ctDNA Assay Reference Panel

• The statement: claim 1unconfirmed — 21 CFR 809.10(c) exempts a properly labeled RUO product from FDA premarket notification/approval. claim 2unconfirmed — FDA's 2013 RUO/IUO distribution guidance sets out when marketing conduct defeats RUO labeling. claim 3unconfirmed — marketing this material to CLIA laboratories to validate clinical MRD tests is the conduct that guidance treats as defeating RUO status.
• What the check found: connectors regulation 809.10 --title 21 --jurisdiction US returned verified for the labeling text only — 809.10(c) exempts a shipment from paragraphs (a)/(b) (IVD (in vitro diagnostic) labeling) and a part 861 standard, and says nothing about premarket notification or approval; claim 1unconfirmed mis-cites its own source on the load-bearing half. No connector in this repo reaches FDA guidance documents (openFDA exposes clearance/ classification records only), so claim 2unconfirmed has never been retrieved by any method, and claim 3unconfirmed — an interpretation of a document that was never read — has nothing to be checked against.
• Who depended on it: A1 In vitro diagnostic products properly…, A2 FDA guidance "Distribution of In Vitro…, A3 Enforcement risk claim — Regulatory (C1, C2), and Commercial (REQ-0029 Does the file's own stated principal…, which states every dollar in its business case depends on the answer).
• What it costs: if the labeling exemption does not reach premarket review, or if this candidate's go-to-market is found to defeat RUO status, the launch route as framed sits inside FDA premarket review or active enforcement exposure — neither of which claim 20unconfirmed's capital estimate or 9-15 month calendar budgets for.
• Outcome: commission
• Because: REQ-0024 Does FDA guidance "Distribution of In Vitro… (regulatory → research) is already open and asks exactly the guidance question that would move claims 2 and 3 — a public fda.gov PDF, not a licensed source, named by Regulatory as "the single cheapest move on this file and the most decisive."

• The statement: claim 16unconfirmed — material with a stated uncertainty budget at 0.001%-0.01% VAF is Poisson-limited at typical 10-30 ng cfDNA inputs, so the lowest panel levels may be irreducibly imprecise rather than merely hard to make.
• What the check found: no connector reaches a bench/physics question. The domain dossier's own targeted literature search ("digital PCR uncertainty measurement ultra-low allele fraction circulating tumor DNA") returned zero PubMed records — confirmed not found, not merely not searched — and Technology independently re-ran two differently-worded queries and got the same clean null.
• Who depended on it: A7 Technical/mechanism risk, and by extension A6 A dPCR replicate series of at least 20… (claim 62unconfirmed, Clinical's sharpened restatement of the same question) — Technology (C1), Operations (C4), and indirectly Clinical (C1) and Commercial (REQ-0030 Is orthogonal dPCR quantification of…).
• What it costs: the entire commercial differentiator over three incumbent vendors who stop at 0.1% VAF (claims 45, 46) — see A6 A dPCR replicate series of at least 20… for the full four-chair consequence chain.
• Outcome: commission
• Because: REQ-0030 Is orthogonal dPCR quantification of… (commercial → technology) already asks this question. No connector or literature search can settle it further; every chair that names a settling artifact names the same one — an internal dPCR replicate series at stated input mass — which is bench work, not a further research lookup.

• The statement: claim 31unconfirmed — commutability against real clinical patient samples can be demonstrated across the major tumor-informed MRD assay chemistries, rather than asserted from matrix choice.
• What the check found: no connector applies. Clinical's own reading is that this is "the file's own honest statement" that the claim is asserted, not demonstrated; the domain dossier records that no independent (non-vendor-self-reported) commutability study was located for any of the three named reference-material vendors either.
• Who depended on it: A8 Commutability against clinical patient… — Clinical (C2), Technology (C2), and Operations (C2), triple- bound.
• What it costs: whether native fragmentomic-profile preservation is a measurable performance advantage over the incumbents' synthetic-matrix products, or an asserted one — the commercial differentiator's second, independent leg alongside A6 A dPCR replicate series of at least 20…/A7 Technical/mechanism risk's VAF-floor question.
• Outcome: commission
• Because: this is a cross-platform bench/field study against real patient specimens, not a database lookup. Technology and Operations both name the settling artifact directly (a commutability study on the same lot the low-VAF characterization is performed against) as buildable next-step work outside this pipeline's connector surface — no ledger request was raised against it for the same reason none was raised for the equivalent bench items in idea-002's own loop record (IR-8 claim 13 came back unconfirmed there).

• The statement: claim 30unconfirmed — pooled human donor plasma can be sourced at a scale and consistency that preserves the native cfDNA fragmentomic profile lot over lot. claim 32unconfirmed — the lowest-VAF panel levels are stable through freeze-thaw, shipping and the stated shelf life within the published uncertainty budget.
• What the check found: neither is checkable by any connector — this is unbuilt-hardware and unrun-procurement territory. Operations's own cold-run analog check (81 recalls, 2011-2026, under the nearest comparable product codes JJX/JJY) found stability-drift and cold-chain/shipping failures among the four repeating root causes across many independent manufacturers of the adjacent product category — not evidence about this candidate, but the closest available manufacturing-history analog for the same failure shape.
• Who depended on it: A10 Pooled human donor plasma can be sourced at… (claim 30unconfirmed) — Technology (C3), Operations (C1). A11 Panel levels at the lowest allele fractions… (claim 32unconfirmed) — Technology (C4), Operations (C3).
• What it costs: A10 Pooled human donor plasma can be sourced at…: the entire native-plasma-vs-synthetic-matrix differentiator has no reliable raw-material input if plasma cannot be sourced consistently. A11 Panel levels at the lowest allele fractions…: an assigned value that does not survive reaching a customer's bench, on the same failure shape recall history already shows recurring in the analog device family.
• Outcome: commission
• Because: both are named, buildable next steps outside this pipeline's connector surface — Technology and Operations each name the settling artifact directly (fragmentomic-profile comparison across independently sourced lots; a stability study spanning the claimed shelf life and shipping conditions). Same shape as A8 Commutability against clinical patient…/IR-3 claim 31 came back unconfirmed: bench-and-field work, not a research lookup, so no ledger request was raised.

• The statement: claim 33unconfirmed — ISO 17034 reference-material-producer accreditation is achievable on the stated budget and timeline.
• What the check found: not checkable by any connector — no registry or market API prices accreditation cost or calendar. Nothing in the file beyond the Generator's own unattributed estimate speaks to it.
• Who depended on it: A12 ISO 17034 reference-material-producer… — Operations (C5).
• What it costs: claim 20unconfirmed's capital estimate — the file's single strongest-scored factor — names ISO 17034-oriented documentation as a cost driver with no quote behind it; an unfavorable finding understates that figure directly.
• Outcome: commission
• Because: a same-day, low-cost human action settles this — a vendor quote and calendar from an accrediting body (e.g. A2LA), which Operations names as the settling artifact. This is a human-attached quote, not a cross-department research question, so no ledger request was raised against it, the same treatment idea-002's IR-1 claims 1 gave a comparable same-day fix.

