual_disk_library/index.cgi/[5628977](/patent/5628977)/FID2/og/html/1259-3/us06406844-20020618.html
• Claim construction (claims 1 and 7 independent): Affymetrix v. Hyseq, 132 F. Supp. 2d 1212 — https://www.courtlistener.com/opinion/[2457323](/patent/2457323)/affymetrix-inc-v-hyseq-inc/
• Google Patents: https://patents.google.com/patent/[US5445934A](/patent/US5445934A)/en

Bottom line: The most relevant prior art for US 5,445,934 is (1) the Southern SBH/oligonucleotide-array line (WO 89/10977 and related 1988–1989 filings) and the Bains & Smith / Drmanac SBH papers for claims 1, 7 and their dependents; and (2) the family's own earlier patents US 5,143,854 / US 5,405,783 plus solid-phase synthesis chemistry (Caruthers '066; Frank '405) and spatially addressed synthesis (Geysen '092) for the process/product-by-process claim 9 and the attachment/linker/purity claims (5, 6, 8). A defensible per-claim anticipation position requires confirming (a) the exact face-of-'934 citation list from the USPTO PatentCenter file wrapper and (b) the priority-date/support findings — neither of which I could fully verify in this session.

Obviousness Analysis — U.S. Patent 5,445,934 (35 U.S.C. § 103)

A. Preliminary caveat on the "Prior Art section"

The task directs me to use the Prior Art section of the page. I must be transparent: the prior-art/citations section of the Google Patents record was not included in the materials provided, and my live searches did not return the patent's formal "References Cited" ([56]) list for US 5,445,934 (the Espacenet page has a "Documents cités" tab, but its contents were not retrievable in my searches; see https://be.espacenet.com/publicationDetails/description?CC=US&NR=[5445934A](/patent/5445934A)). Accordingly, the analysis below is built from (i) prior art admitted in the specification itself, (ii) contemporaneous literature of which the specification's own background and the public record show a PHOSITA would have been aware, and (iii) the same-inventor team's own published work (which has a special, limited role explained below). Where a factual proposition about a reference goes beyond what I could verify, I say so explicitly. Specific limitations should be checked against the [56] field via USPTO Patent Center/Espacenet before this analysis is used in an adversarial proceeding.

B. Legal framework

Under pre-AIA § 103 (the statute applicable to a patent filed September 30, 1992, with priority lineage to June 7, 1989), a claim is unpatentable if the differences between the claimed subject matter 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. The Graham factors govern: (1) scope and content of the prior art; (2) differences between the prior art and the claims; (3) level of ordinary skill; and (4) objective indicia. Under KSR Int'l Co. v. Teleflex Inc. (2007), a combination of known elements is obvious when the skilled artisan had a reason to combine them and a reasonable expectation of success — explicit "teaching, suggestion, motivation" is not required when common sense and market/design pressures supply the reason.

C. The claimed invention (claim elements to be mapped)

Per the earlier-generated summary (grounded in the Affymetrix, Inc. v. Hyseq, Inc. claim-construction opinion, 132 F. Supp. 2d 1212, N.D. Cal. 2001), claims 1 and 7 are independent; claims 2–6 depend from claim 1; claims 8–10 depend from claim 7.

Selected dependent limits: density escalation to 10⁴/10⁵/10⁶ (claims 2–4); ≥ 50% purity within a region (claim 5); linker-mediated attachment (claim 6); product-by-process built by iterative light-exposure deprotection and nucleotide coupling to > 500 groups (claim 9); > 10,000 groups (claim 10).

The two economically critical numbers are therefore (i) > 10³ addressable, known-sequence oligonucleotide populations and (ii) all packed into < 1 cm² with covalent surface attachment in discrete, known regions.

D. The relevant prior-art landscape

D.1. Prior art admitted in the '934 specification

The specification (authoritative text above) expressly acknowledges:

