US 7087732

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

Let me run a few more targeted searches to check for any litigation or 2026 docket activity.

Let me get the actual claim text and abstract to be precise.

US 7,087,732 — Identification Summary

Sources: the authoritative full text at https://patents.google.com/patent/US7087732/en; bibliographic confirmation at https://patents.justia.com/patent/[7087732](/patent/7087732) and https://uspto.report/patent/grant/7,087,732.

Note on the priority line. The record is internally inconsistent in a way you should not auto-correct: the priority-date field reads 1989-06-07, while the priority claims of record are to US 07/492,462 and US 08/456,887. Both are reported verbatim above. US 07/492,462 is the application that issued as US 5,143,854 (Pirrung et al.), the ancestral VLSIPS patent; US 08/456,887 issued as US 6,406,844. The '732 patent sits in that continuation chain.

Abstract

⚠️ Uncertainty flag: the fetched full text was truncated before the abstract and claims sections, so I do not have the verbatim abstract or claim text of US 7,087,732 in front of me. Based on the description text that is available, the disclosure is a light-directed, spatially-addressable parallel chemical synthesis platform built on solid-phase chemistry, photolabile protecting groups, and photolithography, and the title indicates the claimed subject matter is directed to nucleotides/nucleosides and analogs bearing photoremovable protecting groups. I am not asserting a verbatim abstract; confirm against the official USPTO document before relying on exact wording.

Plain-language overview of the invention (from the description): Build many different polymers (peptides, oligonucleotides, etc.) at known, addressable locations on a single substrate. The substrate surface carries reactive groups blocked by photolabile protecting groups. Shining light through a mask removes the protecting groups only in illuminated regions, exposing reactive sites. A monomer (e.g., an amino acid or nucleotide) that itself carries a photolabile protecting group is then flowed over the surface and couples only where light struck. Repeating the illuminate → couple → wash cycle with different masks and different monomers builds a spatially addressable array of diverse sequences. The disclosure also covers binary/gray-code masking strategies for efficient combinatorial synthesis, a flow-cell reactor and fluidics system, data-collection instrumentation, and statistical data-analysis techniques for reading out binding affinities.

Independent Claims — Plain-Language Overview

⚠️ Important limitation on this section. The claims were not included in the retrieved text. What I can tell you is that the recorded claim-language fragments on the Google Patents record indicate this patent issues claims across several statutory classes. The patent's own description recites, in claim-style language ("...is provided", "...comprising the step of..."):

1. Compound claim (the titular subject matter). A compound comprising a nucleotide or nucleoside (and analogs) bearing a photoremovable protecting group, exemplified by a substituted o-nitrobenzyl-type structure of the general form 1*–C₆(R₁)(R₂)(C(R₃)OC(=O)Y)–NO₂, where:

• Y is the oxygen of the carboxyl of a natural or unnatural amino acid, the amino group of a natural or unnatural amino acid, or the C-5′ oxygen of a natural or unnatural DNA/RNA;
• R₁ and R₂ independently are hydrogen, lower alkyl, aryl, benzyl, halogen, hydroxyl, alkoxyl, thiol, thioether, amino, nitro, carboxyl, formate, formamido, sulfido, or phosphido;
• R₃ is alkoxy, alkyl, aryl, hydrogen, or alkenyl.
  A parallel formula fragment recites a side-chain R of a natural or unnatural amino acid with X being a photoremovable protecting group (peptide-side variant).

1. Method of screening linker polymers. Forming a plurality of linker polymers in selected regions of a substrate by recursively irradiating selected regions to remove a protective group and contacting the surface with a monomer; then contacting the resulting linker polymers with a ligand, and contacting the ligand with a labeled receptor.
2. Ordered method of forming polymer sequences (binary synthesis). Serially protecting and deprotecting portions of the polymer sequences to add further portions using a binary synthesis strategy.
3. Apparatus/system claim. A system for determining receptor–ligand affinity: means for applying light to a substrate bearing ligands at predetermined locations, with simultaneous illumination of multiple locations, plus an array of detectors detecting light fluoresced at those locations.
4. Data-analysis method claim. Exposing fluorescently labeled receptors to a substrate bearing ligands at known locations; at multiple data-collection points within each region, measuring fluoresced light; discarding data-collection points deviating from a predetermined statistical distribution; and determining relative binding affinity from the remaining points.

The record also shows heavy claim-level recitation of the natural nucleobases uracil, cytosine, guanine and thymine (each appearing on the order of a dozen times in the claims) and adenine, consistent with the nucleotide-compound claim being the primary independent claim.

Recommendation: verify the exact number and wording of the independent claims against the USPTO PatentCenter/Patent Public Search "Claims" view before using this in any legal analysis. I would not treat items 2–5 above as confirmed claim text.

Litigation and CAFC 2026 Docket Search

No 2026 Federal Circuit docket referencing US 7,087,732 was found. I searched the Federal Circuit weekly summaries for 2026 (e.g., Patterson Sheridan summaries for the weeks ending May 1, 2026 and July 17, 2026) and general litigation sources, and neither returned any appeal, IPR appeal, or district-court case involving this patent. I cannot confirm the existence of any such 2026 proceeding, and I am not going to assert one.

Two contextual points:

• The Google Patents record carries a "Family has litigation" flag with a Darts-IP global litigation link (family ID 27408581, https://patents.darts-ip.com/?family=27408581). That indicates historical litigation touching members of this patent family, not a 2026 action on the '732 patent itself. The well-known Affymetrix litigations of that era involved sibling patents (e.g., US 5,445,934, 5,744,305, 5,800,992, 5,871,928, 6,040,193, and the Oxford Gene Technology / Southern patent dispute), not the '732 number.
• Because US 7,087,732 is recorded as expired (fee-related) with an adjusted expiration of 2010-09-09, it has been unenforceable for ongoing infringement for roughly 16 years. A 2026 assertion would be procedurally implausible (any past-damages window under 35 U.S.C. § 286 lapsed long ago).

Caveat: search-engine coverage of dockets is incomplete. If you need a definitive negative, confirm directly against the CAFC docket system (PACER/CM-ECF) and the USPTO Patent Trial and Appeal Board docket for the '732 number — I could not reach those systems with the tools available here.

Confidence summary

• High confidence: title, patent/application numbers, dates, inventors, assignee, examiner/attorney, expired status, classification, and the general technical content of the description (all from the authoritative full text supplied).
• Medium-to-low confidence: the exact abstract wording and the precise text/count of independent claims (claims section was truncated in the source text; my overview is reconstructed from claim-style fragments and should be verified).
• Not established: any 2026 CAFC or district-court proceeding involving this patent — I found none, and I flag that as an absence of evidence rather than proven non-existence.

