US 6399365

I'll search for authoritative information on this patent number.

US 6,399,365 B2 — Analyst Summary

Search notes up front: My searches returned solid bibliographic and claim-family data for 6,399,365 specifically (Google Patents, FreePatentsOnline, Justia, and the USPTO Official Gazette text for its sibling 6,733,977). I found no CAFC 2026 docket entry naming 6,399,365 — two targeted searches of Federal Circuit 2026 docket material returned zero results. That is consistent with the patent having expired in 2014, but I want to be explicit: I cannot confirm any 2026 appellate activity from the sources available to me. Anything below not directly quoted is inference and is flagged as such.

1. Bibliographic data

Priority-chain caution. The sources are not internally consistent on the earliest date. Google Patents lists a priority date of 1994-06-08 while simultaneously showing "1995-06-07 Priority claimed from US 08/485,452." The chain visible in the sister patent's gazette entry is:

09/907,196 ← continuation of 09/302,052 (filed Apr. 29, 1999, now US 6,287,850) ← continuation of 08/485,452 (filed Jun. 7, 1995, now US 5,945,334)

I am reporting the 1994-06-08 date as the sources state it, not resolving it.

Related continuations/divisionals in the family (from Google Patents "worldwide applications"): US 6,551,817 (from 10/046,623), US 6,733,977 (from 10/229,759), US 2004/0106130 A1 (from 10/619,224), US 7,364,895 (from 10/639,696), plus numerous later continuations through 2010. The 10/639,696 branch shows interference activity at the USPTO in 2006 (declaration of interference Aug. 8, 2006; decision favorable to applicant Aug. 30, 2006).

2. Abstract

Verbatim (FreePatentsOnline / Justia, matching the USPTO abstract):

"A package for hybridization includes a substrate and a housing. The substrate has a first surface that includes an array of probes having biological polymers immobilized thereon. The housing includes a fluid cavity constructed and arranged for hybridization of a target to a probe of the probe array located inside the fluid cavity. The housing also includes a bar code."

Note: the Google Patents text fetched for this patent (the authoritative source supplied) contains the long description and figure list but not the abstract field itself, so the abstract above is cross-sourced — and it is consistent with the description's statement that "the top casing is of sufficient size to accommodate identification labels or bar codes in addition to the cavity."

3. Plain-language overview of the independent claims

The full claim set was not retrievable in one piece; I reconstructed the independent claims from (a) partial quoted text in search results and (b) the dependency references of dependent claims. Claim numbers I can identify as independent are 1, 2, 7, 10, 11, and 41 (dependent claims variously recite "of claim 1, 2, 8 or 10," "of claim 7," "of claim 10," "of claim 11," and "according to claim 41" — the reference to claim 8 implies an additional independent or intermediate claim I could not textually verify; flag this as uncertain).

Claim 1 — the core hybridization package. A package for hybridization with (i) a substrate whose first surface carries a probe array of different biological polymers immobilized on it, and (ii) a probe array density "exceeding 100 different biological polymers per cm²." The quoted fragment is truncated where it recites density; in substance this is the broad apparatus claim to a packaged, high-density bioarray. (Claim text partially retrieved; the balance of the claim is inferred.)

Claim 2 — the bar-coded, optically transparent embodiment. A package for hybridization comprising an optically transparent chip carrying an immobilized biological-polymer probe array, plus a housing with a fluid cavity constructed and arranged so a target can hybridize to a probe inside the cavity, the housing including (a) a bar code and (b) being arranged for use with a detection system. This is the claim that reads most directly on the abstract, and it is the commercially distinctive one: the identifier travels with the reaction chamber, so the scanner can associate the physical array with its experimental identity.

Claim 7 — the array itself (no housing required). An array claim directed to biological polymers on a substrate, with dependents reciting covalent attachment (claim 34), densities exceeding 400 and 1,000 different nucleic acids per cm² (claims 35–36), nucleic acids (37), proteins/polypeptides (38), a long Markush group of binding species (39), and an optically transparent substrate (40). Density recitation in the independent claim is unverified; given claims 35–36 add 400 and 1,000 per cm², the independent claim likely recites a lower bound (possibly the same "exceeding 100 per cm²" language used in claim 1). Uncertain.

Claim 10 — package with fluid communication. A package claim whose dependents recite that the biological polymers are "in fluid communication" (29), are separately located within an area of less than 10⁻² cm² (30), and that the array contains more than 100 (31) or more than 1,000 (32) different sequences. Independent text not retrieved.

Claim 11 — package for supporting a probe array. A package claim with dependents reciting that the housing is constructed to introduce fluid contacting the probe array to hybridize probes to fluid-delivered targets (15), and that the housing includes an alignment structure for use with a detection system, where the detection system is constructed to detect fluorescent light emitted from the array and transmitted through the chip (16). That last limitation ties the claim to back-side (transmission-mode) fluorescence imaging, matching the specification's discussion of optically transparent substrates.

Claim 41 — method of using a probe array. Explicitly quoted:

"providing an array of probes, comprising biological polymers immobilized on a substrate, having a density exceeding 100 different polymers per cm²; providing a bar code associated with said probe array; reading said bar code; aligning said probe array with a detection system; and detecting a signal from said probe array."

Dependents 42 and 43 raise the density threshold to 400 and to "1[00?]0" different polymers per cm² (claim 43's numeral is truncated in the retrieved text — do not treat as verified).

Other limitations appearing in dependent claims (useful for scope): nucleic acids through a linker group, 4–20 nucleotides in length (18–19); each polymer separately located within about 1 µm² to about 1,000 µm² (20); densities exceeding 400 and 1,000 different nucleic acids per cm² (21–22, 27–28); polymers attached "by selectively illuminating said substrate" (23) — i.e., the light-directed synthesis technique; oligonucleotides, proteins/polypeptides, and a Markush group of agonists/antagonists, toxins, venoms, viral epitopes, hormones, receptors, enzymes, substrates, cofactors, drugs, lectins, sugars, oligosaccharides, and monoclonal antibodies (24–26, 39).

4. What the specification actually covers (context for claim reading)

The disclosed device is an injection-molded, multi-piece plastic cartridge that turns a diced VLSIPS™ probe-array chip into a self-contained hybridization flow cell. Key elements described:

• Two-part (later three-part) ultrasonically welded housing; acoustic welding via V-shaped energy directors to form sealed internal channels (FIGS. 8a–9c).
• Chip mounted over a cavity on a ledge, leaving a trough for UV-cure adhesive, with the chip's back surface flush with or below the housing surface for protection (FIG. 6).
• Inlet/outlet ports sealed by septa that are pierced by needles from a fluid-delivery system; a vacuum drawn through a port holds the chip during adhesive cure.
• Asymmetric, non-flush edge and differing alignment holes to key the cartridge into a scanner holder "similar to an audio cassette tape."
• Temperature control: the cavity floor is made thin (~0.05″) and a depression beneath it receives a temperature controller; light-absorptive cavity floor to suppress reflection during imaging.
• FIGS. 31–36: three-casing version with a diffuser/concentrator channel geometry producing laminar flow (channels angled 2°–90°, preferably 5°–45°) and hollow pins as an alternative fluid interface; the intent is to avoid turbulent "hot spots."

Note the fetched Google Patents text also carries a long definitional passage (probe/target definitions; the Pirrung U.S. 5,143,854 and PCT WO 92/10092 synthesis background) that is boilerplate shared across the Affymetrix family rather than unique to this grant.

5. Prior art cited on the face

U.S. 6,214,894 (Zeleny); 6,180,351 (Cattell); 6,140,044 (Besemer); 6,121,048 (Zaffaroni); 6,096,561 (Tayi); 5,961,923 (Nova); 5,945,334 (Besemer — the parent); 5,846,708 (Hollis); 5,807,522 (Brown); 5,800,992. The examiner's primary § 102-type references were thus the applicant's own earlier family members plus the microtiter/matrix-and-memory art (Nova, Zaffaroni) and microarray fabrication art (Brown).