• The statement: claim 20unconfirmed — first revenue is achievable for low-single-digit millions, dominated by dressing/dPCR tooling, donor plasma sourcing, and ISO 17034-oriented documentation, with no instrument development, clinical trial, or premarket submission.
• What the check found: not checkable — an internal cost estimate is not an external fact, and no connector applies. Operations's own reading is that the figure "cannot currently be read as either a pilot-volume cost or a cost that scales to the recurring purchase cadence of the file's own 30-60-laboratory addressable segment" because it names no production volume (A14 Claim 20's capital-intensity estimate for…/claim 58verified).
• Who depended on it: A13 Capital intensity — Operations (C6), and indirectly the file's own Scorer, whose Capital intensity factor (4/5, the strongest-scored factor in the file) rests on this claim alone.
• What it costs: if the figure prices only a demonstration-lot bench build, the file's own stated main reason to be interesting — cheap, fast, no premarket review — loses the "cheap" half even if "fast" and "no premarket review" independently hold.
• Outcome: commission
• Because: REQ-0032 Does the low-single-digit-million capital… (commercial → operations-npi) and REQ-0034 Does the business case assume a target… (operations-npi → commercial) jointly already commission this from two directions — whether the capital figure clears margin at commercial scale, and whether any production-volume assumption exists anywhere in the file to cost against. No further commissioning is needed.

• The statement: claim 9unconfirmed — the Medicare CLFS rate for CPT 0340U was $3,500.00 for the 7/1/21-3/31/22 period.
• What the check found, 2026-09-14: Market Access re-ran payment 0340U --jurisdiction US cold and it returned unconfirmed, `detail: "No local cache at data/hcpcs_payment_rates.csv"` — a missing input file, not a negative finding.
• **What changed, 2026-09-15 (claim 10refuted's sibling gap, split out here since claim 10refuted itself moved to refuted — see IR-20 claim 10 came back refuted below):** REQ-0036 Populate data/hcpcs_payment_rates.csv and… was partially worked. The volume half resolved (see IR-20 claim 10 came back refuted). The rate half did not: the Verifier fetched the CMS (Centers for Medicare & Medicaid Services) CLFS PLA-code data file directly (https://www.cms.gov/.../26clabq3) and confirmed its only download link resolves to an AMA end-user-license click-through page, not a data file — an empirically reproduced access barrier, the same license-gate class this repo already treats as structurally out of reach for CPT/PLA descriptor content (claim 8unconfirmed). A separate, real CMS utilization figure was also recovered as a byproduct (CY2024 claims-averaged Medicare allowed amount, $3,518.68, close to the cited $3,500) but the Verifier is explicit this is not the CLFS schedule rate claim 9unconfirmed asks about and must not be read as confirming it. data/hcpcs_payment_rates.csv remains unpopulated — not for lack of trying, but because the only CMS distribution of this file is gated behind an AMA license click-through.
• Who depended on it: A19 TAM component 1 — Market Access (C2), Commercial (C3).
• What it costs: one of the two price/volume terms of the file's Stage 2 draft TAM arithmetic (~100k tests × $3,500 × 0.5-1.5% ≈ $1.75-5.25M) stays unconfirmed — and, per IR-20 claim 10 came back refuted, the other term is now refuted, not merely missing.
• Outcome: commission
• Because: this is now a structural gap (an AMA license wall), not merely a fixable missing-cache one — closer in kind to claim 8unconfirmed's permanently-unverifiable CPT descriptor than to a simple cache-population task. REQ-0036 Populate data/hcpcs_payment_rates.csv and… stays open on this half specifically for that reason; no further commissioning beyond it is warranted unless a human is willing to accept the AMA license terms to obtain the file, which is a licensing decision for a human, not a research task.

• The statement: claim 13unconfirmed — MolDX Billing and Coding article A58456 ("MolDX: Minimal Residual Disease Testing for Solid Tumor Cancers") exists in the CMS Medicare Coverage Database and states a positive coverage stance.
• What the check found: two distinct gaps, both confirmed by Market Access's own cold re-run: (a) fixable — data/medicare_coverage_policies.csv is absent; (b) structural — the coverage connector keys on a HCPCS (Healthcare Common Procedure Coding System)/CPT code, not a MolDX article ID, so a populated cache would still not confirm this specific article exists.
• Who depended on it: A21 MolDX Billing and Coding article A58456 — Market Access (C4).
• What it costs: the strongest available signal that the downstream clinical test this panel validates is actually reimbursed.
• Outcome: commission
• Because: REQ-0025 Retrieve MolDX Billing and Coding article… (market-access → research) is already open and asks exactly this — a human opening article A58456 directly in the CMS Medicare Coverage Database by its own article ID, since no cache-restoration alone would settle it.

• The statement: claim 15unconfirmed — FDA's 2024 laboratory developed test final rule was vacated by a federal district court in March 2025, keeping LDT-based MRD assays under CLIA/CAP rather than FDA premarket review.
• What the check found: no connector in this repo queries federal court dockets or opinions. Case name, court, docket and date are all unchecked.
• Who depended on it: A22 FDA's 2024 laboratory developed test final… — Market Access (C5).
• What it costs: Market Access names this "material to TAM in both directions" — a reversal changes who buys this panel and under what regulatory posture.
• Outcome: commission
• Because: REQ-0026 Is FDA's 2024 laboratory developed test… (market-access → regulatory) is already open and asks exactly this.

• The statement: claim 19unconfirmed — the durable moat asset is the orthogonal characterization dataset and traceability/uncertainty documentation a laboratory can paste into a CAP validation packet, plus lot-anchoring switching cost, conceding the material itself is reproducible by any competent competitor.
• What the check found: a judgment claim with no primary source and no connector. The load-bearing sub-question — whether a CAP validation packet requires, merely accepts, or ignores third-party traceable reference material — needs a human read of ISO 17034 and the CAP Molecular Pathology (MOL) checklist, neither machine-queryable here.
• Who depended on it: A25 Moat lever — Legal & IP (C2), which states plainly the moat "is a real category of asset only if" the answer is "requires."
• What it costs: the moat/defensibility factor (already scored 2/5, Low) loses even its weak positive basis if the answer is "merely accepts" or "ignores."
• Outcome: commission
• Because: REQ-0028 Read ISO 17034 (legal-ip → research) is already open and asks exactly this.