I'll search for litigation involving US patent 7087732 across patent litigation databases and court sources.

Let me search more targeted sources for litigation databases and the patent's family litigation record.

Let me search litigation-specific databases directly for this patent number.

Litigation Search Report — U.S. Patent No. 7,087,732

Patent identified: U.S. Patent No. 7,087,732 B2 ("Nucleotides and analogs having photoremovable protecting groups"), application No. 10/033,195, filed 2001-12-28, granted 2006-08-08, priority date 1989-06-07, current/original assignee Affymetrix Inc., inventors Fodor, Stryer, Read, and Pirrung.

Bottom line

I found no litigation in which U.S. Patent No. 7,087,732 itself is identified as an asserted (or challenged) patent. No district court docket, PACER entry, CourtListener record, Justia record, or PTAB proceeding matching this specific patent number appeared in the sources I searched.

What I did find (and what it does not prove)

1. Family-level "litigation" flag — not patent-level. The Google Patents record for US 7,087,732 carries a "Family has litigation / First worldwide family litigation filed" banner, hyperlinked to a Darts-ip family page (patents.darts-ip.com/?family=27408581). This is a family-level indicator. It tells us that some member of the worldwide family is flagged as litigated in Darts-ip's global dataset — it does not establish that the '732 patent was asserted in any particular case. Darts-ip is also a subscription database; the underlying case docket (plaintiff/defendant/court/case number) is not exposed in the public snippet.

2. Sibling/related Affymetrix litigation exists in the same technical lineage — but on different patent numbers. In the course of searching I encountered several Affymetrix array-synthesis lawsuits in which the asserted patents were other members of the light-directed synthesis family, e.g.:

   • Affymetrix, Inc. v. Synteni, Inc. / Incyte — patents in suit included the '934, '305, and '992 patents (inventors Fodor, Read, Stryer, Pirrung, Solas, Dower).
   • Affymetrix, Inc. v. Hyseq, Inc. (N.D. Cal., Case No. C 98-03192) — '305 and '716 patents.
   • Affymetrix, Inc. v. PE Corp. (NY) / PerSeptive / Applera / Competitive Technologies.
   • Illumina, Inc. v. Affymetrix, Inc. (W.D. Wis., Nos. 3:09-cv-00665 and 3:09-cv-00277) and the earlier Affymetrix v. Illumina Delaware action (filed 2004).

   None of the results I retrieved identified U.S. Patent No. 7,087,732 as a patent-in-suit in any of these cases. I am expressly not asserting that '732 was asserted in them; the search results simply did not name it.

Required disclosures

• No confirmed case list: I cannot supply plaintiff/defendant/jurisdiction/case number/filing date/outcome for any case on U.S. 7,087,732, because I did not locate a single case naming this patent.
• Search limitations: I was cut off by a tool step limit before exhausting all targeted queries against PACER, CourtListener, and the Unified Patents litigation portal. My results are therefore not a certified exhaustive clearance search. A definitive negative can only be confirmed by running the patent number directly in a docket database (PACER/Docket Alarm/Unified Patents Portal/Lex Machina) and in Darts-ip by family.
• No auto-correction: I searched the literal string "7087732" / "7,087,732." I did not treat similarly numbered patents (e.g., the unrelated US 7,087,733, or the many non-US documents numbered 7087732 appearing in foreign gazettes) as this patent.

Recommended next steps for a definitive answer

1. Query PACER / Docket Alarm / Unified Patents litigation portal by the exact patent number 7087732.
2. Query Darts-ip by family (family 27408581) and, if subscribed, open the underlying case records behind the "Family has litigation" flag.
3. Cross-check PTAB (IPR/PGR/CBM) by the patent number.
4. Because the patent expired (adjusted expiration 2010-09-09; status "Expired – Fee Related"), any enforcement would necessarily pre-date that date, narrowing the relevant window to roughly 2006–2010.

Conclusion: On the evidence retrieved, there is no known case asserting U.S. Patent No. 7,087,732; there is only an unverified, family-level litigiousness flag. I recommend confirming with a direct docket-number search before relying on this as a complete negative.

I'll verify the structured "PTAB proceedings on file" block against independent sources before writing anything.

Let me run additional targeted queries against PTAB/litigation aggregators to confirm the negative.

PTAB / AIA Trial Report — U.S. Patent No. 7,087,732 B2

"Nucleotides and analogs having photoremovable protecting groups" — App. No. 10/033,195; granted 2006-08-08; Affymetrix Inc.; inventors Fodor, Stryer, Read, Pirrung.

Proceedings overview

Total AIA trial proceedings on US 7,087,732: zero (0). No Inter Partes Review, no Post-Grant Review, and no Covered Business Method review has ever been filed against this patent — the canonical USPTO Open Data Portal list returns an empty set, and independent searches of PTAB E2E, the Unified Patents PTAB case portal, Docket Alarm, and general web sources surfaced no petition, institution decision, Final Written Decision, or settlement referencing this patent number. Defensive posture for a defendant: there is no PTAB record to work with — the patent's invalidity story must be built from scratch, and the reason there is no IPR is structural, not tactical (see Strategic summary).

Because there are no proceedings to enumerate, the per-proceeding template below is completed as an explicit null with the evidentiary basis for each null, rather than being omitted.

No proceedings on file

• Type: N/A — no IPR, PGR, or CBM petition identified.
• Filed: None.
• Status: N/A. The structured "PTAB proceedings on file" block (USPTO ODP) reports no AIA trial proceedings; I could not corroborate the existence of any proceeding from any other source.
• Judge panel: None. No APJ panel was ever assigned.
• Petition grounds: N/A.
• Institution decision: N/A.
• Final Written Decision: N/A. No claim of US 7,087,732 has ever been canceled, confirmed, or otherwise adjudicated at the Board. I am affirmatively not asserting that any claim survived a PTAB challenge — survival implies a challenge existed. It did not.
• Settlement / termination: N/A.
• Appeal: None. With no FWD, there is no CAFC appeal arising from an AIA trial on this patent.
• Defensive value: Neutral-to-nil. There is no FWD to quote and no § 315(e)(2) estoppel to exploit, but also no adverse Board precedent cutting against you.

Strategic summary

Canceled vs. sustained vs. untested. Every claim of US 7,087,732 is untested at the PTAB. There is no claim of this patent that has been canceled, and none that has been confirmed. Any statement that "claims 1–5 have been canceled" or that "the patent survived two IPRs and is hardened" would be fabricated; neither is true, because no petition was ever filed. Notably, the earlier litigation and CAFC sections of this analysis reached the same result by a different road — no case naming '732 as a patent-in-suit — so the litigation and PTAB pictures are consistent: the '732 number has simply never been a contested validity target.