6. Uncertainty statement

• No CAFC 2026 docket activity for 6,399,365 was found, and I am not asserting there is none — only that my searches returned nothing. A patent listed as expired since 2014 would not ordinarily generate new 2026 appellate activity, but I cannot rule out unrelated citation.
• The only litigation record surfaced is the Delaware District Court case 1:04-cv-00901 (Google Patents/Unified Patents) and Darts-IP family ID 27048348. I did not retrieve the pleadings, asserted claims, or outcome.
• Independent claims 1, 7, 10, and 11 are only partially verified in their exact text; the pieces I quote are authoritative, and the remainder is inferred from dependent-claim references. The reference to a claim 8 in dependents 17–20, 24–26, and 33 indicates an independent or intermediate claim I could not retrieve at all.
• Claim 43's recitation of "1…" different polymers per cm² is truncated in the source and should not be quoted as a specific number.

I'll research litigation involving US patent 6399365. Let me start with several targeted searches.

Good start — Affymetrix v. Illumina appears to involve the '365 patent. Let me dig deeper into specific cases.

I've confirmed the primary case. Let me verify additional details, including any appeal and other suits.

Litigation Involving U.S. Patent No. 6,399,365

Based on my searches (Google Patents' litigation links via Unified Patents/Darts-ip, CourtListener, and contemporaneous Delaware docket documents and press coverage), one infringement action is documented in which US 6,399,365 ("the '365 patent") was actually asserted. I found no other suit in which the '365 patent was a patent‑in‑suit.

Case 1 — Affymetrix, Inc. v. Illumina, Inc.

Procedural timeline (as reflected in the docket and press reports):

• Feb. 15, 2006 — Court allowed Illumina to file amended answer/counterclaims adding inequitable conduct, antitrust, and unclean‑hands allegations.
• Mar. 2006 — Affymetrix dropped U.S. 6,607,887 from the case.
• Aug. 16, 2006 — Markman order issued (reported in Affymetrix, Inc. v. Illumina, Inc., 446 F. Supp. 2d 277 (D. Del. 2006)), construing disputed terms in the remaining patents, including the '365 patent.
• March 2007 — Jury verdict in favor of Affymetrix; the jury found Illumina's arrays, scanners, software, and related products infringed one or more claims of all five remaining patents‑in‑suit (including the '365 patent); found a 15% royalty rate and awarded more than $16 million in damages for 2002–2005.
• Jan. 16, 2008 — The parties filed a Stipulated Order of Dismissal (Doc. 471‑2) dismissing all claims and counterclaims with prejudice, each side bearing its own costs and fees.

Later resolution: A separate report (Chinese IP newsletter, citing a $90 million payment by Illumina to Affymetrix and mutual covenants not to sue) indicates the broader dispute was ultimately settled. I could not verify the final settlement terms or any Federal Circuit appeal within my available searches, so I flag those as unconfirmed.

Important exclusions (not '365 patent cases)

To avoid confusion, the following matters surfaced in my search but do not involve US 6,399,365:

• Illumina, Inc. v. Affymetrix, Inc. (W.D. Wis.) concerning U.S. 7,510,841 and 7,612,020 — different patents (summary judgment of non‑infringement for Affymetrix; cases dismissed).
• MIT / E8 Pharmaceuticals v. Affymetrix (D. Mass.) concerning a Housman SNP‑detection patent — different patent.

Caveats

1. Primary source is the Google Patents litigation link, which lists exactly one case for this family: Delaware District Court case 1:04‑cv‑00901 — the Affymetrix v. Illumina action above. Unified Patents' portal entries for US‑6399365‑B2 surface the same Affymetrix family.
2. Some items are drawn from secondary/press sources (Business Wire via BioSpace; the California Biotech Law blog; an LA Times/Bloomberg item; a Chinese IP newsletter). Where a press release rendered the '531 patent as "5,535,531" rather than 5,545,531, I have not auto‑corrected it — the version used in the court's own proposed jury instructions is 5,545,531, and I note the discrepancy.
3. My search was cut short before I could independently confirm (a) any Federal Circuit appeal of the 2007 judgment, (b) the precise date/terms of the final settlement, or (c) whether any additional defendant was ever sued on the '365 patent. I cannot state with high confidence that this is the only litigation ever involving the '365 patent; I can only say it is the only one my searches documented.

Top recommendation for verification: pull the full docket for D. Del. 1:04‑cv‑00901‑JJF via PACER/Unified Patents' litigation portal, and check the Darts‑ip family record (family 27048348) linked from the Google Patents page, to confirm the case list is complete.

I'll start by verifying the canonical PTAB list against public sources.

The canonical block shows no AIA trials. Let me stress-test that with targeted searches before I rely on it.

Key lead: Illumina sought ex parte reexamination of the '365 in 2007. Let me check whether any AIA trial (IPR) exists and confirm the reexam trail.

Proceedings overview

Total AIA trial proceedings on US 6,399,365: 0. The USPTO Open Data Portal "PTAB proceedings on file" block supplied with this task is empty, and my independent searches surfaced no IPR, PGR, or CBM naming this patent as the challenged patent (status breakdown: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0). I could not query PTAB E2E or the PTAB Decisions database directly in this session, so I state the zero count as the canonical ODP position, corroborated but not exhaustively verified.

The bottom-line defensive posture is therefore not "the patent survived IPRs and is hardened" and not "claims 1–5 are canceled." It is a third thing, and it is stronger for a defendant than either: the patent is expired (anticipated expiration 2014-06-08, per the Google Patents legal-status record), and no AIA trial was ever filed against it. A current demand letter citing any claim of the '365 — claims 1–44 appear in the printed patent — is asserting a patent on which essentially no recoverable relief is available, and the absence of any IPR is at least partly explained by the calendar (see below), not by the claims being untouchable.

No proceedings to enumerate

Because the structured list is empty, there are no per-proceeding sections to write: no proceeding number, petitioner, panel, institution decision, FWD, settlement, or appeal exists for me to report. I will not manufacture proceeding numbers, petitioners, or verdicts to fill the template.

What I can document are the non-AIA challenges to this patent. These are strictly outside the requested scope (ex parte reexamination is a central-examination vehicle, not an AIA trial) and I flag them as such because a defendant reading a demand letter will encounter them.

Ex parte reexamination request (NOT an AIA trial) — requested by Illumina, Inc.

• Type: Ex parte reexamination (35 U.S.C. §§ 301–307), pre-AIA regime. Not an IPR/PGR/CBM; no PTAB panel, no estoppel.
• Filed: reported as on or about 2007-10-19, the same date Illumina filed requests against four other Affymetrix patents asserted against it.
• Status: Unconfirmed for the '365. Internet-publication reporting (patentdocs.org, 2008-01-02) states that Illumina had "also filed a request for reexamination of the fifth patent (U.S. Patent No. 6,399,365) asserted by Affymetrix," while the four other requests were ordered on 2007-12-21. I found no source confirming that the '365 request was granted, and no reexamination certificate content for '365. Treat the request as reported-but-unverified and the outcome as unknown.
• Defensive value: if a reexamination certificate exists for '365, the printed claims in the patent as granted may no longer be the operative claims. This is the single highest-value document to pull (see next steps). If no certificate exists, the claims stand as printed.

District court assertion — Affymetrix, Inc. v. Illumina, Inc. (D. Del.)