• The statement: claim 11unconfirmed — reference-material and external QC spend runs 0.5%-1.5% of laboratory test revenue for high-complexity molecular assays. `claim 12unconfirmed` — the addressable segment is the 30-60 US laboratories running or developing tumor-informed MRD assays.
• What the check found: both returned unverifiable from market by design — no free, authoritative, machine-queryable source for either figure.
• Who depended on it: A27 TAM component 4, A28 TAM component 3 — Commercial (C2, C5).
• What it costs: the spend-share and population terms of the file's only worked TAM arithmetic; Commercial's own note is that a materially lower true share could push the already-marginal TAM score (2/5) across, not merely near, the kill floor.
• Outcome: commission
• Because: these are purchasable, well-defined figures (a licensed lab-economics report; a MolDX/CAP/NY-CLEP directory enumeration), not findings the connector surface refutes — the textbook unconfirmed-is-a-gap-in-the-check case the protocol names. No request existed for either; the front end raised REQ-0037 Obtain a licensed lab-economics report on… (front-end → research) while opening this loop record.

• The statement: claim 21unconfirmed — MRD laboratories prefer a third-party commutable panel over the in-house contrived controls they already build, and will anchor a published LoD claim to an external lot. claim 26unconfirmed — assay-development scientists select and purchase reference material directly as a consumable, without a capital-procurement process.
• What the check found: both returned unverifiable from adoption by design — stakeholder-future claims with no free, authoritative source; the connector's own instruction not to substitute a model's estimate of stakeholder behaviour was followed.
• Who depended on it: A32 Assumption, A33 Assay-development scientists select… — Commercial (C10, C9).
• What it costs: Commercial's own "What would change my mind" states a negative answer on switching (claim 21unconfirmed) alone "is enough to close Routes A and B/C without spending anything further on sizing" — the single most consequential open demand-side question in the file by Commercial's own account.
• Outcome: commission
• Because: Commercial's own position names the exact settling instrument — a short structured survey of 10-15 MRD laboratory directors and assay-development scientists — as the cheapest move that would resolve C6, C7, C9 and C10 (A29 By year 3, A30 A molecular laboratory or assay…, A33 Assay-development scientists select…, A32 Assumption) together, but had not filed it as a ledger request. The front end raised REQ-0038 Run a short structured survey of 10-15 MRD… (front-end → research) for it while opening this loop record.

• The statement: claim 52unconfirmed — a QC material whose entire purpose is validating a tumor-informed ctDNA MRD assay is not itself "intended for use in the diagnosis, monitoring, or screening of neoplastic diseases" within 862.9(c)(1)/866.9(c)(1), so the 510(k) exemptions at 862.1660 and 866.5910 are not withdrawn. `claim 53unconfirmed` — 866.5910's cystic-fibrosis-scoped generic type is not read under 866.9(a) as a different intended use from this candidate.
• What the check found: the Verifier re-ran both underlying regulation texts (connectors regulation 862.9, connectors regulation 866.9, title 21, both verified, source_ref eCFR versioner issue 2026-09-11) and both re-confirmed the neoplastic-disease clause and the different-intended-use clause verbatim — consistent with what claims 5 and 22 already established on an earlier pass. But the Verifier is explicit that **neither call, nor any connector in this repo, answers what the claims actually assert**: whether this specific device is or is not "intended for" neoplastic-disease diagnosis/monitoring/ screening, and whether an oncology cfDNA MRD panel is or is not "a different medical purpose" from the CF-scoped 866.5910 generic type, are FDA intended-use interpretive judgments that regulation text does not settle. The Verifier's own words: "the chair's condition restates the open question as a resolved fact in the candidate's favour; nothing checked this pass moves it from open to resolved in either direction." This is not new information — it is the same open question claims 5 and 22 already flagged, confirmed still open by a second, independent connector run.
• Who depended on it: A4 A quality control material whose entire…, A5 21 CFR 866.5910's generic type — Regulatory (C3, C4), and indirectly the entire clinical-QC second-SKU economics, which the candidate's own "Indication-scope note" describes as "largely 510(k)-exempt" on a reading these two claims do not establish.
• What it costs: the same consequence already named in A4 A quality control material whose entire…/A5 21 CFR 866.5910's generic type — if either exclusion is read as reaching this material, the second SKU is a 510(k) submission rather than a registration-and-listing exercise, a materially different capital and calendar picture. This review changes nothing about that consequence; it confirms the question deciding it is still open.
• Outcome: commission
• Because: this is a structural gap, not a data gap — it needs a human FDA regulatory determination (an actual Q-sub or classification request), which no research report can substitute for, but a precedent search can narrow the uncertainty before committing to one. No existing request reaches this: REQ-0024 Does FDA guidance "Distribution of In Vitro… (regulatory → research) settles a distinct regulatory question (RUO/IUO marketing conduct, claims 1-3), not this one. The front end raised REQ-0039 Is an FDA pre-submission (front-end → regulatory) while updating this loop record, asking both whether a Q-sub is the right next step and whether any existing FDA classification precedent on comparably-scoped QC materials narrows the answer first.

• The statement: claim 54unconfirmed — no live, enforceable US patent held by a named incumbent claims a method reading on this candidate's mechanism. `claim 56unconfirmed` — the sub-0.01% VAF characterization/spiking protocol occupies available (unclaimed) patent scope.
• What the check found: nothing was run for either — both name no patent number, so there is no query to issue, and PATENTSVIEW_API_KEY is confirmed unset in this checkout this pass (re-checked). The patent connector takes a single patent_id and returns filing/grant status for that family only; it performs no landscape or freedom-to-operate search and cannot evaluate a claim about the absence of any blocking patent across eight named incumbents, or about open claim scope across an unbounded prior-art landscape. The Verifier states this is the identical structural gap to claim 18unconfirmed — the original, unscoped FTO (freedom to operate) claim already on file before this round table.
• Who depended on it: A23 No live, A24 The specific sub-0.01%-VAF… — Legal & IP (C1, C3), the two conditions its own abstains verdict is built on.
• What it costs: per A23 No live/A24 The specific sub-0.01%-VAF…, both Route A's moat and Route D's licensing asset are unresolved in either direction — Legal & IP's own words, unchanged by this pass: "any FTO line in a score, deck, or business case is an invention until then."
• Outcome: commission
• Because: REQ-0027 Run a patent/prior-art landscape search (legal-ip → research) is already open and asks exactly this — a keyword/assignee patent-prior-art landscape search over the named incumbents, explicitly scoped to settle both the FTO question (C1/claim 54unconfirmed) and, as a named byproduct, the open patent-scope question (C3/claim 56unconfirmed). No further commissioning is needed.