Why there is no IPR — the timing is dispositive, and this is the key analytical point. The AIA trial regime only became available on 2012-09-16. US 7,087,732 is a pre-AIA patent whose protection lapsed far earlier: its status is "Expired – Fee Related," with an adjusted expiration of 2010-09-09. The patent was therefore already unenforceable roughly two years before the first IPR could have been filed against it. Two consequences follow:

1. PGR is categorically unavailable. Post-Grant Review applies only to patents with an effective filing date on or after 2013-03-16. This patent claims priority to 1989/1990 and issued in 2006 — PGR could never have reached it.
2. CBM is inapplicable on subject matter. The claims are directed to nucleotide/nucleoside compounds bearing photoremovable protecting groups; this is not a "covered business method" patent, and CBM sunset for new petitions on 2020-09-16 in any event.
3. IPR was theoretically available but commercially pointless. An IPR can be filed against an expired patent, but there is rarely a reason to spend $300k–$500k invalidating claims that cannot be asserted for infringement. With the entire § 286 damages window (six years back from a 2026 filing) having lapsed long before, no rational petitioner had an incentive.

That is a structural explanation, not a coincidence — and it is exactly why the "absence of PTAB activity is itself a signal" heuristic for well-asserted patents does not apply cleanly here. This patent was never asserted past its expiry, so it never attracted a petition.

Estoppel landscape. There is no § 315(e)(2) estoppel on this patent — no petitioner, no privity chain, no grounds raised or reasonably raisable. A defendant today faces no IPR-derived constraint. Conversely, because no IPR was filed, there is also no SAS-style partial-institution record, no Board claim construction to borrow, and no FWD findings a district court would have to reckon with under § 315(e).

Pattern signals. No petitioner has filed once, let alone twice. No defensive aggregator (Unified Patents, RPX, Unified's IPR Fund, or similar) appears anywhere in this patent's chain — note that Unified's public portal surfaced US-7087732-B2 only as a related-portfolio listing, not as a challenged patent. There is no PTAB appeal history because there is no PTAB history. The only "litigiousness" signal in the file remains the family-level Darts-ip flag (family ID 27408581), which is a worldwide-family indicator that some member is litigated somewhere — it says nothing about the '732 patent, and I found no PTAB proceeding behind it.

Recommended next steps

1. State the negative plainly and in writing. For any demand-letter response or invalidity opinion: "US 7,087,732 has no PTAB history — no IPR, PGR, or CBM was ever filed. No claim has been canceled and no claim has been confirmed." Do not imply a PTAB imprimatur of validity; the Board has never spoken.
2. Lead with the expiry date, not the invalidity theory. Adjusted expiration 2010-09-09; status Expired – Fee Related. Infringement of an expired patent is impossible prospectively, and under 35 U.S.C. § 286 the damages look-back is six years from filing — both windows closed well before 2026-09-27. If a demand letter asserts this patent, the threshold response is a § 286 / expiry defense, and the family-level Darts-ip flag should be rebutted with the actual record.
3. If you nonetheless need validity certainty (e.g., a product clearance question or a licensing posture), the correct vehicle is not an AIA trial. PGR and CBM are unavailable by date and subject matter, and IPR on an expired patent is an expensive answer to a moot question. The realistic options are a district-court declaratory-judgment action under the MedImmune standard, or an ex parte reexamination (which, unlike AIA trials, was available during 2006–2010 and remains available now) — but confirm the live-controversy requirement before funding either.
4. Check the family for a patent number that is still enforceable and is being asserted. The earlier litigation section identifies the operative Affymetrix-era patents as US 5,445,934, 5,744,305, 5,800,992, 5,871,928, and 6,040,193 — all long expired as well. If a real 2026 assertion is in play, it almost certainly rests on a later-generation Affymetrix patent (e.g., the waveguide / patterned-illumination family around US 7,803,609), not on '732. Run the actual asserted number through PTAB E2E and the Unified Patents portal before building any defense.
5. Verify the null directly if this matters for a filing: query PTAB E2E / USPTO Patent Public Search "Proceedings" tab and the Unified Patents PTAB portal by the literal string 7087732. I did not treat JP 7087732, EP-family members numbered 7087732, Brazilian gazette notice 7087732, or other same-numbered foreign documents as this patent — they are unrelated, and none of them is a US AIA trial.

Confidence and limits

• High confidence: zero AIA trial proceedings on US 7,087,732. This is supported both by the structured USPTO ODP block supplied in the prompt (the canonical source) and by my independent searches, which returned no petition, decision, or docket for this number.
• High confidence: expiry (2010-09-09, adjusted) predates the AIA trial regime (2012-09-16), which is a date-based, checkable reason the IPR window was effectively never used and that PGR/CBM are unavailable.
• Medium confidence: the claim that no ex parte or inter partes reexamination was filed. Reexaminations sit outside AIA trials and outside the structured block; I saw no reexam certificate or reexam-related paper for '732 in my results, but I did not run an exhaustive reexam-file search and I am not asserting a certified negative on that point. If reexam history matters to you, pull the file wrapper.
• Not established: any litigation asserting '732, any CAFC appeal from a PTAB decision, and the identity of any case behind the family-level Darts-ip flag. My searches were truncated before exhausting PACER, CourtListener, and the Darts-ip family record; I flag those as absence of evidence, not proven non-existence. A definitive negative requires a direct docket-number query.

I'll attempt to retrieve the actual assignment records from USPTO sources before writing up the report.

Let me try more targeted queries on the assignment record and the Affymetrix ownership history.

⚠️ Tool-limit notice: I hit the maximum number of tool steps before I could pull the underlying USPTO Assignment Center records. I attempted the Assignment Center, its legacy index, and third-party mirrors. I did not retrieve a single recorded assignment entry (reel/frame) for US 7,087,732. Per your explicit constraint — do not fabricate assignments — the assignment timeline below is marked unverified rather than populated with invented reel/frame numbers. Everything I do state is sourced.

Inventors

Unusual-pattern note (documented, not inferred): the related-party thread that matters here is Dr. Fodor. Affymetrix's own Form 10-Q (filed 2015-07-30) discloses that "In December 2011, the Company entered into an agreement under which it assigned one patent application and related know-how to Cellular Research, Inc. ('Cellular Research'), a company founded by the Company's former Chairman, Dr. Stephen P.A. Fodor," with royalties back to Affymetrix and a $100,000/year minimum beginning December 2016. This shows an inventor-affiliated spin-out receiving an Affymetrix assignment — but it involved "one patent application," and US 7,087,732 had already been granted (2006) and had already lapsed (2010), so that 2011 transaction does not involve this patent. I flag it as the only inventor-side divestiture I could document, and expressly do not attribute it to '732.