• Case: 1:04-cv-00901 (D. Del.), filed 2004-07-26, per the litigation link on the Google Patents record and patentdocs.org.
• Patents asserted: five Affymetrix patents; '365 was among them, and was the one Illumina reportedly could not get into reexam.
• Claim construction: the published opinion at 446 F. Supp. 2d 277 (D. Del. 2006) construes claims of the '432 and '243 patents — not the '365. Do not import those constructions onto the '365 claims.
• Defensive value: useful as a roadmap of how Affymetrix pleaded and how Illumina counter-positioned, and as a source of admissions, but it produced no claim-level adjudication of the '365.

Strategic summary

Claim status: entirely UNTESTED at the PTAB, and the whole patent is expired. No claim of the '365 has been canceled, confirmed, or construed by the Board, because no claim was ever challenged in an AIA trial. Every claim — 1 through 44 as printed, subject to the unresolved reexamination question above — stands unadjudicated. Nothing before the PTAB has narrowed this patent, and nothing before the PTAB has blessed it either. For a defendant, the operative narrowing event is not a PTAB decision but the anticipated expiration date of 2014-06-08 shown in the Google Patents legal-status record, which the assignment/security-interest entries (2012 GE Capital lien; 2015 release) are consistent with.

Estoppel landscape: there is none, and that cuts in the defendant's favor. 35 U.S.C. § 315(e)(2) estoppel only attaches to a petitioner that obtained an institution decision. With no IPR petitioner in the chain, no party is estopped, and no prior-art ground has been "used up" at the Board. Every § 102/§ 103 reference — including art that a hypothetical IPR petitioner could have raised — remains fully available. Practically, though, estoppel is not the leverage point here; § 286 is. Any complaint filed today (2026-09-11) can only reach infringement committed on or after 2020-09-11, and the last possible infringing act on an expired patent was 2014-06-08. That leaves zero recoverable damages, and an expired patent cannot support prospective injunctive relief. If the demand letter is aimed at current product sales, the theory is not merely weak, it is temporally impossible.

Pattern signals: the absence of PTAB activity is informative but partly a timing artifact. AIA trials became available 2012-09-16; the '365 was already ~18 years old and expired 2014-06-08. That leaves a roughly 21-month window in which an IPR petition was even practically sensible against a patent whose remaining term was shrinking and whose value was largely in past damages — a poor economic case for a petitioner, which is the most likely reason no IPR exists. There is no repeat-petitioner pattern (no petitioner at all), no Affymetrix PTAB-appeal pattern to report for this patent, and no evidence of a defensive-aggregator challenge. I note that Unified Patents maintains a public patent profile page for US-6399365-B2, but a profile page is a database entry, not a filed proceeding; nothing in it evidences a Unified Patents petition. Do not treat it as one.

Recommended next steps

1. Do not assume a reexam certificate, and check. A patent owner asserting '365 should be required to produce the reexamination certificate (if any) for the '365, or the PTO file history showing termination. Pull it from USPTO Patent Center / the '365 file wrapper and confirm whether the operative claims differ from the printed claims. My sources report only that Illumina filed a request; the outcome is unverified.
2. Lead with expiration, not invalidity. Quote the Google Patents legal-status entry: "2014-06-08 — Anticipated expiration," status "Expired - Lifetime." Combined with the § 286 six-year damages bar, any demand premised on ongoing or recent sales should be rejected outright; any demand premised on 2010–2014 sales faces a § 286 time bar measured from today's date.
3. If past sales are somehow at issue, the invalidity record available to you is richer than the PTAB record: the D. Del. litigation (1:04-cv-00901), the Illumina reexam filings, and the full pre-AIA prior-art landscape are all unencumbered by § 315(e)(2) estoppel.
4. If you want belt-and-suspenders on the zero-PTAB finding, run the case number and patent number directly through PTAB E2E (https://ptacts.uspto.gov/ptacts/) and the PTAB Decisions search at https://www.uspto.gov/patents/ptab/decisions. I was unable to complete that direct query in this session, and an expired-but-recently-asserted patent occasionally draws a late petition; if one exists, it would have been filed before expiration and would change step 2's analysis only for the challenged claims.

Confidence statement. High confidence: no AIA trial was docketed per the ODP structured list, and the patent's anticipated expiration is 2014-06-08. Medium confidence: that no IPR was ever filed (based on searches, not a direct PTAB E2E query). Low confidence / unverified: the disposition of the reported Illumina ex parte reexamination request on the '365. I have not verified the full claim set of the '365, so I have deliberately not stated whether any specific claim is canceled, confirmed, or amended.

I'll research the assignment records before writing anything up.

Let me try to locate the underlying USPTO assignment records and any litigation/ownership context.

I ran out of search budget before I could pull the Assignment Center entries directly, so what follows distinguishes carefully between what is documented and what I could not verify. No correspondent-of-record data was retrievable, and I will not invent it.

Inventors

Named on the face of the patent as listed on the Google Patents record for US 6,399,365. The application (09/907,196, filed 2001-07-17) is a continuation of US 08/485,452 (filed 1995-06-07), which itself claims priority to 1994-06-08, and the same three inventors appear on the sibling European filings (EP 1 562 045 A3 lists Besemer, Goss, Winkler).

Unusual patterns: none of the classic precursors. There is no evidence any inventor departed Affymetrix near filing, and no evidence of a later inventor-side assignment activity. I could not locate per-inventor employment records or exit dates; treat the "presumed Affymetrix" column as an inference from the assignee of record, not a documented fact.

Original assignee

Affymetrix, Inc. (Santa Clara, California) — named as assignee on the issued patent and still shown by Google Patents as current assignee.

• Primary business: manufacture of DNA microarrays / "GeneChip" probe arrays and the instrumentation and software around them (founded 1992 by Stephen Fodor; IPO 1996).
• Product embodying the claims: Yes, in substance. This patent family claims a package/cartridge with a reaction cavity, fluid inlets, seals, and a temperature-control interface for a probe array chip — i.e. the hybridization cartridge and fluidics hardware that Affymetrix shipped with its GeneChip systems (the specification itself describes the chip package being mounted "on a hybridization station" with needles puncturing septa). It is an apparatus/hardware patent, not a bare method claim.
• Current status: Acquired by Thermo Fisher Scientific — announced 2016-01-08, closed 2016-03-31 for ~$1.3B. Affymetrix, Inc. survived as a Thermo Fisher subsidiary (products later rebranded "Applied Biosystems"), which is consistent with no recorded patent assignment to Thermo Fisher appearing on this patent's legal-events record.
• The patent itself is Expired – Lifetime; Google Patents shows an anticipated expiration of 2014-06-08 (20 years from the 1994 priority date).

Assignment timeline

Important caveat up front: I was not able to open the USPTO Assignment Center record for US 6,399,365 during this session. The Google Patents legal-events record for this patent shows only two post-issuance recorded events, and both are security-interest instruments, not ownership transfers. The reel/frame below is corroborated from an INPADOC record for a sibling Affymetrix patent covered by the same portfolio lien; I could not confirm the identical reel/frame against this patent number itself.

• 2012-06-25 (executed) / recorded 2012-06-27 — Reel 028465/0541 (sourced from Espacenet INPADOC for a sibling Affymetrix filing; not verified for this patent number)

  • Conveyance: Security Agreement (liens / collateral, not a transfer of title)
  • Assignor: Affymetrix, Inc.
  • Assignee: General Electric Capital Corporation, as Agent
  • Correspondent: Not retrieved. I could not pull the attorney/firm of record from the Assignment Center; no correspondent is stated in the public legal-events feeds.
  • Context: Securitization. Portfolio-wide lien granted to secure the $100M credit facility dated June 25, 2012 among Affymetrix as borrower, GE Capital as Agent, and Silicon Valley Bank as Syndication Agent (credit agreement text is public in SEC/contract filings). Patent title never left Affymetrix.

• 2015-11-13 (recorded) — Reel/frame not retrieved

  • Conveyance: Release by Secured Party
  • Assignor: General Electric Capital Corporation, as Agent
  • Assignee: Affymetrix, Inc.
  • Correspondent: Not retrieved.
  • Context: Lien release / internal. GE Capital released its security interest, returning Affymetrix to an unencumbered record. This is the termination of a financing lien, not the acquisition of ownership.