• The statement: claim 55unconfirmed — no CPT/HCPCS code, NCD, LCD, or Medicare/Medicaid coverage determination applies to a reference-material panel of this kind, because it is purchased as a laboratory operating consumable rather than billed to a third-party payer.
• What the check found: nothing to query — the product has no code and no name under which to look one up, and a null connector return would not establish "none is achievable" in any case, since that is a forward-looking negative about a code that does not exist. The Verifier states plainly: "this is not a new gap; it restates claim 14unconfirmed, which is already unconfirmed on identical grounds." Structural, not fixable by any means available here.
• Who depended on it: A18 No CPT/HCPCS code — Market Access (C1).
• What it costs: nothing new. Market Access's own position already anticipated exactly this outcome and does not wait on it: "I read [claim 14unconfirmed] as correctly framed but unfalsifiable as worded... the more precise and checkable version of the same fact is what I am binding as [C1]... On this half, the Scorer's Factor 2 treatment — n/a, not 0 — is consistent with what my field's tools would tell it if asked, and I have no quarrel with it." Market Access's conditional verdict turns on the recurring-demand question (A19 TAM component 1-A22 FDA's 2024 laboratory developed test final…), not on this claim ever resolving to verified — it is structurally unable to, by the chair's own design, and the chair already treats that as the expected and correct state.
• Outcome: absorb
• Because: the only dependant is Market Access's C1, and Market Access's own verdict does not wait on this claim resolving — it already proceeds on the reading that "no code exists" is the correct and permanent state for a laboratory consumable, the same reading the Scorer independently reached for Factor 2. Nothing in this register or any position is blocked behind claim 55unconfirmed ever moving off unconfirmed, and reworking the claim into a checkable form (as Market Access already attempted once, deliberately) would only reproduce the same structurally unfalsifiable negative. No further commissioning serves a purpose.

• The statement: claim 57unconfirmed — the reference-material/external-QC spend pool is correctly sized as a share of downstream Medicare CLFS test revenue for CPT 0340U, rather than by a bottom-up per-laboratory spend estimate.
• What the check found: no connector applies — this is a claim about which of two TAM sizing methodologies is more correct, not an external fact a database holds. The Verifier notes both methods depend on the same unconfirmed inputs regardless of which is chosen: claim 9unconfirmed (CLFS rate, fixable — missing cache), claim 10refuted (Medicare volume, fixable — missing cache), and claim 11unconfirmed (QC-spend-share, permanently unverifiable market claim). "Choosing the top-down method over the bottom-up one does not resolve any of those three gaps."
• Who depended on it: A26 The reference-material and external-QC… — Commercial (C1).
• What it costs: per A26 The reference-material and external-QC…, the file's only worked TAM arithmetic (~$1.75-5.25M) stays without independent cross-check, possibly mis-sized by an unknown factor.
• Outcome: commission
• Because: this claim cannot be settled ahead of, or independently of, its own component inputs — REQ-0036 Populate data/hcpcs_payment_rates.csv and… (front-end → research, claims 9/10) and REQ-0037 Obtain a licensed lab-economics report on… (front-end → research, claims 11/12) are already open and commission exactly those inputs. Once they resolve, the top-down-vs-bottom-up cross-check A26 The reference-material and external-QC… needs becomes a human comparison of two now-evidenced figures, not a further connector question; no separate request is needed for the methodology choice itself.

• The statement: claim 60unconfirmed — the panel's characterization/value-assignment process at 0.001%-0.01% VAF can be performed repeatably across production lots without the value-assignment-error, stability-drift, or vial-to-vial-imprecision failure modes that account for a large share of FDA recalls under product codes JJX and JJY.
• What the check found: a fused claim, and the Verifier split it. The historical premise is corroborated: `connectors recalls --product-code JJX/JJY --since 10y --jurisdiction US` returned 22 and 62 (50-of-62 shown) records respectively, and reading the reasons, a substantial share do cite exactly the three named failure modes by name (value-assignment errors, stability drift, vial-to-vial imprecision), alongside a separate, also-substantial cluster of unrelated causes (cold-chain/logistics, mislabeling, breakage) — "a large share" is a defensible, not an overstated, characterization. The forward-looking half — whether this candidate's still-undesigned process will avoid those same failure modes — is "not verified and not verifiable here": no connector predicts manufacturing outcomes, and the tag reflects the claim as a whole, held at unconfirmed.
• Who depended on it: A16 The panel's characterization and… — Operations (C9).
• What it costs: per A16 The panel's characterization and…, the same failure patterns that dominate the closest comparable device family's recall history recur here, at the VAF range this candidate's own file already flags as most fragile — unresolved either way by this pass.
• Outcome: commission
• Because: the settling artifact is bench/build-plan work the product does not yet have — Operations's own C9 names it directly: "a documented lot-release/value-assignment SOP with inter-lot CV data at the claimed uncertainty budget, spanning at least three independently produced lots." That is downstream of the product existing at all (contingent on A6 A dPCR replicate series of at least 20…/A7 Technical/mechanism risk resolving first), the same shape as IR-3 claim 31 came back unconfirmed/IR-4 claims 30 and 32 came back unconfirmed's bench items in the first pass; no ledger request is raised for work that cannot be commissioned until there is a lot to test.

• The statement: claim 62unconfirmed — a dPCR replicate series of at least 20 replicates per level, at the panel's own stated 10-30 ng input mass, demonstrates a defensible uncertainty budget at the lowest target VAF levels rather than being dominated by Poisson counting-statistics noise.
• What the check found: not checkable — a bench/physics claim about an experiment that has not been run. The Verifier states this is "a more specific, protocol-level restatement of claim 16unconfirmed, this file's already-unconfirmed named largest technical risk," and declined to run literature/literature-detail as a substitute, for the same reason recorded at claim 16unconfirmed: a paper hit count on dPCR replicate design or Poisson-limited detection would neither confirm nor refute this specific, not-yet-generated dataset.
• Who depended on it: A6 A dPCR replicate series of at least 20… — Clinical (C1), the file's own named largest technical risk, independently bound by four chairs (see A6 A dPCR replicate series of at least 20…/A7 Technical/mechanism risk's full consequence chain in the register and IR-2 claim 16 came back unconfirmed above).
• What it costs: unchanged from A6 A dPCR replicate series of at least 20…/A7 Technical/mechanism risk/IR-2 claim 16 came back unconfirmed — the entire commercial differentiator over three incumbents who stop at 0.1% VAF.
• Outcome: commission
• Because: REQ-0030 Is orthogonal dPCR quantification of… (commercial → technology) already asks this precise bench question, and every chair naming a settling artifact for this family (A6 A dPCR replicate series of at least 20…, A7 Technical/mechanism risk, A9 A single pooled-donor-plasma…) names the same one — an internal dPCR replicate series at the stated input mass. This is the same review outcome as IR-2 claim 16 came back unconfirmed, applied to a second, independently-bound claim citing the identical underlying experiment; it does not need a second request, only a second review, since check_loop.py tracks failure coverage per claim number rather than per underlying question.