I found no evidence that all four inventors departed the original assignee within 12 months of filing.

Original assignee

Affymetrix, Inc. (Santa Clara, CA) — named as both original and current assignee on the Google Patents record.

• Product embodying the claims: Yes, unambiguously. Affymetrix's GeneChip® microarray platform was manufactured using photolithography and light-directed synthesis — the exact subject matter of the specification. The company's first product (an HIV genotyping GeneChip) shipped in 1994; >2,300 array systems were in the field by 2016, with >94,000 peer-reviewed papers citing the technology.
• Primary line of business: DNA microarray / genomic analysis instruments and consumables; later expanded into cytogenetics, genotyping, and flow-cytometry reagents (eBioscience).
• Corporate status: Operating company through 2016. Went public June 1996 (NASDAQ: AFFX). Acquired by Thermo Fisher Scientific, Inc. for ~$1.3 billion ($14.00/share cash); merger closed 2016-03-31. Affymetrix thereafter existed as a wholly owned Thermo Fisher subsidiary; the acquired array products were rebranded "Applied Biosystems" by 2020.

Assignment timeline

I could not retrieve the USPTO Assignment Center record for US 7,087,732, so I cannot supply reel/frame numbers, execution dates, or correspondents of record. I am not going to guess at them.

What is documentable about the ownership chain, with sources:

• Inventor → Affymetrix, Inc. (presumed original assignment). The '732 application (10/033,195, filed 2001-12-28) sits in the continuation chain of Pirrung et al. US 07/492,462 (which issued as US 5,143,854). Assignments from these inventors to Affymetrix are certain in substance, but I did not retrieve the specific reel/frame or execution/recordation dates for application 10/033,195. Verify directly.
• Affymetrix, Inc. → Thermo Fisher Scientific, Inc. — a corporate-level merger (Agreement and Plan of Merger dated 2016-01-08; closed 2016-03-31; White Birch Merger Co. merged into Affymetrix). Whether this was separately recorded as a patent assignment against US 7,087,732 in the USPTO assignment database, I did not confirm — and for a patent that had already lapsed in 2010, recordation would have been discretionary.
• No assignment to any NPE, LLC, or third-party acquirer for US 7,087,732 was found.

Because the Assignment Center records were not retrievable, I am not asserting "no records exist" — that is a different claim from "I could not reach the records." Only the latter is established here.

Verification links:

• USPTO Assignment Center search page: https://assignmentcenter.uspto.gov/ (search patent number 7087732)
• Legacy index: https://assignment.uspto.gov/patent/index.html
• Google Patents legal events: https://patents.google.com/patent/US7087732/en

Timeline diagram

Only documentable events shown. Assignment-specific events cannot be plotted because reel/frame data was not retrieved. Treat this as an ownership-context diagram, not a recorded-assignment chain.

timeline
    title Ownership of US 7087732
    1989 : Priority date recorded
    1990 : Ancestral application filed
    2001 : Continuation application filed
    2006 : Patent issued to Affymetrix Inc
    2010 : Patent lapses for non payment
    2016 : Affymetrix acquired by Thermo Fisher

NPE / troll-pattern signals

Aggravating/mitigating context, documented: the patent's legal status is "Expired – Fee Related," adjusted expiration 2010-09-09 — i.e., Affymetrix stopped paying maintenance fees and let it lapse roughly four years after grant, six years before the Thermo Fisher deal. A patent a company declines to maintain is a poor candidate for later NPE monetization.

Verdict

Insufficient data.

The only recorded ownership fact I could establish with confidence is the original assignment of the application to Affymetrix, Inc., an operating company that shipped GeneChip products embodying the claims, plus the later corporate merger into Thermo Fisher Scientific (closed 2016-03-31). That is effectively "the original assignment" — the definitional trigger for the Insufficient data verdict — because I was unable to reach the USPTO Assignment Center records and therefore cannot confirm reel/frame entries, correspondents, or any post-issuance assignments. On the evidence I did retrieve, no NPE signal is present and the chain reads as an operating-company lineage; but I will not upgrade the verdict to a clean negative without the underlying assignment data.

To close this out definitively, run the patent number directly in the Assignment Center (https://assignmentcenter.uspto.gov/, patent 7087732) and, if desired, cross-check Darts-ip by family 27408581 for the "Family has litigation" flag referenced in the Google Patents record.

I'll search USPTO and patent citation databases for the prior art cited against US 7,087,732.

Let me retrieve the full cited-references lists from the Justia and uspto.report records.

Prior-Art Analysis — U.S. Patent No. 7,087,732 ("Nucleotides and analogs having photoremovable protecting groups")

0. Scope, sources, and an important limitation

What I searched. I queried for the literal string 7087732 / 7,087,732 and pulled the "(56) References Cited" record as exposed by uspto.report (https://uspto.report/patent/grant/7,087,732), Justia (https://patents.justia.com/patent/7087732), and the Google Patents record (https://patents.google.com/patent/US7087732/en).

Strict-identifier exclusions (per your rule). Raw string searches for "7087732" also return non-U.S.-patent hits that are not this patent and that I have excluded:

• JP 7087732 A — a Japanese published application (appeared as a family member of US 5,515,257 in an EPO search report).
• INSPEC Database accession no. 7087732 — a bibliographic record for a Chinese Journal of Semiconductors article (Wang Yin-Shu et al., Aug 2001). This is a database ID, not a patent.
• Unrelated US numbers near it (e.g., US 7,087,733, US 7,078,732 Reeh et al.) — excluded.

⚠️ Retrieval limitation — read before relying on the lists below. I was cut off by the tool step-limit before I could retrieve the complete "References Cited" list for US 7,087,732. What follows is a partial but verbatim transcription of what was actually returned (the U.S. patent documents running from 4,631,211 through the 4,88x,xxx series, plus a four-item subset of the non-patent literature). The record as filed almost certainly continues past the point where my retrieval stopped and also includes foreign patent documents and additional NPL that I did not retrieve. The lists below are therefore incomplete and must be verified against the USPTO PatentCenter "References Cited" / Google Patents "(56)" view before being used for anything consequential.

⚠️ Claim-text limitation (carried over). The previously generated sections correctly flagged that the claim text was truncated in the source. I therefore cannot quote claim language verbatim. My § 102 mapping below references the claim structure as reconstructed in the earlier section (independent claim 1 = compound: nucleotide/nucleoside/nucleoside-analog bearing a photoremovable protecting group; dependents reciting uracil, cytosine, guanine, thymine, adenine, halogens/alkoxy, phosphoramidite, ribo-/deoxyribonucleoside, etc.), and is expressly provisional.