• No other recorded assignments. I found no assignment to any LLC, holding company, trust, or licensing vehicle, and no assignment to Thermo Fisher Scientific on this patent's record.

• Original issuance-level assignment (inventors → Affymetrix): standard for a 1994-era Affymetrix filing and consistent with Affymetrix being the assignee of record from issuance, but I did not retrieve the reel/frame for the inventor assignment and will not state one.

Timeline diagram

timeline
    title Ownership of US 6399365
    1994 : Priority date
    1995 : Parent application filed
    2001 : Continuation filed
    2002 : Issued to Affymetrix Inc
    2004 : Family litigation filed in Delaware
    2012 : Security agreement to GE Capital
    2015 : GE Capital lien released
    2016 : Affymetrix acquired by Thermo Fisher

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. No transfer from Affymetrix to any "IP / Licensing / Holdings / Ventures" entity is on the record. The only two recorded events (2012-06-27 security agreement; 2015-11-13 release) are a lien and its discharge, and both name Affymetrix, Inc. as the party with beneficial title.

2. Known asserter in the chain — Not present. Neither Affymetrix, Inc. nor General Electric Capital Corporation is on any public NPE/asserter list (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Round Rock, Spangenberg entities, or the Unified Patents / RPX high-frequency plaintiff lists). GE Capital here is a commercial lender taking collateral — the signature of a secured creditor, not an asserter.

3. Repeat correspondent across the chain — Unclear / insufficient data. I was unable to retrieve the correspondent of record for either entry, so I cannot test for recurrence. This is the one piece of the requested analysis I could not complete from available data.

4. Cascading transfers — Not present. Two events, ~3.4 years apart, both financing-related, both reversing each other. No chained LLC hops, no sub-24-month cluster.

5. Pre-litigation transfer — Not present. There is no assignment within 6 months before any suit. The only litigation flag on this record is a family-level filing in Delaware District Court, case 1:04-cv-00901 (per the Unified Patents litigation data linked from the Google Patents page) — eight years before the 2012 lien and eleven years before the 2015 release, and in the wrong direction. Note: I could not confirm which family members were actually asserted in that case, so treat the 2004 litigation as family context rather than confirmed assertion of this patent.

6. Bankruptcy fire-sale — Not present. No bankruptcy of Affymetrix; it was acquired in a solvent, all-cash $1.3B merger in 2016. No IP sale-in-bankruptcy is on the record.

7. Privateering — Not present. No transfer to a third-party NPE asserting on Affymetrix's behalf. Affymetrix's own enforcement posture (2004 Delaware action, the 2005 Applied Biosystems non-exclusive license) is classic operating-company behavior, and the patent expired in 2014 before any NPE could acquire it usefully.

8. Defensive aggregator — Not present. The chain does not terminate at RPX, AST, LOT, Unified, or OIN. It terminates where it started: Affymetrix, Inc. (now a Thermo Fisher subsidiary under the Applied Biosystems brand).

Verdict

Operating-company assertion.

Justification: the record shows no ownership assignment at all for US 6,399,365 — the only two recorded post-issuance events are the 2012-06-27 security agreement to GE Capital as Agent (reel 028465/0541, collateral for the $100M June 25, 2012 credit facility) and the 2015-11-13 release of that lien, both of which leave Affymetrix, Inc. as the owner. Affymetrix is an operating company that manufactured and sold GeneChip arrays and the hybridization cartridge hardware this patent claims, and it litigated against competitors (family-level Delaware case 1:04-cv-00901) rather than licensing through shells; the 2016 Thermo Fisher acquisition was a solvent cash merger of a going concern, and the patent expired 2014-06-08. Caveat: I could not open the Assignment Center record or retrieve either correspondent of record, so the "repeat correspondent" signal is unresolved rather than negative.

Verify at USPTO Assignment Center: https://assignmentcenter.uspto.gov/ (search patent number 6399365), and cross-check the public legal-events feeds at https://patents.google.com/patent/US6399365/en and the 2004 family litigation entry at https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1%3A04-cv-00901.

I'll verify the '365 patent's front-page references before analyzing them.

Let me pull the actual front-page reference list from the FreePatentsOnline record for 6399365.

Let me verify the Zeleny reference number and pull the foreign/other references.

Prior Art Cited on the Face of US 6,399,365 — Reference-by-Reference § 102 Analysis

Scope and sourcing note. I was not able to query the USPTO's PatentCenter/Patent Full-Text database directly through my tools. The reference list below is built from the FreePatentsOnline record for US 6,399,365 (its "US Patent References" table), cross-checked against Google Patents' citation data for the same document, and against the USPTO Official Gazette text for the sibling patent US 6,733,977 that recites the same priority chain. Where a filing date is not visible in those records, I mark it (filing date unverified) rather than guess.

One correction to the previously generated section. That section listed the Zeleny reference as 6,214,894. The actual citation on the face of the '365 patent is US 6,215,894 B1, "Automatic imaging and analysis of microarray biochips" (General Scanning, Inc.; issued Apr. 10, 2001). Two independent sources confirm the 6,215,894 number. Per the operating rule, I treat the search results as ground truth and flag the earlier figure as an error, not a variant.

1. The verified reference list (all U.S. patents)

A contradiction I could not resolve, flagged rather than papered over: one indexed copy of the document hosted at patentimages.storage.googleapis.com/.../US6399365.pdf is captioned "US 6,399,365 C1" (i.e., a reexamination certificate) and carries an entirely different list of "U.S. PATENT DOCUMENTS" running from US 3,690,836 (1972) through US 5,304,786 (1994), together with non-patent literature including an entry bates-stamped with an "IAFP…" production prefix and a citation to Abstracts, Biochip Arrays Technologies, Fabrication & Applications, May 10, 1995 (Stouffer Mayflower Hotel, Washington, DC). That list does not match the ten references above and is not consistent with a bioarray front page. I could not determine whether this is (a) a genuine ex parte reexamination certificate for the '365 patent with its own considered-references list, (b) an IDS/production compilation from the Affymetrix v. Illumina litigation, or (c) a search-engine content mix-up. I am not treating it as authoritative and I am not asserting that a C1 certificate exists. It should be verified against the USPTO file wrapper before any reliance.

2. The controlling legal issue: reference dates vs. effective filing date

This is the threshold question and it dominates the § 102 analysis.

The '365 patent issued June 4, 2002 from App. No. 09/907,196, filed July 17, 2001. It is a straight continuation chain:

09/907,196 (2001) ← 09/302,052 (Apr. 29, 1999, now US 6,287,850) ← 08/485,452 (Jun. 7, 1995, now US 5,945,334) ← 08/255,682 (Jun. 8, 1994, abandoned — the CIP parent)

So claims supported by the '452 disclosure get a June 7, 1995 date; claims supported only by new matter added in the '452 get that date and claims supported by the '682 get June 8, 1994. (The file wrapper would be needed to allocate subject matter between the '682 and the '452; I could not retrieve it.)

Now compare reference dates:

• US 6,215,894 (Zeleny) — filed Feb. 26, 1999.
• US 6,180,351 (Cattell) — issued Jan. 30, 2001; application date in the 1998–99 window (filing date unverified).
• US 5,961,923 (Nova) — issued Oct. 5, 1999; filing date in the mid-1990s (unverified).
• US 5,800,992 (Fodor) — issued Sept. 1, 1998, from a June 7, 1995 filing.