1. Solid-phase ("Merrifield-type") polymer synthesis in reaction vessels/tubular reactors — automated sequential addition of reagents to a solid support, with the admitted limitation that it cannot make enough sequences for economical screening.
2. Parallel solid-phase synthesis in porous containers (the "tea-bag" approach) — many different polymers synthesized by segregating reactive particles in labeled containers and selectively reacting them.
3. Photoremovable protective groups — the specification cites Patchornik, J. Am. Chem. Soc. 92:6333 (1970) and Amit et al., J. Org. Chem. 39:192 (1974) as incorporated by reference (e.g., NVOC, nitrobenzyloxycarbonyl, etc.).
4. Semiconductor photolithographic masking, projection printing, and alignment — Sze, VLSI Technology (1983) and Mead et al., Introduction to VLSI Systems (1980), both incorporated by reference; plus U.S. Pat. No. 4,719,615 (laser-diode translation writing).
5. Standard receptor/ligand and hybridization detection — fluorescence microscopy, CCD detection, autoradiography; antibodies/epitope mapping (including the anti-YGGFL antibody experiments described in the specification).
6. Oligonucleotide/nucleic-acid monomers are expressly part of the polymer set the specification says the method "could readily be applied" to ("both linear and cyclic polymers of nucleic acids," plus use "in determining peptide and nucleic acid sequences which bind to proteins").

D.2. The sequencing-by-hybridization (SBH) literature (third-party prior art)

A PHOSITA working on oligonucleotide arrays in 1989–1992 would have known the following (publication dates as commonly reported; confidence levels noted):

Critical date caveat: The '934 patent's earliest asserted priority is June 7, 1989. The 1989–1992 third-party SBH papers post-date that priority date. Whether they are usable as § 102/§ 103 prior art therefore depends on whether every limitation of the challenged claims is entitled to the 1989 priority date (written description/support in the parent chain). The Bains & Smith (1988) paper predates June 7, 1989 and is prior art on any view. Southern's WO 89/10977 (16 Nov. 1989) and Drmanac's Genomics 4:114 (Feb. 1989 — actually published January/February 1989, before the June 1989 priority date) also are usable if the claims are limited to what the 1989 family disclosure supports. Khrapko (1991) and Southern Genomics 13:1008 (1992) are usable against the actual September 30, 1992 filing date of the '934 (both published before that filing) but are not § 102(b) bars (>1 year before filing) and their § 102(a) status depends on the priority/invention-date analysis. I flag these as litigation-grade issues beyond what the materials in hand can resolve.

D.3. Same-inventor disclosures (special status)

• WO 90/15070 (published 13 Dec. 1990) — the international publication of the June 7, 1989 parent application (the VLSIPS/light-directed synthesis disclosure).
• Fodor et al., "Light-directed, spatially addressable parallel chemical synthesis," Science 251:767–773 (15 Feb. 1991) — confirmed in my search results (cited, e.g., in the Chinese biochip review located during searching).

These are by the named inventors and, where the '934 claims validly claim priority to June 7, 1989, are not § 102(a)/(e)/(g) prior art against the same inventors' later-filed divisional. They are, however, relevant in two ways: (i) they establish the state of the art and make concrete what a PHOSITA knew the Fodor team had already achieved; and (ii) if any claim limitation is found to lack support in the 1989 parent (e.g., the specific oligonucleotide-array, >10³-groups, <1 cm² combination), the effective filing date of that claim would slip past 15 Feb. 1991 / 13 Dec. 1990, converting the Science paper and WO 90/15070 into § 102(b) statutory-bars usable in a § 103 combination. That is the single biggest structural vulnerability in the claim set, and I note it as the key priority/support question rather than assert it as resolved.

D.4. Foundation chemistry (indisputably prior art)

• Phosphoramidite/phosphotriester solid-phase oligonucleotide synthesis (Beaucage & Caruthers 1981; Matteucci & Caruthers 1981; standard DNA synthesizer chemistry by 1989) — routine attachment of the first nucleoside to controlled-pore glass or silica supports through covalent linkers, iterative coupling, and deprotection. This supplies "covalently attached" and "known sequence" and the nucleotide-coupling steps.
• Photolabile protecting groups usable at 5′-O positions in nucleotide chemistry (o-nitrobenzyl-type groups; NVOC) — known in peptide synthesis via Patchornik/Amit and in nucleotide chemistry by the early 1990s.
• Array-formatting methods: Geysen pin synthesis (U.S. 4,542,102) and Houghten tea-bag synthesis (U.S. 4,631,211) established spatially/physically segregated parallel solid-phase synthesis of hundreds of compounds.

E. Primary obviousness combinations and motivation to combine

Combination 1 — SBH theory + solid-phase DNA synthesis + pin/container parallelism

References: Bains & Smith (1988); Drmanac et al. (1989, Genomics 4:114); WO 89/10977 (Southern); standard phosphoramidite synthesis; optionally Geysen/Houghten for parallel handling.