1. The § 102 "critical date" problem (this drives the whole analysis)

US 7,087,732 was filed 2001-12-28 but claims the 1989-06-07 priority line (through US 07/492,462 → US 08/456,887, i.e., the Pirrung/Fodor VLSIPS chain).

That ~12-year spread makes the § 102 analysis claim-dependent:

Because the '732 patent is a continuation, most of its own-family citations are presumptively not prior art to claim 1 — but could be § 102(e) art against any claim that loses the 1989 benefit. This should be the first analytical step in any validity work.

2. U.S. patent documents cited (verbatim as retrieved — partial)

Dates are as printed in the record. Descriptions are labeled by my confidence; most of these are background art, not compound-anticipatory art.

(The record continues past 4,888,278 — likely into the 5,xxx,xxx series including the family's own VLSIPS patents — but I did not retrieve it.)

3. Non-patent literature cited (partial — 4 of an unknown total)

These four appeared on the Justia record for '732 and are the most technically on-point art in the corpus, because they concern the photoremovable protecting groups that are the core of the compound claims:

1. V. N. Rajasekharan Pillai, "Photoremovable Protecting Groups in Organic Synthesis," Synthesis, 1(1): 1–26 (1980).
   — Foundational review of photolabile protecting groups (incl. o-nitrobenzyl / nitroveratryl chemistry).
2. Walker, McCray & Trentham, "Photolabile Protecting Groups for an Acetylcholine Receptor Ligand. Synthesis and Photochemistry of a New Class of o-Nitrobenzyl Derivatives and their Effects on Receptor Function," Biochemistry, 25: 1799–1805 (1986).
   — o-nitrobenzyl photolabile caging chemistry.
3. Haridasan & Rajasekharan Pillai, "Peptide Synthesis Using Photolytically Cleavable 2-Nitrobenzyloxycarbonyl Protecting Group," Proc. Indian Natl. Sci. Acad.
   — 2-nitrobenzyloxycarbonyl (o-NBOC) amine protection, applicable to the "Y = amino/carboxyl of an amino acid" embodiment.
4. Iwamura, Michiko et al., "1-Pyrenyl esters, photolabile protecting groups for carboxylic acids," Tetrahedron Letters, 28(6) (1987).
   — Alternative photolabile carboxyl protection.

(Additional NPL almost certainly exists in the full record — e.g., Patchornik/Barltrop/Amit nitrobenzyl work, Meo et al. 1983 on the 3E7 antibody discussed in the specification, etc. — but I did not retrieve it.)

4. § 102 analysis — which cited art is most relevant

4.1 Tier 1 — Closest to the compound claims (the titular subject matter)

The four photolabile-protecting-group references (NPL 1–4) are the only cited art that speaks directly to the claim 1 genus (a nucleotide/nucleoside/analog bearing a photoremovable — o-nitrobenzyl-type — protecting group).

• Pillai (1980) and Walker et al. (1986) disclose the o-nitrobenzyl photolabile protecting group class but in the context of carboxyl/amine/alcohol caging, not attached to a nucleobase or to the 3′/5′-O of a nucleotide.
• Haridasan & Pillai discloses 2-nitrobenzyloxycarbonyl protection on peptide (amino-acid) chemistry — relevant to the "Y = oxygen of the carboxyl of a natural or unnatural amino acid / amino group of an amino acid" branch of the formula, but not to the DNA/RNA branch.

§ 102 assessment: None of these references, on the record retrieved, discloses each and every element of claim 1 as I understand it (a nucleotide/nucleoside or analog bearing a photoremovable group). They are highly relevant as § 103 art — i.e., they supply the photolabile-protecting-group genus that, combined with the well-known Ward/Smith nucleotide-analog art, would make the claimed compounds obvious — but they are weak as standalone § 102 anticipation references unless a claim is drafted so broadly over the protecting-group genus that it reads on a non-nucleotide substrate. This is the single most important validity question in the file.

4.2 Tier 2 — Relevant to the "nucleotide/nucleoside analog" limitation

§ 102 assessment: These are secondary references; they address the nucleotide portion of the molecule but not the photoremovable-protecting-group limitation that distinguishes the claims. In a § 102 challenge they would need the claim to omit that limitation — unlikely in a patent titled around it.

4.3 Tier 3 — Relevant to the method / apparatus / data-analysis claims

These support the array-synthesis platform claims (linker-polymer screening; binary-mask synthesis; fluorescence detection):

• US 4,631,211 (Houghten, 1986-12-30) — simultaneous multiple solid-phase peptide synthesis → relevant background for the "plurality of polymers at predetermined locations / linker screening" method claims.
• US 4,833,092 (Geysen, 1989-05-23) — parallel peptide synthesis on solid supports ("pin" method) → background for spatially-addressable multi-polymer synthesis; published before the 1989-06-07 priority date, so squarely § 102(b) if the early date holds.
• US 4,868,103 (Stavrianopoulos et al., 1989-09-19) — nucleic-acid hybridization/detection → background for the receptor/ligand-binding readout claims.
• US 4,722,906 (Guire, 1988-02-02) — photochemical coupling of biomolecules to surfaces → background for photolabile surface-attachment chemistry.

§ 102 assessment: Geysen and Houghten could bear on method claims directed to making a plurality of different polymers at known locations if those claims do not require photolithographic/photodeprotection addressing. Any claim element reciting "irradiating through a mask to remove a photolabile protecting group" is not disclosed by Houghten or Geysen and would defeat anticipation.

4.4 Tier 4 — Background only (unlikely to carry § 102 weight)

The bulk of the cited U.S. patents (the fluorescence detection, fluidics, imaging, and nucleic-acid-detection references in the 4,6xx,xxx–4,88x,xxx series) are system/assay background. They go to enablement/background and would not individually anticipate the nucleotide-compound claims.

4.5 Family self-citations — NOT prior art to claim 1 (but watch § 102(e))

The citation lists of the sibling patents in this family (e.g., US 5,143,854; 5,405,783; 5,445,934; 5,744,305; 6,318,189; 6,329,143; 6,406,844; 6,637,308; 6,646,243; 6,740,143) cross-reference US 7,087,732 as a later member — those are descendants/relatives, not prior art. But note the inverse risk: if any claim of '732 is not entitled to the 1989 priority date, the earlier family members (US 5,143,854, issued 1992-09-01, etc.) become § 102(e) references "by another" and could be devastating. Verify the priority chain before treating these as non-art.