• The statement: claim 63unconfirmed — a single pooled-donor-plasma lot can simultaneously retain native background sufficient for commutability and carry an orthogonally dPCR-quantified ultra-low VAF spike whose uncertainty budget is not degraded by that background.
• What the check found: not checkable — a bench/technical-design claim no connector evaluates. The Verifier notes it "does not resolve, and is not resolved by, claim 30unconfirmed (donor plasma lot-to-lot fragmentomic consistency) or claim 31unconfirmed (cross-platform commutability) — all three are open technical questions about the same material, none checkable here."
• Who depended on it: A9 A single pooled-donor-plasma… — Technology (C5), the integration condition — Technology's own words: this sits "at the seam between C1 and C2," and nothing in the file puts the two questions together anywhere else.
• What it costs: per A9 A single pooled-donor-plasma…, commutability (A8 Commutability against clinical patient…) and the low-VAF uncertainty budget (A6 A dPCR replicate series of at least 20…/A7 Technical/mechanism risk) could each resolve favorably in isolation and the product could still fail at the seam between them — a distinct failure mode none of the other reviews in this register would catch.
• Outcome: commission
• Because: the settling artifact is named directly by Technology — "a single bench study reporting both a commutability metric and a dPCR-derived uncertainty budget on the same lot, at the same VAF level" — and it is bench/bioinformatics work outside any connector's reach. REQ-0030 Is orthogonal dPCR quantification of… is adjacent (it tests the dPCR uncertainty-budget half alone) but does not reach the joint question this claim asks; no ledger request currently commissions the combined study, and none can usefully be raised until a lot exists to characterize both ways at once — the same reasoning IR-3 claim 31 came back unconfirmed, IR-4 claims 30 and 32 came back unconfirmed and IR-17 claim 60 came back unconfirmed already apply to comparable pre-hardware bench items.

The first refuted finding in this loop. Every review through IR-19 claim 63 came back unconfirmed routed commission because every failure was unconfirmed — a gap in the check. This one is different: the check ran, cleanly, against real data, and the claim as worded is wrong.

• The statement: claim 10refuted — annual Medicare claim volume for CPT 0340U exceeds 100,000 tests.
• What the check found: data/medicare_procedure_volumes.csv was populated 2026-09-15 from a real CMS release (MUP_PHY_R26_P05_V10_D24_Geo.csv, "Medicare Physician & Other Practitioners — by Geography and Service," CY2024 claims; URL and sha256 recorded in data/provenance.json, retrieved 2026-09-15T05:10:22+00:00). claim 10refuted's own connector command, run cold against that cache — `connectors procedures 0340U --expect-volume 100000 --jurisdiction US — returned refuted`: "0340U (national, office/non-facility 25,746): claimed 100,000 overstates the cached figure of 25,746 by 288%, outside the 25% tolerance." The Verifier independently re-ran the same command a second time in the course of checking an unrelated claim (68) and got the identical figure, and separately opened the cached CSV by hand and confirmed Tot_Srvcs=25,746, Tot_Benes=11,922 for the national, office/non-facility row, with no national facility row present in the release. This is not a connector artifact or a scope mismatch on the check's part — it is a clean read of real, dated, provenance-tracked CMS data against a claim that named a specific number and a specific jurisdiction (Medicare), and the number is wrong by roughly 4x, not merely short. What this finding does not establish: the connector's own caveat, preserved in the claim's note, is that "a claim can legitimately exceed this Medicare-FFS figure if it counts non-Medicare volume." Medicare FFS excludes Medicare Advantage (Medicare Advantage), commercial, Medicaid, VA, and cash-pay. claim 10refuted's own text says "Medicare claim volume," not "total US claim volume," so this escape hatch is not available on the claim's own wording — the claim was about Medicare specifically, and Medicare specifically is refuted. What stays genuinely open is a different proposition nobody has asked or answered anywhere in this file: whether total US cross-payer volume for 0340U (Medicare FFS + Medicare Advantage + commercial + Medicaid + cash-pay) clears 100,000, or any other bar the TAM chain needs.
• Who depended on it: A20 TAM component 2 — Market Access (C3) and Commercial (C4), both of which already flagged, before this result came back, that even a favorable answer would only ever be "a Medicare FFS floor, never a market size." Indirectly, A19 TAM component 1 (claim 9unconfirmed, IR-7 claim 9 came back unconfirmed) and A26 The reference-material and external-QC… (claim 57unconfirmed, IR-16 claim 57 came back unconfirmed), since all three feed the same Stage 2 draft TAM arithmetic (~100k tests × $3,500 × 0.5-1.5% ≈ $1.75-5.25M).
• What it costs: the specific number is gone — the file's only worked TAM arithmetic loses the volume anchor it was built on, and a same-shaped calculation using the actual Medicare-FFS figure (25,746 × $3,500 × 0.5-1.5% ≈ $450K-$1.35M) would land roughly a quarter of the original range, on the Medicare-FFS slice alone. What it does not cost: the Verified Composite's TAM factor was already left blank at Stage 4, not scored, precisely because none of its four components (claims 9, 10, 11, 12) had a verified figure behind them — the composite score does not move, because it was never counting this number in the first place, whether unconfirmed or refuted. Market Access's conditional verdict and Commercial's conditional verdict both already treated the whole downstream-reimbursement chain as unestablished, not as resting on this one figure specifically.
• Outcome: commission
• Because: I considered pivot and absorb first and rejected both. Not absorb — Market Access's C3 and Commercial's C4 are real, named dependants, and my own register text (A20 TAM component 2) already stated plainly what breaks: "a low or absent figure removes the one anchor the downstream-test-revenue chain has at all." That is not nothing, and an absorb that said so would be the shrug-with-a-heading the protocol warns against. Not pivot — a pivot changes the shape of the opportunity (intended use, target, payer route, site of care), recorded against a frontmatter field. Nothing about this finding narrows or redirects what this candidate is or who it sells to; it changes one input to a business-case spreadsheet that was already excluded from the scored composite and was already understood by both dependent chairs, before the check ran, to be a floor rather than a market size. Re-sizing a TAM estimate downward is ordinary commercial work, not a change of shape — treating it as cause for a pivot here would be the "redesigned because something looked worse" move the protocol distinguishes from an actual pivot, just run in the opposite (worse-news) direction. Commission, on the genuinely different question, not the settled one. Per the protocol's own asymmetry note, re-raising a request to re-check Medicare-FFS volume for 0340U would be re-litigating a clean, dated, provenance-tracked finding — that number is settled and should not be re-asked. But the claim's dependants never actually needed the Medicare-FFS number specifically; they needed to know whether the downstream test market is large enough to support a meaningful QC-material spend pool, and Commercial's own C1 (claim 57unconfirmed) already named the surviving path around this exact dead end: a bottom-up per-laboratory spend estimate instead of a top-down Medicare-volume-anchored one. REQ-0037 Obtain a licensed lab-economics report on… (front-end → research, already open) commissions exactly the two inputs that bottom-up method needs (addressable lab count, QC-spend-as-percentage-of-revenue). No new request is raised for total cross-payer 0340U volume specifically: granular multi-payer claims volume for a single PLA code is not published by any free, authoritative source I can identify, and is the same shape of permanently-unconfirmable market-class question claims 11 and 12 already occupy — raising it as a fresh request would manufacture a second version of a gap this loop record has already named once.