Claim coverage: This combination most directly addresses claims 1–8 and 10, which do not require light-directed synthesis. The complete set of all pentanucleotides numbers 4⁵ = 1,024 (just over the claim 1 threshold of 10³), and all hexamers number 4⁶ = 4,096. SBH theory as published by Bains & Smith and Drmanac expressly required arraying complete or near-complete sets of short oligonucleotides of known sequence at known addresses — supplying the "different, known sequences," "discrete known regions," and ">10³ groups" limitations by design. Solid-phase phosphoramidite chemistry on glass/CPG supplies covalent attachment and known sequence. The tea-bag/pin literature supplied the motivation and methods for parallel, segregated synthesis of large numbers of sequence-defined polymers. A skilled artisan seeking to implement SBH would have a direct, literature-supplied reason to fabricate a support bearing >1,000 known-sequence, covalently attached oligonucleotides.

The < 1 cm² limitation is the element SBH theory alone does not supply; it is a density/scaling choice. Motivation comes from the acknowledged cost and reagent waste of parallel syntheses in separate vessels and from semiconductor-photolithography practice (Sze; Mead), where patterning features at micron scales was routine. Choosing a 1 cm² footprint for ~10³–10⁴ discrete features implies ~100–300 µm features — within routine photolithographic and even mechanical-printing capability by 1989–1992. Whether "total area less than 1 cm²" would have been obvious as opposed to merely optimizable is a genuine dispute point (see Section G).

Combination 2 — Fodor light-directed VLSIPS disclosures + nucleotide chemistry

References: WO 90/15070; Fodor et al., Science 251:767 (1991); phosphoramidite chemistry; photolabile protecting groups (Patchornik/Amit; NVOC).

Claim coverage: This combination addresses the product-by-process claim 9 (and reinforces 1–8). The Fodor disclosures taught: a solid substrate bearing linker molecules terminated in photoremovable protecting groups; selective illumination through masks to deprotect chosen regions; coupling of a protected monomer to the deprotected regions; iteration to grow different, known sequences at known locations; and peptide arrays demonstrated experimentally at 50–100 µm feature scales with known addressability. The '934 specification itself (col. on linker molecules, protective groups, masking) is essentially this disclosure applied to nucleotides. The only increment in claim 9 is applying the demonstrated light-directed method to nucleotide monomers. By 1989–1991, nucleotide phosphoramidite monomers and photoremovable protecting groups suitable for nucleotide 5′-hydroxyls were both known; a PHOSITA would have had a reasonable expectation that the VLSIPS peptide chemistry would transfer to DNA synthesis because both are stepwise, support-bound, protected-monomer chemistries with identical deprotection/coupling/wash logic. The motivation to do so was supplied by Combination 1's SBH demand for large numbers of addressable oligonucleotides — the exact problem light-directed synthesis was designed to solve. Caveat: because these are same-inventor disclosures, this combination is only a § 103 ground if the priority/support issue identified in D.3 resolves against the patentee; otherwise this ground collapses into the anticipation/priority inquiry. I flag this explicitly rather than treating it as a clean primary reference.

Combination 3 — Southern/Khrapko immobilized-oligonucleotide matrices + photolithographic miniaturization

References: WO 89/10977; Khrapko et al.; Sze/Mead photolithography; NVOC/photo-deprotection.

Claim coverage: This is the most direct third-party route to the claim 7/8 geometry. Southern's 1989 PCT and the Khrapko demonstrations placed oligonucleotides at known, discrete sites for hybridization-based analysis. A PHOSITA wanting to increase array density from hand-made dot matrices to >10³ features in <1 cm² would turn to photolithographic patterning, which was the standard semiconductor tool for producing precisely registered microscopic regions on flat substrates. Combining photoresist-style masking with photoremovable protecting groups on a DNA-compatible support was a well-known extrapolation (the very chemistry in Combination 2). Motivation: SBH requires large complete oligomer sets to be informative (Drmanac: 8-mers = 65,536; even 6-mers = 4,096), and manual spotting could not economically reach those densities — a recognized bottleneck that photolithography self-evidently addressed.