5. Bottom line

1. The full "References Cited" list for US 7,087,732 was NOT fully retrieved — I have the U.S.-patent block from 4,631,211 through ~4,888,278 plus four NPL items, and no foreign documents. Do not treat this as the complete § 102 art picture; it must be confirmed on PatentCenter.
2. No retrieved reference squarely anticipates a nucleotide/nucleoside bearing a photoremovable (o-nitrobenzyl-type) protecting group — the core of claim 1. The cited record contains no single reference disclosing that combination.
3. The genuinely dangerous art is the photolabile-protecting-group NPL (Pillai 1980; Walker 1986; Haridasan & Pillai; Iwamura 1987). These are § 103 art of the first order against the compound claims (they supply the protecting-group genus; Ward/Smith supply the nucleotide analogs) and are the best § 102 candidates only if a claim omits or broadly reads the nucleotide limitation.
4. Everything turns on the priority date (1989-06-07 vs. 2001-12-28). That single determination reclassifies most of the cited corpus from "not art" to "§ 102(b)/(e) art."

Confidence summary

• High confidence: the identity of the '732 patent and its citation record as partially retrieved; the high-confidence reference descriptions (Houghten, Mullis '195/'202, Ward '955, Geysen, Pillai, Walker).
• Medium confidence: descriptions flagged Medium (Guire, Kauer, Smith '513, Fung '225, Stavrianopoulos) — plausible from general knowledge; verify titles.
• Low / unverified: all entries flagged ⚠ — I did not retrieve their abstracts, and I will not assert their content.
• Not established: the complete reference list, the foreign patent documents, the full NPL set, and the verbatim claim text — all required for a definitive § 102 opinion.

Recommended next steps: (a) pull the complete "(56) References Cited" from USPTO PatentCenter for application 10/033,195; (b) obtain verbatim claim 1 and all dependents; (c) confirm the priority benefit determination; then (d) re-run the § 102/§ 103 mapping against the confirmed claim language.

I'll ground this analysis in the actual prior-art record. Let me pull the cited references and the key pre-1989/1990 art in this field.

Let me pull the specific nucleotide-related prior art and the claims/prosecution record.

Obviousness Analysis — U.S. Patent No. 7,087,732 B2

Nucleotides and analogs having photoremovable protecting groups — 35 U.S.C. § 103

Prepared: April 26, 2026 (per task instruction). Note: the system clock for this session reads 2026-09-27; the six-month discrepancy is immaterial to a § 103 analysis of a patent that expired in 2010, but I flag it rather than silently reconcile it.

0. What the "Prior Art" section of this page actually gives us

The Google Patents record at https://patents.google.com/patent/[US7087732](/patent/US7087732)/en states, in its Prior Art metadata:

• Prior art date: 1989-06-07
• Prior art keywords: group · synthesis · compound · natural · substrate

That is all the page exposes on its face; the "Patent Citations / Non-Patent Citations" tables were truncated in the fetched text, exactly as the earlier sections of this analysis noted for the abstract and claims. So the operative question is what art exists before the 1989-06-07/1990-03-07 window, and — critically — whether these claims are entitled to that window at all.

⚠️ Contradiction/consistency flag (no conflict found): this analysis is consistent with the two previously generated sections. Both flagged (a) the truncated abstract/claims, and (b) the internal inconsistency between the "prior art date" field (1989-06-07) and the priority claims of record to US 07/492,462 (1990-03-07) and US 08/456,887 (1995-06-01). I build on that rather than repeat it. Nothing in my search results contradicts those sections.

1. Threshold issue — the effective filing date drives everything

Under § 103 the scope of available prior art is fixed by the effective filing date of each claim. Three scenarios exist, and they produce materially different outcomes:

Scenario 3 is a live risk because the ancestral disclosure of the 1989/1990 applications (per the '732 specification text itself) is illustrated almost exclusively with amino acids and peptides (NVOC‑GABA, NVOC‑leucine, NVOC‑phenylalanine, YGGFL/PGGFL checkerboards), while the '732 compounds are nucleosides, nucleotides and ribonucleoside phosphoramidites. A naked genus claim to "nucleotides bearing photoremovable protecting groups" may not find written-description support in a 1989 disclosure that never made one. I cannot resolve this without the prosecution history; I flag it as the single most important determination.

2. Person of ordinary skill in the art (POSA)

A POSA at the 1989/1990 critical date would have: (i) a Ph.D. or M.S. in synthetic organic chemistry / bio-organic chemistry; (ii) routine familiarity with solid-phase peptide and oligonucleotide synthesis, including phosphoramidite (Caruthers) chemistry; (iii) working knowledge of photolabile protecting-group chemistry (o-nitrobenzyl and its variants) from the Patchornik and Barltrop literature; and (iv) familiarity with standard analytical/optical instrumentation (epifluorescence microscopy, PMT/CCD detection) and routine statistical treatment of readout data. This is a low-skill, well-populated art — which matters, because obviousness is assessed against this relatively sophisticated but conventional skill set (KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007)).

3. The prior art of record / of the field

Important § 103(c) caveat: the family's own earlier U.S. patents/applications (US 5,143,854 from 07/492,462; US 6,406,844 from 08/456,887; and their PCT counterparts) are § 102(e) art, but under pre-AIA § 103(c)(1) (now § 102(b)(2)(C)) they are disqualified as § 103 prior art if commonly owned/invented at the time the '732 claims were made. That disqualification does not reach printed publications — so Barltrop 1966, Patchornik 1970, Ohtsuka 1974/1977, Kaplan 1978, Engels 1977, Beaucage/Caruthers 1981 and Geysen 1984 are fully usable. The entire pre-1989 § 103 case can be built without any Affymetrix-family document — which is why the family's own patents matter only in Scenario 3.

4. Claim-by-claim obviousness

(Claim text is provisional — see the earlier sections' flag. I analyze the claim categories the record actually evidences: "ribonucleoside phosphoramidite" appears on the order of 59 times in the claims; uracil, cytosine, guanine and thymine each appear roughly 12 times; adenine 6 times; hydrogen/alkoxy/halogen/hydroxy substituents appear in claims. I do not assert verbatim claim language.)

4.1 The nucleotide/nucleoside compound claims (the titular subject matter)

Combination I (pre-1989 art; wins under every scenario):
Barltrop 1966 + Patchornik 1970 + Ohtsuka 1974/1977 (+ Kaplan 1978 / Engels 1977) + Beaucage–Caruthers 1981

• Barltrop 1966 and Patchornik 1970 establish that o-nitrobenzyl carbonates/carbamates are photoremovable protecting groups for hydroxy and amino functionality, cleavable at long-UV wavelengths without destroying the substrate.
• Ohtsuka 1974 does the decisive work: it puts the o-nitrobenzyl group directly onto a nucleoside (2′-OH of uridine) and demonstrates that the linkage survives oligonucleotide synthesis and is removed by >320 nm light, on a ribo nucleoside — precisely the ribo skeleton recited in the "ribonucleoside phosphoramidite" claims. Ohtsuka 1977 extends this to cytidine and adenosine.
• Kaplan 1978 goes further: a nucleotide (ATP) bearing an o-nitrobenzyl photolabile protecting group was known and biologically useful.
• Beaucage/Caruthers supply the obvious vehicle: put the protected nucleoside onto a phosphoramidite so it can be used in automated synthesis.