Consequence: several of these references were filed or issued after the June 1994/June 1995 priority dates. They cannot be § 102(a) or § 102(b) art against claims entitled to those dates. They are available only as § 102(e) art — and § 102(e) requires both (i) that the reference be "by another," and (ii) that the '365 claim in question not be entitled to a date earlier than the reference's § 102(e) date. Zeleny, filed February 1999, is the clearest example: it is § 102(e) art only if the asserted claim is supported solely by matter added no earlier than the 1999 applications, which is a difficult position given that the bar-code language appears in the shared specification.

I flag this because an examiner's listing of a reference on the face of a patent is not a finding of anticipation. Most of these were almost certainly cited as background or as § 103 art, and I do not have the office actions to confirm the statutory basis actually applied.

3. Reference-by-reference analysis

1. US 6,215,894 B1 — Zeleny et al., "Automatic imaging and analysis of microarray biochips"

• Citation: US 6,215,894 B1; General Scanning, Inc.; filed Feb. 26, 1999; issued Apr. 10, 2001.
• Description: A scanner system for microarray biochips in which a machine-readable identifier (explicitly a bar code, preferably also human-readable numerals) is imprinted on the biochip and logically linked to a computer folder holding that experiment's parameters, reagents, fluorescent tags, and scanning/analysis protocols. The scanner reads the bar code, auto-retrieves the protocol, sets itself up accordingly, scans, and stores the resulting image map under the identifier.
• Potentially anticipatory of: Claim 41 is the closest read — its steps of "providing a bar code associated with said probe array; reading said bar code; aligning said probe array with a detection system; and detecting a signal from said probe array" map almost element-for-element onto Zeleny's disclosure. Claim 2 (housing including a bar code, arranged for use with a detection system) is also implicated, though Zeleny's identifier is on a bare microscope slide, not on a housing bearing a fluid cavity — so claim 2's housing/cavity element is unmet.
• Candid caveat: Zeleny does not disclose a sealed hybridization cavity, septa, ultrasonic welding, or a temperature controller, so it cannot anticipate the package claims (1, 10, 11). And its Feb. 1999 filing date is post-priority (see § 2).

2. US 6,180,351 B1 — Cattell et al., "Chemical array fabrication with identifier"

• Citation: US 6,180,351 B1; issued Jan. 30, 2001 (filing date unverified; 1998–99 window).
• Description: Fabrication of chemical arrays (biopolymer arrays) on a substrate together with an identifier — e.g., a bar code — carried on the array substrate for tracking and later association of the array with its intended chemistry; contemplates reading the identifier to correlate an array with the experiment.
• Potentially anticipatory of: Claims 2 and 41 on the identifier/bar-code element. Like Zeleny, it addresses the substrate-borne identifier concept but not the reaction-chamber-borne identifier that is the commercial core of claim 2, and not the density or housing limitations.
• Timing: Same post-priority problem as Zeleny.

3. US 6,140,044 A — Besemer, Winkler & Goss, "Method and apparatus for packaging a probe array" — not § 102 art

• Citation: US 6,140,044 A; Affymetrix, Inc.; issued Oct. 31, 2000; same 1994-06-08 family root.
• Why it is not prior art: The inventive entity is identical to the '365 patent (Besemer, Goss, Winkler). A reference is not "by others" under § 102(a) or § 102(e), and a patent to the same inventors on the same disclosure cannot be § 102(b) art either. It is at most a double-patenting / § 103 consideration (and would be commonly owned for § 103(c) purposes). No anticipation of any claim.

4. US 6,121,048 A — Zaffaroni et al., "Method of conducting a plurality of reactions"

• Citation: US 6,121,048 A; issued Sept. 19, 2000 (filing date and assignee unverified in my sources).
• Description: Conducting multiple chemical/biological reactions in a plurality of reaction regions on a single device with reagent introduction and removal — multi-chamber fluid handling art.
• Potentially anticipatory of: In principle only the most generic "housing with a fluid chamber" concept in claims 1 / 10 / 11, and even then only if the claim were read to omit the probe array — which it does not. Zaffaroni discloses reaction chambers, not immobilized probe arrays inside a cavity covered by a chip. I assess anticipation as very unlikely; this is § 103 material at best.

5. US 6,096,561 A — Tayi, "Scheduling operation of an automated analytical system"

• Citation: US 6,096,561 A; issued Aug. 1, 2000; classified in 436/518.
• Description: Scheduling and resource-management software for an automated clinical analyzer (sample/reagent throughput scheduling).
• Potentially anticipatory of: Nothing. This is automated-analyzer logistics art with no probe array, no bar-coded array, no hybridization cavity. It cannot anticipate any of claims 1–43. It is background showing the "high-throughput analyzer" context the specification invokes.

6. US 5,961,923 A — Nova, Senyei et al., "Matrices with memories and uses thereof"

• Citation: US 5,961,923 A; issued Oct. 5, 1999; US Class 422/68.1 (filing date unverified).
• Description: Solid-support matrices (beads/supports) each bearing machine-readable memory/identification so that a reaction history is retained and travels with the individual support element — the "identifier travels with the material" concept.
• Potentially anticipatory of: The identifier-associated-with-analyte element of claims 2 and 41, at a conceptual level. It does not disclose a hybridization cavity, a planar high-density probe array, or detection of a signal from an array.
• Assessment: Weak as anticipation; realistically § 103 art. Its issuance date (1999) is again post-priority for the 1994/1995 claims.

7. US 5,945,334 A — Besemer et al., "Apparatus for packaging a chip" — not § 102 art

• Citation: US 5,945,334 A; Affymetrix, Inc.; filed June 7, 1995 as Ser. No. 08/485,452; issued Aug. 31, 1999. This is the direct priority parent of the '365 patent.
• Why it is not prior art: Same three inventors (Besemer, Goss, Winkler) and, more fundamentally, the '365 patent is a continuation of the same disclosure. A parent application's own disclosure cannot anticipate its continuation — the continuation claims, by definition, are supported by it and take its benefit date. § 102(b) requires a patent "by another"; this is not. No anticipation of any claim.
• Title discrepancy to note: FreePatentsOnline and the USPTO record title '5,945,334 as "Apparatus for packaging a chip." A Google Patents snippet rendered the same number under the title "Hybridization device and method" — which is the title of the sibling US 6,733,977. I read the FPO/USPTO title as correct and note the mis-association rather than adopting it.

8. US 5,846,708 A — Hollis et al., "Optical and electrical methods and apparatus for molecule detection"

• Citation: US 5,846,708 A; issued Dec. 8, 1998 (filing date unverified).
• Description: Detecting molecules by optical and electrical (electronic) means — sensing hybridization/detection events on a chip-scale device.
• Potentially anticipatory of: The detection element of claims 11 (claim 16, reciting an alignment structure for a detection system that detects fluorescent light emitted from the array and transmitted through the chip) and 41 ("detecting a signal"). It does not disclose the back-side/transmission-mode geometry recited in 16, nor the bar code, nor the housing.
• Assessment: § 103 background. Not anticipatory.

9. US 5,807,522 A — Brown & Shalon, "Methods for fabricating microarrays of biological samples" — the strongest § 102 candidate

• Citation: US 5,807,522 A; Leland Stanford Junior University; issued Sept. 15, 1998 (application filed in the 1994–95 window; commonly reported as June 7, 1995 — filing date unverified).
• Description: The "Stanford" spotted-microarray method — mechanically depositing biological samples (cDNAs, etc.) as array elements on a substrate to produce a microarray, with arrays of high element count. This is the classic § 102 platform reference for high-density biological-polymer arrays per se.
• Potentially anticipatory of:
  • Claim 7 (array of biological polymers on a substrate) and its density sub-claims 35 / 36 (>400 and >1,000 different nucleic acids per cm²), if Brown's examples are shown to reach those densities — Brown's microarrays are typically many thousands of spots, so this is plausible on the density element.
  • Arguably claim 1 / claim 10 / claim 31 / claim 32 on the density element alone.
• Why it likely fails: Brown uses mechanical spotting, so dependent claim 23 (polymers attached "by selectively illuminating said substrate," i.e., light-directed synthesis) is not anticipated. More importantly, Brown discloses no housing, no fluid cavity, no seal, no bar code and no method of reading a bar code — so it cannot anticipate claims 1, 2, 10, 11 or 41 as a whole. It potentially anticipates only the bare-array claim 7 if claim 7 carries no housing limitation.
• Timing wrinkle: If Brown was filed June 7, 1995 — the very day 08/485,452 was filed — then against claims entitled to the June 8, 1994 '682 date, Brown's § 102(e) date is later than the claimed priority date, and Brown would not be § 102(e) art at all. Against claims entitled only to June 7, 1995 it is a same-day problem. This must be resolved from the actual Brown filing receipt and the '365 file history.