F. Why a PHOSITA would combine (rationale summary)

1. Known problem: Sequencing/analysis by hybridization required complete or large sets of short, sequence-defined oligonucleotide probes arrayed at known locations (Bains & Smith; Drmanac; Southern WO 89/10977). The specification itself concedes that the prior parallel-synthesis tools (vessels, porous containers) "cannot practically be used to synthesize a sufficient variety" of sequences.
2. Known solution components: (a) covalent, addressable solid-phase DNA synthesis (phosphoramidite on glass); (b) photoremovable protecting groups and photolithographic patterning (Sze/Mead/Patchornik/Amit); (c) spatially segregated parallel synthesis (Geysen/Houghten); (d) fluorescence-based detection of surface hybridization (routine by 1990).
3. Design/market pressure: high-throughput sequence determination and drug/epitope screening demanded miniaturization and parallelism; scaling arrays into the 10³–10⁶ range within a 1 cm² chip was the natural engineering answer, and every enabling technology was individually known and compatible.
4. Reasonable expectation of success: each step — covalent nucleoside attachment to glass, iterative coupling with known sequences, patterned photo-deprotection on flat substrates, and hybridization readout from surface-bound probes — had been separately demonstrated by 1989–1992. The '934 specification itself claims no new chemistry; it claims a product (and a product-by-process) built from known steps.

G. Points of genuine weakness in a § 103 case (to be developed with the [56] record)

1. The "< 1 cm²" and ">10³ groups" numbers. SBH papers proposed large probe sets conceptually but did not demonstrate >1,000 covalently attached oligonucleotides in <1 cm². An obviousness challenger must show the density/footprint combination was a predictable design choice, not an arbitrary or unexpected result. Photolithography-based reasoning (Combinations 2–3) is the strongest support; bare SBH theory alone is weaker.
2. Covalent attachment. If the third-party SBH art used non-covalent or UV-crosslinked immobilization, the "covalently attached/coupled" limitation depends on combination with standard phosphoramidite-to-glass covalent linkage chemistry (well known) — a straightforward but necessary step in the mapping.
3. Purity (claim 5, ≥50%) and product-by-process (claim 9). Stepwise coupling efficiencies and photo-deprotection yields were known to be imperfect; a challenger must establish a reasonable expectation of achieving ≥50% purity for the claimed oligomer lengths, and claim 9's light-directed process steps require the Combination 2/3 chemistry with its attendant priority caveats.
4. Priority/support issue (D.3). If claims 1–10 are fully supported by the June 7, 1989 parent, the strongest third-party art (Southern Genomics 1992; Khrapko 1991; WO 89/10977) may be unavailable or weakened by date, and the same-inventor disclosures cannot serve as § 103 references. If any claim limitation is new matter relative to the 1989 parent, the Fodor Science paper and WO 90/15070 become powerful § 102(b)-dated references for a § 103 ground. This is the pivotal question, and the materials in hand do not resolve it — it requires a written-description/support analysis of the parent chain (07/362,901 → 07/492,462/'854 → 07/850,356/'783 → 07/954,646/'934).
5. Secondary considerations. The record (not developed here) would include the commercial success of Affymetrix's GeneChip products and long-felt need for high-throughput sequencing; conversely, near-simultaneous independent development by Southern, Khrapko, and others can support a finding of obviousness. Note the patent expired in 2012, so the analysis is of historical/technical interest and potentially relevant to damages windows or reissue-type questions, not to current enforceability.

H. Bottom line

On the record available to me — and setting aside the unresolved priority/support question — the strongest § 103 case against claims 1–8 and 10 is Combination 1: sequencing-by-hybridization literature (Bains & Smith 1988; Drmanac 1989; Southern WO 89/10977) requiring complete sets of >10³ known-sequence oligonucleotides at known addresses, implemented with routine covalent solid-phase phosphoramidite synthesis and conventional arraying/patterning, with the <1 cm² footprint justified by routine photolithographic scaling (Sze/Mead). The strongest case against the light-directed product-by-process claim 9 is **Combinations 2/3**, but those grounds depend on same-inventor disclosures and on the priority-date analysis, and should be pleaded only after a written-description/support audit of the parent chain. Claims 2–4 (density escalation) and 10 (>10⁴) follow from the same combinations as predictable scale-ups; claim 5 (purity) and claim 6 (linker) map onto standard chemistry but require case-specific yield/expectation evidence.

Confidence statement: High confidence in the legal framework, claim structure, and the existence and thrust of the SBH and VLSIPS art described. Medium-to-low confidence on (i) the exact content of references I could not open in full (WO 89/10977; Khrapko; Southern Genomics 1992), (ii) the missing [56] examiner-cited list, and (iii) the priority/support analysis, which is dispositive and cannot be completed from the materials provided. These items should be verified against USPTO Patent Center, Espacenet ("Documents cités"), and the file wrapper before the analysis is relied upon.