• The statement: claim 47unconfirmed — by year 3 a third-party commercial reference panel captures at least a low double-digit share of addressable US MRD-laboratory reference-material spend, rather than losing it to in-house contrived controls or the three incumbent vendors. claim 48unconfirmed — a laboratory or assay manufacturer will pay a per-panel price materially above the incumbents' list price (LGC Seraseq, Horizon/Revvity HDx, Twist Bioscience) for a panel characterized into the sub-0.01% VAF range.
• What the check found: both routed to adoption, both returned unverifiable, both with the same detail — asserts what clinicians, payers or patients would do; no free, authoritative, machine-queryable source answers that; **stays unconfirmed permanently unless a human attaches real research.** The Verifier substituted no opinion of its own for stakeholder behaviour on either, which is the correct handling and worth naming because these are exactly the two claims where a plausible-sounding guess would have been easiest to write. It also recorded a housekeeping finding the register should carry: both claims had been sitting under the literal string unverified where the grammar has four valid tags and that is not a spelling of any of them — corrected on this pass, and a reminder that a claim can be malformed as well as unchecked.
• Who depended on it: register rows A29 By year 3 and A30 A molecular laboratory or assay… — Commercial (C6, C7). Per Commercial's own "What would change my mind," these are two of the four conditions (with C9/A33 Assay-development scientists select… and C10/A32 Assumption) that one lab-director survey would settle together as **the entire demand side of this business case**. A31 The gross margin earned on the panel depends on A30 A molecular laboratory or assay… downstream: a gross-margin condition cannot even be posed without a defensible price point.
• What it costs: whether anyone buys this, and at what price. Every revenue and margin figure in the business case sits on top of these two, and they are the demand half of a file whose supply half (A6 A dPCR replicate series of at least 20…–A9 A single pooled-donor-plasma…, the sub-0.01% characterization) is separately unbuilt. The asymmetry is worth stating: claim 48unconfirmed failing kills the premium pricing the whole model assumes; claim 47unconfirmed failing means the product is sellable but not at a share that supports a standalone business. Those are different outcomes and the survey distinguishes them.
• Outcome: commission
• Because: unverifiable from adoption is the protocol's cleanest commission case — the gap is in the check, permanently, and the instrument that closes it is primary research rather than any connector this repo will ever hold. REQ-0038 Run a short structured survey of 10-15 MRD… is that commission, front end → research, open: a structured survey of 10–15 MRD laboratory directors and assay-development scientists covering what reference material they use today, what they pay for it, whether they would switch to or pay a premium for a sub-0.01% VAF panel, and whether they can purchase it directly.
• On adopting an existing request rather than raising a new one — the IR-6 claim 20 came back unconfirmed test, applied. idea-001's IR-6 claim 20 came back unconfirmed recorded that folding a commission into another review's request is only honest if the request's question carries it, and that record's own failure was folding the RTM (remote therapeutic monitoring) question into a request whose question never did. So: REQ-0038 Run a short structured survey of 10-15 MRD…'s **settles line names only claims 21 and 26**, and on that line alone this would be the same mistake. Its question asks what laboratories pay today and whether they would pay a premium for a sub-0.01% VAF panel — which is claim 48unconfirmed almost verbatim, and is the switching behaviour claim 47unconfirmed's share forecast rests on. The question carries it, so the request is adopted and no second survey is raised against the same fifteen people. **What REQ-0038 Run a short structured survey of 10-15 MRD… cannot do is settle claim 47unconfirmed outright**: stated willingness to switch bounds a year-3 share forecast, it does not establish one, and the honest reading of a favourable survey is that claim 47unconfirmed becomes better-founded and stays unconfirmed. Recorded so that a clean survey result is not later read as having verified a forecast.

• The statement: claim 49unconfirmed — gross margin on the panel, once donor-plasma sourcing, orthogonal dPCR characterization and ISO 17034-oriented documentation costs are included, is high enough to support a standalone reference-material business at the addressable laboratory count and purchase cadence.
• What the check found: routed to market, returned unverifiable — no connector prices costs or margins, and the claim is internal financial modeling rather than an external fact any typed connector holds. The Verifier added the dependency that matters more than the routing: this claim inherits the unresolved TAM component gaps (claims 9–12) and the un-costed capital estimate (claim 20unconfirmed). It is not one unverified number; it is an arithmetic over four of them.
• Who depended on it: register row A31 The gross margin earned on the panel — Commercial (C8). Nothing depends on A31 The gross margin earned on the panel in turn, and that is the point of it: Commercial describes this as a pure unit-economics kill, independent of whether the product can be built (A6 A dPCR replicate series of at least 20…–A9 A single pooled-donor-plasma…) or sold (A29 By year 3, A30 A molecular laboratory or assay…) at all.
• What it costs: the whole business, on a path none of the other rows cover. A candidate can survive its technology risk and its demand risk and still not clear margin at the addressable count and cadence — 30–60 US laboratories buying on some cadence nobody has established is a small denominator, and claim 11unconfirmed's spend-share multiplier (A28 TAM component 3) is itself permanently unverifiable. This is the row that would kill the opportunity quietly, after everything else came back fine.
• Outcome: commission
• Because: REQ-0032 Does the low-single-digit-million capital… is the vehicle and its question carries this one on the same IR-6 claim 20 came back unconfirmed test applied above: it asks whether the capital estimate funds production "at a scale and cost-of-goods consistent with a per-panel price that clears margin against the incumbent vendors' list prices, or... pilot-volume tooling only," and its settles line names Commercial's gross-margin condition explicitly — which is C8, which is this claim. Open, commercial → operations-npi, no second request needed.
• The dependency is the deliverable here, and it is stated rather than buried. A31 The gross margin earned on the panel cannot be answered before its inputs are: the price comes from A30 A molecular laboratory or assay… (IR-21 claims 47 and 48 came back unconfirmed, REQ-0038 Run a short structured survey of 10-15 MRD…), the production scale and cost-of-goods from REQ-0032 Does the low-single-digit-million capital…'s own answer and REQ-0034 Does the business case assume a target…'s production-volume assumption, the addressable count from A27 TAM component 4/A28 TAM component 3 (REQ-0037 Obtain a licensed lab-economics report on…, permanently unverifiable), and the capital base from claim 20unconfirmed (IR-6 claim 20 came back unconfirmed). **So this commission is real but it is last in a chain of five**, and a reader who sees commission on all three of A29 By year 3, A30 A molecular laboratory or assay… and A31 The gross margin earned on the panel should not read three parallel pieces of work. Absorbing was not available: Commercial wrote this as a kill condition, and a kill condition with no dependants listed is still a kill condition.