10. US 5,800,992 A — Fodor et al., "Method of detecting nucleic acids"

• Citation: US 5,800,992 A; Affymetrix, Inc.; issued Sept. 1, 1998, from a June 7, 1995 filing in the VLSIPS™ family.
• Description: Detecting nucleic acid targets by hybridizing them to a high-density array of oligonucleotide probes at known locations on a substrate and reading out a labeled (fluorescent) signal.
• Potentially anticipatory of: The hybridization-and-detection core of claims 10 and 11 and the "hybridizing a target to a probe... and detecting a signal" steps of claim 41. It is a strong § 102(e) reference on the array-side limitations (density >100, >400, >1,000 per cm²; known locations; fluorescent detection), and it squarely supports the "different biological polymers/sequences" element.
• Why it likely fails: It does not disclose the bar code (claims 2, 41), the multi-piece welded housing with a fluid cavity (claims 1, 2, 10, 11), the alignment structure keyed to a detection system (claim 11/16), or the method step of reading a bar code (claim 41). It is therefore a § 102(e)/§ 103 reference on the array-half of the claims, not a full anticipant.
• "By another" status: Inventors are Fodor et al., not Besemer et al., so it is "by another" notwithstanding common assignment — but common ownership again limits the § 103(c) use.

4. Foreign references and other documents

I could not retrieve the "Foreign Patent Documents" or "Other References" blocks for US 6,399,365 from the sources available. What I can say with confidence about the family's foreign filings:

• WO 95/33846 A1, "Bioarray chip reaction apparatus and its manufacture" (published Dec. 14, 1995) is the PCT counterpart of this work.
• EP 0 695 941 A1/B1, "Method and apparatus for packaging a chip" (published Feb. 7, 1996) is the corresponding EPO filing.

Both are the applicant's own parallel family filings on the same June 1994 priority, and therefore are not § 102 prior art against the '365 claims — they are the same invention by the same inventive entity. If the '365 patent's face in fact recites either of these, they should be read as family/priority cross-references, not as anticipating references.

The unattributed PDF listing (US 3,690,836 through 5,304,786 plus the IAFP-stamped non-patent literature) is, if genuine, a much older and broader mechanical/fluid-handling art set — largely 1970s–1980s cartridge, cuvette and specimen-processing patents — which would be § 103 background at best. I have not verified it and I am not treating it as part of the '365 face.

5. Bottom line

The defensible high-level conclusion: this reference set was assembled as a § 103 combination, not as a § 102 anticipation set. The array art (Brown, Fodor, Hollis) supplies the probe-array and detection halves; the identifier art (Zeleny, Cattell, Nova) supplies the bar-code/tracking half; the microfluidics art (Zaffaroni, Tayi) supplies the automated-reaction context; and the applicant's own earlier patents (Besemer '334/'044) are the priority parents rather than prior art. The novelty that survived into the 2007 jury verdict in Affymetrix v. Illumina (D. Del. 1:04-cv-00901, asserted claims 36 and 41) is precisely the combination — a bar-coded, alignment-keyed hybridization cartridge — that none of the cited references individually shows.

6. Uncertainty statement (read before relying)

• Filing dates marked "(filing date unverified)" — Cattell, Nova, Hollis, Zeleny's claim-level date, Brown's application date, Zaffaroni, Tayi and Fodor's exact application dates — must be pulled from the actual front pages before any § 102(e) date conclusion is drawn. I have given issue dates I could verify and flagged the rest.
• The effective-filing-date allocation between the June 8, 1994 '682 application and the June 7, 1995 '452 application is unresolved. Every § 102 analysis above turns on it. It requires the '365 file wrapper (and the '682/'452 disclosures) to settle.
• I did not retrieve the office actions, so I cannot state which references the examiner applied under § 102 versus § 103, or whether any rejection was made and overcome. The phrase "potentially anticipates" in my tables is an analytical judgment about disclosure content, not a statement of what the examiner held.
• The "US 6,399,365 C1" reexamination-certificate PDF and its divergent reference list are unresolved and may be a litigation document-production artifact rather than a prosecution document. Verify against the USPTO file wrapper.
• The prior generated section's reference to "6,214,894 (Zeleny)" is, on the evidence of two independent sources, a misnumbering of US 6,215,894. Separately, that section's characterisation of the examiner's primary references as "the applicant's own earlier family members plus the microtiter/matrix-and-memory art (Nova, Zaffaroni) and microarray fabrication art (Brown)" is consistent with what I verified, and I have built on it rather than restated it.
• Minor housekeeping: the task header gives the current date as April 26, 2026 while my operating context is September 11, 2026. Nothing in this analysis depends on which is correct.

Coverage note: this section addresses prior art only. It does not revisit the bibliographic, claim-construction, or litigation sections generated earlier, and it does not opine on validity, infringement, or the Affymetrix v. Illumina outcome beyond what prior art scope bears on.

I'll pull authoritative data on the references cited on the face of the '365 patent before doing the §103 analysis.

Critical finding in that data — a reexamination. Let me verify that and pull the remaining references.

Obviousness Analysis — U.S. 6,399,365 B2 under 35 U.S.C. § 103

Scope note. I worked from the ten references listed in § 5 (Prior art cited on the face) of the previously generated summary, plus the two references the '365 specification itself incorporates by reference (Pirrung 5,143,854; Fodor 5,800,992, which appears on the face as well). My searches were cut off before I could retrieve the disclosures of Zeleny 6,214,894, Cattell 6,180,351, Zaffaroni 6,121,048, Tayi 6,096,561, Nova 5,961,923, and Hollis 5,846,708 — so for those I rely on the face-of-patent listing and the earlier section's characterization, and I flag where my analysis would change with their actual text. Everything I state as a disclosure of Brown, Fodor, Besemer 5,945,334, Besemer 6,140,044, and WO 95/33846 is from retrieved text.

0. Two corrections to the earlier sections — flagged per operating rules

(a) A reexamination occurred, and claims 7, 41, and 43 are gone. The previously generated summary analyzed claim 7 (the array claim) and quoted claim 41 (the method claim) as live claims. That is contradicted by the patent document itself. The indexed PDF of US 6,399,365 contains a reexamination certificate stating:

"AS A RESULT OF REEXAMINATION, IT HAS BEEN DETERMINED THAT: Claims 7, 34–37, 40–45, 55 and 58 are cancelled. Claims 1–6, 8–33, 38–39, 46–54 and 56–57 were not reexamined."
— https://patentimages.storage.googleapis.com/b9/50/42/f7f72dd9ad13e0/US6399365.pdf

This is corroborated by the reexamination request itself: Illumina filed ex parte reexamination requests on October 19, 2007 against four asserted Affymetrix patents, and, per patentdocs.org, "Illumina … has also filed a request for reexamination of the fifth patent (U.S. Patent No. 6,399,365) asserted by Affymetrix" (https://patentdocs.org/2008/01/02/uspto-to-reexam-2/). Caveat: I saw only a truncated snippet of the certificate and did not retrieve the grounds of rejection or the date of the certificate. The cancellation language could reflect an examiner's § 103 rejection, the patentee's own cancellation, or a mix — I cannot say which. But the cancellation of claim 7 and claims 41–45 is the single most probative fact available here, because those are precisely the array-per-se and method-of-use claims, and they are the claims the prior art most cleanly reaches.