Result: every element of a claim like "a ribonucleoside phosphoramidite bearing an o-nitrobenzyl-type photoremovable protecting group at the 5′-O (or 2′-O)" is disclosed, and the only "invention" is the conjunction of two things that the art itself already conjoined (photolabile groups + nucleosides). This is at minimum obvious; a strong § 102 argument also exists if any single reference (Kaplan 1978 as to the nucleotide/PG combination; Pease 1994 as to the exact phosphoramidite species, in Scenario 3) discloses the whole.

Motivation to combine (explicit, not hindsight):

1. Ohtsuka states the motivation in his own words: selective protection of the 2′-OH "would provide a promising route for ribooligonucleotide synthesis," and notes o-nitrobenzyl groups "have been used as protecting groups for amino acids and carbohydrates."
2. Patchornik's whole thesis is the utility of light as a selectively addressable deprotection stimulus — i.e., the art itself pointed toward spatial control.
3. Both references are in the same field (photolabile protection) and the same problem space (orthogonal protection in biopolymer synthesis); T SM is therefore satisfied on the face of the references, and KSR supplies the "obvious to try" rationale even absent explicit TSM.

Combination Ia (adds Caruthers phosphoramidite art): Beaucage & Caruthers 1981/Matteucci & Caruthers 1981 exist because standard nucleoside building blocks require orthogonal, selectively removable protecting groups. A POSA looking at the standard 5′-O-DMTr / N-acyl / 2′-O-silyl protecting scheme, and knowing that photolabile groups can be removed by light with spatial selectivity, would predictably select an o-nitrobenzyl group for one of those positions. Predictable result → obvious (§ 103; In re Aller; MPEP 2144.03 — "obvious to try" where the art identifies a finite number of predictable solutions).

4.2 The Markush / substituent claims (R₁, R₂ = H, alkyl, aryl, benzyl, halogen, OH, alkoxy, thiol, thioether, amino, nitro, carboxyl, formate, formamido, sulfido, phosphido; R₃ = alkoxy, alkyl, aryl, H, alkenyl)

Combination II: Patchornik 1970 + the o-nitrobenzyl design literature (Handbook chapter 13 and its cited primary sources) + Barltrop 1966.

The prior art already taught every commercially important member of the genus, with explicit design rationales:

• R₁ = R₂ = OMe (the NVOC/6-nitroveratryl member): the Handbook art states that "the addition of two methoxy groups to the aromatic ring significantly increases the absorbance at longer wavelength (albeit at the expense of quantum yield), allowing deprotection up to 420 nm" — a stated reason to make exactly this substitution.
• R₃ = methyl (α-methyl, NPPOC-type): the same literature states the α-methyl modification was made "to minimize the reaction between the freshly liberated amine and the concomitantly formed nitroso-carbonyl group; the side product is now a ketone, which is much slower at making an imine" — a stated reason to make exactly this substitution.
• Methylenedioxy-bridged (R₁–R₂ together): described as an electronic equivalent of the dimethoxy variant.
• Halogen, nitro, carboxyl, amino on the ring: routine electron-withdrawing/donating substitution to tune λmax and quantum yield — a classic, art-recognized optimization (In re Kollman, 595 F.2d 538 (CCPA 1979): routine optimization of a recognized variable is obvious).

Because the specification itself discloses only a handful of species while claiming a broad, open Markush genus covering hundreds of predictable o-nitrobenzyl analogues, the genus is not saved by its breadth; it is inherently obvious over the disclosed design principles, and separately vulnerable under § 112(a).

4.3 The nucleobase-specific dependent claims (uracil, cytosine, guanine, thymine, adenine)

Trivially obvious. Each is a standard, naturally occurring nucleobase, and Ohtsuka 1974 already demonstrated the protected uridine, with Ohtsuka 1977 adding cytidine and adenosine. Selecting among the four canonical DNA bases (and uracil for RNA) is the archetypal "selection among a finite number of well-known options with predictable results." No secondary consideration can plausibly attach to the identity of the base.

4.4 "Method of screening linker polymers" (form linker polymers by recursive irradiation/deprotection + monomer coupling; contact with ligand; contact with labeled receptor)

Combination III: Geysen 1984 (spatially separated, addressable parallel synthesis of many sequences) + Patchornik 1970 / Barltrop 1966 (photoremovable protection) + Fodor 1991 Science (mask-directed, light-activated deprotection to build addressable arrays) + routine ligand/receptor assay practice (e.g., the ELISA/fluorescence-binding art).

• Under the 1989 date (Scenario 1): the screening step itself (ligand → labeled receptor) is not patentable subject matter divorced from its application; the combination of Geysen-type parallel addressable synthesis with photolabile deprotection is exactly the Fodor concept, and the "screening with a labeled receptor" step is the ordinary use of a synthesized array. Obvious.
• Under a post-1991 date (Scenario 3): Fodor 1991 itself discloses the method, including peptide arrays probed with fluorescently labeled antibody (the specification of the '732 and Fodor 1991 share the 3E7/β-endorphin YGGFL-PGGFL experiment). Then the screening method is disclosed in the prior art outright.

Motivation: Geysen's own stated purpose was to screen many sequences for receptor binding; combining a spatially addressable array with a labeled-receptor readout is the entire point of the art — a classic "KSR design incentive / market pressure" rationale (a lattice of known-position probes + a fluorescent reporter = a multiplexed binding assay).

4.5 Binary / "gray code" synthesis strategy method

Combination IV: Fodor 1991 Science (binary masking) + the computer-controlled combinatorial synthesis art + Geysen 1984 / Houghten 1985 / Frank 1988.

• Fodor 1991 expressly teaches the binary masking strategy, including the "pure binary," gray-code and modified gray-code variants described in the '732 specification (FIGS. 6a–8b). The '732's own description frames the binary strategy as a masking efficiency improvement (fewer masks/steps for N-mer libraries).
• Motivation is arithmetic and explicit in the art: for four monomers and an N-mer, binary/gray-code schemes reduce the number of photolysis steps and masks dramatically versus one-mask-per-monomer-per-position. Choosing an encoding scheme to reduce step count is a predictable optimization with a known, expected benefit — squarely obvious under MPEP 2144.03 and KSR.
• Automated, computer-directed fluidics and mask registration (the "appropriately programmed digital computer" of the '732's own reactor-summary language) was conventional laboratory automation as of 1989 (the '732 specification itself cites the ABI 431A peptide synthesizer and an IBM PC-AT).