• The statement: claim 1unconfirmed: a properly labeled RUO product is not subject to FDA premarket notification or approval. claim 3unconfirmed: marketing this material to CLIA laboratories to validate clinical MRD tests is the conduct the 2013 guidance treats as defeating RUO status. claim 130unconfirmed: the P0 the original shape intended use, which included "ongoing performance monitoring" of assays run on patient specimens, is consistent with 809.10(c)(2)(i)'s research-phase condition. The context rows are claim 2unconfirmed, claim 64unconfirmed and claim 65unconfirmed.
• What the check found:
  • claim 1unconfirmed: regulation 809.10 --title 21 returned verified (eCFR issue 2026-09-22) for the text of the condition, "in the laboratory research phase of development, and not represented as an effective in vitro diagnostic product". The opening words of 809.10(c) exempt only from paragraphs (a)/(b) and a part 861 standard. The regulation never mentions premarket review, so the claim's load-bearing half is not supported by its own source.
  • Claims 64-65: REQ-0024 Does FDA guidance "Distribution of In Vitro… retrieved the 2013 guidance (Docket 2011-D-0305). The Verifier re-fetched it and confirmed the quoted text as accurate. Section IV.B names three kinds of conduct: soliciting clinical laboratories, "assistance in performing clinical validation", and validate-then-offer-as-LDT messaging. Section II p.6 says: "Mere placement of an RUO or IUO label... does not render the device exempt... FDA may determine that the device is intended for use in clinical diagnosis based on... how the device is marketed." Both are held unconfirmed on evidence class only, because no typed connector reads a guidance PDF.
  • claim 130unconfirmed: the regulation half is verified. Whether the P0 the original shape sentence meets the condition is an FDA interpretive judgment, and no connector makes one.
  • Regulatory's reading, which cites these tags and issues none: the sentence says "not for use in diagnostic procedures" and then describes quality control of patient runs, which is a use in diagnostic procedures. Route A plus the claim 19unconfirmed pitch is close to all three kinds of conduct in IV.B.
• Who depended on it: A1 In vitro diagnostic products properly…, A2 FDA guidance "Distribution of In Vitro…, A3 Enforcement risk claim, A35 The P0 intended use of idea-003 - which…, and Regulatory behind them (C1, C2, C5). Commercial's whole Route A depended on it too: REQ-0029 Does the file's own stated principal… said "every dollar in the money model depends on" this, and it was answered against Route A as scoped. So did the "no premarket submission" clause of claims 20 and 50.
• What it costs: the P0 the original shape launch route. That route is an RUO label on a product sold to CLIA laboratories for validation and daily QC of patient runs, and in Regulatory's field it is not a compliant posture. Incumbents do sell similar material RUO (claim 123unconfirmed), so it is sellable. It is also the exposure a diligence room discounts, and its size is unknown until REQ-0068 Since the November 2013 RUO/IUO… reports how often FDA enforces.
• Outcome: pivot
• Because: Commission was considered first, because the tags are unconfirmed. Two vehicles exist, and neither would move the block. REQ-0068 Since the November 2013 RUO/IUO… (enforcement letters) measures how often FDA acts, not what the rule is. A 513(g) on the P0 the original shape sentence would pay FDA to rule on a clause the chair reads as contradicting itself and the company can simply delete. The unconfirmed here is an evidence-class limit on text the Verifier confirmed as quoted (claims 64-65), plus an interpretive judgment (claim 130unconfirmed). It is not a missing input, so commissioning would be re-litigating a finding under a gap's label. Absorb is not available, because Route A and claims 20/50 depend on it. So this is a pivot. Regulatory named two shapes without the problem, and **the choice between them was made on the evidence, not on the chair's ordering**:
  • (b) Clinical QC material at launch under 866.5910. It rests on one verified precedent, claim 131verified: FDA's SCU code, "Dna And Rna Oncology Controls", is Class II under 866.5910 and marked 510(k) exempt. It also rests on the file's one verified regulatory demand driver, CLIA's control-procedure requirement (claim 7verified), which is exactly the run-by-run use (b) lets the label state. Its open dependencies are claim 52unconfirmed (the neoplastic-disease limitation, which only a 513(g) settles — IR-26 claims 52 and 53, REQ-0081 Draft the 21 CFR 513) and the quality-system cost (REQ-0070 If idea-003 launches as a clinical…, accepted and not yet answered).
  • (a) Narrow RUO, development phase only. Nothing verified supports it. The customer count (claim 154unconfirmed) is unconfirmed. Whether the largest MRD developers are reachable at all is REQ-0079 Under the development-phase-only RUO shape…, still open. Commercial's REQ-0069 If idea-003 launched strictly RUO - sold… answer is that (a) supports a first sale but not the standalone recurring business: it removes the daily-control line, which Commercial computes as most of the recurring revenue (claim 151unconfirmed). It also gives up claim 12unconfirmed's 30-60 clinical labs, the file's whole obtainable segment. (b) is not free, and its costs are recorded in P2 clinical QC material at launch rather than left out. Operations says it adds a QMS layer, removes no condition, and moves claims 20 and 50 up. Clinical says a daily control at a derived ultra-low level must print its per-tube zero-copy probability on the label. If A4 A quality control material whose entire… fails there is a 510(k) before revenue. P2 clinical QC material at launch takes (b). (a) is recorded under "Pivot options on the table", with the triggers that would bring it back (REQ-0070 If idea-003 launches as a clinical… or the 513(g) coming back against (b), or A56 If this candidate's daily-use single-level… failing). See P2 clinical QC material at launch.