(b) The priority date is outcome-determinative, and the earlier section did not resolve it. The '365 is a continuation of 09/302,052 (now 6,287,850) ← continuation of 08/485,452 (now 5,945,334) ← continuation-in-part of 08/255,682, filed June 8, 1994 (per the (63) data in the patent PDF). If the claims get the 1994-06-08 benefit, the § 102(b) critical date is 1994-06-08 and Brown 5,807,522 (filed June 7, 1995), the 5,945,334 (issued Aug. 31, 1999), 6,140,044 (issued Oct. 31, 2000), and WO 95/33846 (published Dec. 14, 1995) all fall outside § 102(b). Conversely, if the bar-code and ">100 different polymers/cm²" limitations lack § 112 support in the 1994/1995 applications, the effective filing date is July 17, 2001, and every one of those references becomes § 102(b) art. A challenger will litigate this first. Note that the listed expiration date of June 8, 2014 implies the 1994-06-08 date was honored — but expiration and prior-art status are separate questions.

Consequence for this analysis: I therefore build the § 103 case primarily on art that is prior regardless of which priority date controls — the pre-1994 Affymetrix synthesis art (5,143,854, 5,800,992, PCT WO 92/10092) and the flow-cell/assay-format art — and then show the incremental exposure from each additional reference if the later date controls.

1. Legal standard and the level of ordinary skill

Framework. Graham v. John Deere Co., 383 U.S. 1 (1966) (scope/content of art; differences; level of skill; secondary considerations); KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) (the "predictable use of prior art elements according to their established functions"; a known problem in the field supplies the reason to combine; a "finite number of identified, predictable solutions" makes a combination obvious; "obvious to try" where the prior art gives a finite set of identified options). The '365 is a pre-AIA patent, but a reexamination in 2008–2010 would have been conducted under KSR.

Level of ordinary skill (as of 1994–1995). A person having ordinary skill would hold a B.S. or M.S. in mechanical/biomedical engineering, analytical chemistry, or a biological science, with 2–5 years' experience in laboratory reaction-vessel or microfluidic cartridge design, or a Ph.D. in molecular biology/biochemistry with hands-on experience running solid-phase hybridization assays and the instruments that read them. Such a person would know: (i) photolithographic ("VLSIPS") array synthesis on planar substrates (Pirrung/Fodor); (ii) capillary/spotted microarray fabrication (Brown); (iii) fluorescent scanning of planar arrays through the substrate (Fodor; the '734 scanner); (iv) microtiter plates, flow cells, and sample cassettes with inlet/outlet ports and septa; (v) injection-molded part design and ultrasonic welding (ordinary mechanical skill); and (vi) bar-code labeling of laboratory vessels in clinical and research automation (ubiquitous by 1994).

2. What each reference supplies

Because I could not retrieve six of the ten references, I separate verified from unverified.

Verified (retrieved disclosure):

Cited on the face but NOT verified by me (disclosure unknown): Zeleny 6,214,894; Cattell 6,180,351; Zaffaroni 6,121,048; Tayi 6,096,561; Nova 5,961,923; Hollis 5,846,708. The earlier section characterized Nova and Zaffaroni as "microtiter/matrix-and-memory art" and Hollis as optical-detection art; I could not confirm that, and I do not rely on it below except where noted as a hypothesis.

Also available as admitted background: 5,143,854 (Pirrung) and PCT WO 92/10092 — the '365 specification calls these "the pioneering techniques" for forming sequences on a substrate, and incorporates them by reference. These are 1992-era publications and are prior art on any priority theory.

3. The combinations

Combination 1 — The array/density claims: Brown 5,807,522 + Fodor 5,800,992 (+ 5,143,854)

What it covers: every claim reciting a substrate bearing an array of different biological polymers at known locations at a density exceeding 100 (and 400, and 1,000) per cm² — i.e., cancelled claims 7, 34–37 and the density sub-limitations of the surviving claims (21–22, 27–28, 31–32).

Why it is obvious:

• Fodor 5,800,992 claim 5 expressly recites "at least 10³ distinct polynucleotide in a surface area of about 1 cm²" at "separate, defined position[s]." That is a literal disclosure of the '365's most aggressive density recitation (>1,000 nucleic acids/cm²). There is no numerical gap to bridge — the reference exceeds the claim.
• Brown supplies the alternative fabrication route (mechanical deposition rather than light-directed synthesis) and expressly frames the array as "discrete analyte-assay regions … each … a selected, analyte-specific reagent," i.e., different polymers at known positions.
• Motivation: both references are in the same field (high-throughput nucleic-acid analysis), address the same problem (how to interrogate many sequences in parallel on one support), and Brown's own stated purpose is the automated production of such arrays. A PHOSITA with Brown's apparatus and Fodor's demonstrated densities would have had every reason to use either to make a >1,000/cm² array. This is the KSR paradigm: the elements (solid support + immobilized diverse probes + known positional addressing) each perform their established function, and the only variable is a density figure that the art already exceeded.

Why this matters: this combination is presumably exactly what the examiner applied in reexamination, which is the most plausible explanation for the cancellation of claim 7 and claims 34–37/40.

Combination 2 — The package/housing claims: {Brown 5,807,522 or Fodor 5,800,992} + {a fluid-chamber reference from the face of the patent}, in view of the flow-cell art admitted in the '365 specification

What it covers: the "housing having a fluid cavity … constructed and arranged for hybridization … located inside the cavity" element of claims 1, 2, 8/10–11, 46–54, and 56–57; and the "arrays in fluid communication" limitation (claim 29).

The admission in the applicant's own specification. The '365 specification quotes U.S. 5,631,734 (itself incorporated by reference) as follows:

"The substrate may be mounted on a flow cell. The flow cell includes a body having a cavity on a surface thereof. The cavity is between about 50 and 1500 µm deep … The bottom of the cavity is preferably light absorbing … The substrate seals the cavity except for an inlet port and an outlet port."

This is an applicant admission, in the specification, that a body + cavity + inlet port + outlet port + substrate-sealing-the-cavity + light-absorbing cavity floor assembly was known. (MPEP § 2129; an applicant's characterization of the art in the specification is usable evidence of what a PHOSITA knew.) The only thing the '365's housing adds over this admitted flow cell is plasticity of form — injection molding, ultrasonic welding, and casing geometry — which is ordinary mechanical engineering.

Why it is obvious:

• The problem was expressly identified in the art. 6,140,044 and WO 95/33846 both state the object: "a commercially feasible device for use in high throughput assay systems," and the family's background (retrieved from the sibling 2010/0298165 publication) states: "As commercialization of such substrates becomes widespread, an economically feasible and high-throughput device and method for packaging the substrates are desired."
• The '734 flow cell, the microtiter-plate art (Zeleny/Cattell/Zaffaroni/Tayi/Nova, per the face), and the prior chip-packaging patents (5,945,334, 6,140,044) bracket the solution on both sides.
• Motivation, in KSR terms: the known problem ("how do we run hybridization on a planar array in a controlled, sealed, thermostatted volume?") directly supplied the reason to combine the array of Brown/Fodor with the chamber architecture of the flow-cell/microtiter art. The results were entirely predictable: sealed fluid volume, temperature control, septum-pierced ports, and light-absorbing cavity floor were each known techniques applied to their known functions (MPEP § 2144.03).
• The finite number of identified solutions point is unusually strong here: the '365's own FIGS. 18–27 enumerate five alternative chip-mounting schemes (two concentric ledges + adhesive; adhesive film/tape; clamp with fingers + gasket; etc.). An inventor who lists five interchangeable mounting approaches in one disclosure is demonstrating that the mounting problem had a small, predictable solution space.