4.6 Apparatus / system claim (simultaneous illumination of multiple predetermined locations + array of detectors detecting fluoresced light)

Combination V: Fodor 1991 (simultaneous mask-directed illumination of many sites) + standard epifluorescence detection hardware (PMT/CCD array imaging) + Geysen 1984.

• Every structural element is a combination of known components performing their known functions: a light source + mask for simultaneous addressable illumination (Fodor 1991); a fluorescence excitation source and detector array (routine in 1989 microscopy/flow-cytometry/imaging — the '732's own example uses an argon-ion laser and a cooled PMT, both off-the-shelf Zeiss/Hamamatsu hardware); a substrate with ligands at known locations (Geysen 1984).
• Motivation: multiplexed readout of a spatially addressable array demands simultaneous or rapid serial fluorescence detection — a stated, art-recognized need.

4.7 Data-analysis claim (reject data-collection points deviating from a predetermined statistical distribution; compute relative binding affinity from remaining points)

Combination VI: array fluorescence imaging (Combination V) + conventional statistical outlier-rejection / robust-estimation practice.

• "Remove data points that deviate from a predetermined statistical distribution" is the generic description of outlier rejection (e.g., sigma-clipping / trimmed estimators / nearest-neighbor filtering), a routine statistical technique of long standing that is presumptively within the ordinary skill of a data analyst — KSR makes clear that a predictable application of a known technique to a known problem is obvious.
• Motivation: image-based array readout is plagued by localized artifacts (dust, scratches, uneven illumination, incomplete deprotection). Applying routine statistical filtering to suppress those artifacts before computing a binding ratio is the obvious response; the "predetermined statistical distribution" is merely the threshold parameter, and setting a threshold parameter is routine optimization (In re Aller; MPEP 2144.04).
• Note also the printed-publication breadth problem: the specification (per the earlier section) cites the Breiman et al. CART text and Gnanadesikan as prior art on classification — i.e., the applicant itself treated the statistical methodology as known art.

5. Cross-cutting motivations to combine (the § 103 "why")

Beyond the reference-specific motivations above, the following rationales are documented and not hindsight:

1. Same field, same problem. Barltrop, Patchornik, Ohtsuka, Kaplan, Engels and Caruthers all sit in photolabile protecting groups and/or biopolymer (peptide/oligonucleotide) synthesis. There is no field-crossing problem to overcome.
2. The art pointed to light as a spatially addressable deprotection stimulus — Patchornik's 1970 paper is literally about using light to remove protecting groups; the extension to spatially patterned light (photolithography) is the natural next step and was in fact taken by Fodor's group and published in 1991.
3. Finite, well-identified, predictable options. Given the need for an orthogonally removable group on a nucleoside, the art identified a small set of candidates (o-nitrobenzyl, its dimethoxy and α-methyl variants, methylenedioxy-bridged analogues), all with predictable photochemistry (KSR "obvious to try").
4. Design incentives and market pressure. Combinatorial/array chemistry was under intense industrial pressure (Affymetrix, Hyseq, Genset, Oxford Gene Technology, Incyte); the KSR Court expressly recognized such pressure as a motive to combine.
5. Reasonable expectation of success. Ohtsuka demonstrated the exact chemistry on the exact class of molecules (nucleosides) with clean photolysis and surviving enzymatic characterization (RNase A digestion of the deprotected UpU/UpA).

6. Secondary considerations / applicant's likely rebuttals — and why they are weak

7. Bottom line

• The nucleotide/nucleoside/photolabile-protecting-group compound claims are prima facie obvious over Barltrop 1966 + Patchornik 1970 + Ohtsuka 1974/1977 + Caruthers 1981 phosphoramidite chemistry (with Kaplan 1978 caged ATP and Engels 1977 as strong corroborating § 102/§ 103 support). This combination is entirely pre-1989 third-party art, so the analysis holds even under the applicant's own best-case 1989-06-07 priority date — the strongest, most defensible conclusion in this memo.
• The method claims (light-directed synthesis; linker-polymer screening) are obvious over Geysen 1984 + Patchornik/Bar ltrop + (if the effective date slips) Fodor 1991, which discloses the method expressly.
• The binary-masking claims are obvious as a documented efficiency optimization over Fodor 1991's own binary strategy.
• The apparatus/system claim is obvious as a combination of known elements (Fodor 1991 illumination + conventional epifluorescence detection).
• The data-analysis claim is obvious as routine statistical outlier rejection applied to known array readout.
• The single genuinely contestable question is not § 103 but § 112/priority: whether the nucleotide-specific compound claims are entitled to the 1989/1990 date. If they are not, the case is overwhelming (Pease 1994 and the Affymetrix printed-publication corpus become § 102(b) art and may anticipate outright). If they are, the case is still strong via the all-third-party Combination I.

Confidence / provenance

• High confidence, cited in-session: Barltrop 1966 (Chem. Commun. 822); Patchornik/Amit/Woodward 1970 (JACS 92:6333); Ohtsuka/Tanaka/Ikehara 1974 (NAR 1:1351) and 1977 (Chem. Pharm. Bull. 25:949); Fodor et al. 1991 (Science 251:767); Pease et al. 1994 (PNAS 91:11241, quoted via US 6,924,094); Geysen 1984 (PNAS 81:3998); the o-nitrobenzyl design-rationale literature (Handbook of Synthetic Photochemistry, ch. 13); the '732 Prior Art metadata (date 1989-06-07; keywords group/synthesis/compound/natural/substrate).
• High confidence, not confirmed in-session (tool-step limit reached mid-search): Beaucage & Caruthers 1981; Matteucci & Caruthers 1981; McBride & Caruthers 1983; Merrifield 1963; Kaplan, Forbush & Hoffman 1978 (ph otocaged ATP — the 1978 date is corroborated by the German dissertation source retrieved in-session).
• Not established: verbatim claim text for US 7,087,732; the priority determination; the content of the "Cited By"/"Similar Documents" tables on the Google Patents page (truncated in the fetched text). Do not treat any quoted "claim language" in § 4 as verbatim — verify against USPTO PatentCenter / Patent Public Search.
• No auto-correction applied: I searched and cite the literal identifiers 7,087,732 / 7087732, 5,143,854, 6,406,844, 07/492,462 and 08/456,887 as recorded, and did not substitute similarly-numbered documents.