Combination 3 — The bar-code claims: Combination 2 + the established practice of machine-readable identification on laboratory vessels

What it covers: claim 2 ("the housing also includes a bar code" + "arranged for use with a detection system"); cancelled claim 41 ("providing a bar code associated with said probe array; reading said bar code; aligning said probe array with a detection system").

Why it is obvious:

• The specification itself treats the identifier as an optional packaging convenience, not a technical contribution: "the top casing is of sufficient size to accommodate identification labels or bar codes in addition to the cavity." Under KSR, where the patentee frames a feature as a desirable-but-conventional option, that is strong evidence the feature is the "predictable use of prior art elements according to their established functions."
• Bar-code labeling of sample vessels, microtiter plates, and slides in automated clinical and genomic workflows was widespread in the early 1990s. The claimed function — read the code, then align and scan the array, so the instrument knows which array it is looking at — is precisely the function bar codes already served in laboratory automation.
• The limitation is not vulnerable to a "printed matter" rejection (cf. In re Gulack), because claim 2 requires the housing be "arranged for use with a detection system" and claim 41 requires the code to be read and the array aligned — a functional cooperation between the indicium and the apparatus. So the correct attack is § 103 over known labeling practice, not § 101/printed matter.

Honest weakness in this combination: the face of the '365 does not appear to include a bar-code-specific reference. I verified Nova, Zaffaroni, Tayi, Cattell, and Zeleny citation on the face but not their disclosures. Whoever asserts invalidity of claim 2 must supply a barcode/housing reference or rest on official notice / "known technique" reasoning. Notably, claim 2 falls in the 1–6 range that the 2008 reexamination did not reach — i.e., this combination was apparently never tested by the PTO. That is a real evidentiary gap, and I flag that claim 2 is likely the most defensible claim in the patent on this record.

Combination 4 — The fluorescence-through-the-chip limitation: Combination 2 + Fodor 5,800,992 + the 5,631,734 detection disclosure

What it covers: the claims requiring the array signal to be detected through the chip (claim 16 per the earlier section: detection system "constructed to detect fluorescent light emitted from the array and transmitted through the chip") and the "optically transparent chip" element of claim 2.

Why it is obvious: Fodor 5,800,992 teaches fluorescent-label hybridization detection; the '365 specification's own recitation of the 5,631,734 disclosure teaches a laser "directed at the inner surface of the substrate," a dichroic mirror, a confocal pinhole, and a photomultiplier tube. Glass microscope slides and cover slips as transparent array substrates are expressly described in that same incorporated passage ("the substrate is made of a conventional microscope glass slide or cover slip … about 0.7 mm thick"). Back-side (transmission-mode) excitation of a transparent array substrate is therefore squarely in the admitted art. The combination adds nothing beyond selecting a transparent substrate so light can pass through it — a result dictated by the physics of the detection method, not an inventive choice.

Combination 5 — Double patenting / the family references (not a § 103 combination — flag this)

The 5,945,334 and 6,140,044 are cited on the '365 face, but they share the same inventive entity (Besemer, Goss, Winkler) and the same assignee. Under pre-AIA § 102(a)/(e), a reference must be "by another"; same-inventive-entity family members are therefore not § 102(a)/(b)/(e) prior art against the '365 if the '365's claims are entitled to the 1994/1995 priority. Their real exposure is obviousness-type double patenting across the continuation chain, which is a separate doctrine and would turn on whether the '365's claims are patentably distinct from the 5,945,334/6,287,850 claims given a substantially identical specification. I am explicitly not asserting a § 103 theory grounded on these family members as "prior art," because the "by another" requirement blocks it on the priority facts as I understand them. If, however, the priority chain fails (see § 0(b) above), these references flip into § 102(b) art and Combination 1 collapses into near-anticipation of the array and package claims.

4. Claim-by-claim mapping

5. Rebuttals a patentee would raise, and how they fare

1. "The family references are not prior art." Correct, if the 1994-06-08 benefit holds — see § 0(b). But this does not save the array claims, because 5,143,854, 5,800,992, and WO 92/10092 are pre-1994 prior art regardless.
2. "The references are non-analogous." Fails. Brown, Fodor, the microtiter art, and the flow-cell art are all in the same field of endeavor — high-throughput analysis of biological samples on solid supports. In re Bigio / In re Clay do not help.
3. "No motivation to combine." Weak against KSR. The problem was expressly articulated in the prior art (6,140,044: "a commercially feasible device for use in high throughput assay systems"), the solution set was small and identified (FIGS. 18–27 of the patent itself), and the design constraints (sealed volume, bubble removal, temperature control, correct orientation in the reader) were all known engineering problems.
4. Secondary considerations. If the patentee invokes the 2007 Delaware jury verdict (Affymetrix v. Illumina, C.A. 1:04-cv-00901-JJF) as objective evidence of non-obviousness, note that (i) a jury finding of infringement is not a finding of validity unless the jury was asked and answered that question — I could not verify from my sources that it was; (ii) any commercial-success or licensing evidence must have a nexus to the claimed subject matter, and the reexamination cancelled the array and method claims, undercutting nexus for those; and (iii) the case ended in a stipulated dismissal with prejudice on January 16, 2008, months after the reexamination requests were filed — consistent with a negotiated resolution rather than appellate affirmance of validity. I could not verify any Federal Circuit appeal or the final settlement terms; treat the "fully adjudicated validity" narrative as unconfirmed.
5. Bar code. The strongest surviving defense. The bar-code element was apparently not challenged in the 2008 reexamination, and I have not located a bar-code-specific reference among the face-of-patent citations. A challenger must either find one or persuade the tribunal to take official notice of the ubiquity of bar-coded laboratory vessels circa 1994.

6. Bottom line

• The prior art renders the array-per-se and method-of-use subject matter obvious. Fodor 5,800,992 claim 5 recites >10³ distinct polynucleotides in ~1 cm² at defined positions — a literal disclosure exceeding the '365's >100, >400, and >1,000 per cm² recitations — and 5,143,854/WO 92/10092 (expressly acknowledged as "pioneering techniques" in the '365's own background) supply light-directed synthesis at known locations. Brown 5,807,522 supplies the alternative mechanical route. This is consistent with, and most plausibly explains, the cancellation of claims 7, 34–37, and 40–45 and claim 58 in reexamination.
• The package claims are the invention's actual contribution, and they are the claims the reexamination did not reach (1–6, 8–33, 38–39, 46–54, 56–57). Their § 103 exposure comes from the flow-cell/microtiter art combined with the admitted 5,631,734 flow cell, plus ordinary mechanical skill — but on the references on the face of this patent, that case is materially weaker than the case against the array and method claims.
• Claim 2 (bar code) is the most defensible claim on this record, because the absence of a bar-code reference from the cited art was apparently never remedied in reexamination.
• Everything hinges on the priority date. If the '365's claims do not get the June 8, 1994 benefit, 5,945,334, 6,140,044, WO 95/33846, 5,807,522, and 5,800,992 all become § 102(b) art and the package claims become very hard to defend.

Verification steps I recommend before anyone relies on this: (1) pull the full reexamination certificate and the examiner's reasons for cancelling claims 7, 34–37, 40–45, 55, 58 from the '365 file wrapper (reexam request filed by Illumina ~Oct. 19, 2007); (2) retrieve the disclosures of Zeleny 6,214,894, Cattell 6,180,351, Zaffaroni 6,121,048, Tayi 6,096,561, Nova 5,961,923, Hollis 5,846,708 — my searches were truncated before I could, and my analysis does not depend on their content; (3) check the D. Del. 1:04-cv-00901-JJF docket for whether the jury was asked a validity question and for the final stipulated dismissal terms (Doc. 471-2).