• The statement: claim 62unconfirmed: a 20-replicate dPCR series at 10-30 ng input shows a defensible uncertainty budget at 0.001%-0.01% VAF. claim 16unconfirmed: that range is Poisson-limited at those inputs. claim 133unconfirmed: at 10-30 ng, one variant is expected in ~0.3-0.9 molecules at 0.01% and ~0.03-0.09 at 0.001%, so most inputs contain none.
• What the check found:
  • claim 133unconfirmed: the Verifier recomputed it and found it internally correct (3,030-9,091 GE; N = 0.30-0.91 and 0.03-0.09; e^-N = 74%-40% and 97%-91%). It is held unconfirmed only because its inputs are not connector results. The 10-30 ng range is claim 16unconfirmed's, and the 3.3 pg/genome constant is not returned by any connector.
  • Claims 62 and 16: no bench data exists, which is unchanged since IR-2 claim 16 came back unconfirmed and IR-18 claim 62 came back unconfirmed.
  • What is new is the verified floor around them. claim 100verified: in an interlaboratory ddPCR reference-material study, LoB/LoD/LoQ were 0.01%/0.02%/0.1%, and gravimetric agreement held only down to 0.1%. claim 99verified: the best dPCR reference procedure's false-positive rate reaches 0.003%, with LoQ 0.2%.
  • The research report's §5 recorded no NMI or vendor value assignment below ~0.1% with a stated uncertainty, as not found after a search.
  • Technology's arithmetic, which is judgement and not a claim: direct assignment at 0.001% with 30% uncertainty needs ~26 µg of cfDNA per arm at 0.003% background.
• Who depended on it: A6 A dPCR replicate series of at least 20… (Clinical C1), A7 Technical/mechanism risk (Operations C4; Technology C1 until 2026-09-26), A38 At a cfDNA input of 10-30 ng (Clinical C4). Beyond the register, it is the mechanism: clause "each level orthogonally quantified by digital PCR". Commercial's REQ-0030 Is orthogonal dPCR quantification of… framed a negative answer as removing "the product this business case is pricing".
• What it costs: the P0 the original shape mechanism's verb at the bottom levels, not its range. Clinical: "It is the wrong verb." Technology, on the same clause: "cannot work as framed." Both say 0.001% can be reached by derivation, with a dilution-and-heterogeneity budget instead of a measurement budget. Nothing about the matrix, the buyer or the intended use depends on this review.
• Outcome: pivot
• Because: Commission was considered first. The vehicle would be claim 62unconfirmed's bench series, and Clinical, which bound it, now says the arithmetic predicts the result ("Poisson-dominated whoever runs it") and that it "should not be where the next unit of money goes." Funding an experiment whose outcome the verified background floor (claims 99, 100) already predicts is paying to confirm the block. The unconfirmed on claim 133unconfirmed is a gap in the check (a constant no connector returns), not doubt about the arithmetic. P1 derived-value bottom levels also does not rest on it: if 133 were wrong, the verified background leg and Technology's vial-heterogeneity reasoning would still rule out direct measurement at 0.001%. Absorb is not available, because the mechanism: frontmatter asserts the thing. So this is a pivot. Three chairs name the same narrowest exit: Clinical C5, Technology C1, and Operations ("moves me toward supports"). Levels below 0.1% are value-assigned by gravimetric dilution from a dPCR-quantified parent and labelled as derived. The range is kept. Not taken, and why. A floor at 0.01% is Technology's preferred shape on feasibility. But no failed statement is about 0.01% specifically, and whether derivation holds its budget at 0.001% is a live condition (A39 Each of this panel's levels below 0.1% VAF…, A43 At this panel's lowest assigned level). If that condition fails, it is the pre-named trigger for the 0.01% floor. A floor at 0.1% makes this a fourth entrant into a three-vendor market, which Technology and Clinical both say should be evaluated as a different product. Buffer format and an aggregate tumor-fraction build are recorded as options, not taken. See P1 derived-value bottom levels.

• The statement: claim 134unconfirmed: below-0.1% levels can be dilution-assigned within ≤35% relative expanded uncertainty (k=2), including heterogeneity and LoB. claim 135unconfirmed: spike and endogenous cfDNA recover equally across three extraction kits. claim 138unconfirmed: there are ≥100 expected tumor molecules per vial at the lowest level, summed over certified variants. claim 139unconfirmed: background at certified positions in the unspiked pool is ≤ one-tenth of the lowest level.
• What the check found: each Verifier note says the same thing in its own words. These are forward-looking conditions on a lot that has not been made, no connector can observe them, and they need the candidate's own data. claim 138unconfirmed's arithmetic was checked and is internally consistent: 15,000 GE × 0.001% = 0.15 molecules per variant, so ~667 variants are needed for 100 molecules. Nothing was found against any of them. Nothing could have been.
• Who depended on it: A39 Each of this panel's levels below 0.1% VAF… (Clinical C5, Technology C1). This row is P1 derived-value bottom levels's own mechanism, and Clinical names it as what moves it to conditional. A40 Across at least three cfDNA extraction… (Clinical C6, Technology C8), A43 At this panel's lowest assigned level (Technology C6), A44 In the unspiked pooled-donor-plasma lot… (Technology C7). A9 A single pooled-donor-plasma… / claim 63unconfirmed (Technology C5, IR-19 claim 63 came back unconfirmed) is the same lot's joint reading.
• What it costs: if A39 Each of this panel's levels below 0.1% VAF… fails, P1 derived-value bottom levels fails the way P0 the original shape did and the next shape is the 0.01% floor. If A40 Across at least three cfDNA extraction… fails, the assigned value is kit-dependent at every level and the native-plasma matrix goes. If A43 At this panel's lowest assigned level fails at 0.001%, certification must become aggregate tumor fraction or the floor rises. If A44 In the unspiked pooled-donor-plasma lot… fails, positions drop out of certification pool by pool.
• Outcome: commission
• Because: these are textbook unconfirmed: the gap is in the check, because the object does not exist yet. The only artifact that settles them is one pilot-lot programme, which both chairs describe. Clinical's version is weeks of dPCR: one parent, two or three dilution steps, ≥10 units per level, ≥60 wild-type reactions. Technology's version adds a deep-sequenced spike line, an error-corrected background map of the pool, and a multi-kit extraction study, at 9-18 months. Running it is a decision for a human, and no department here can make it. What can be commissioned is the thing a human needs in order to decide: REQ-0082 Specify the single pilot-lot bench… (front end → technology) asks for the programme's design, a pre-specified pass/fail criterion per claim (including claim 63unconfirmed), and its cost and calendar. REQ-0074 Extend report ctdna-mrd-assay-validation.tec… (open) says whether A39 Each of this panel's levels below 0.1% VAF… would be a first. REQ-0075 What is one unit of this panel (open) defines the unit, which A43 At this panel's lowest assigned level's per-vial count needs. Stated plainly so that REQ-0082 Specify the single pilot-lot bench…'s answer is not later read as settling anything: a protocol settles none of these claims. It makes them settleable.

• The statement: claim 52unconfirmed: this material is not "intended for use in the diagnosis, monitoring, or screening of neoplastic diseases" under 866.9(c)(1), so the 866.5910 exemption is not withdrawn. claim 53unconfirmed: 866.5910's CF-scoped type is not read under 866.9(a) as a different intended use.
• What the check found: unchanged since IR-13 claims 52 and 53 came back unconfirmed. Regulation text cannot settle an intended-use judgment. What changed is the precedent. claim 131verified puts an oncology-controls code, SCU, inside 866.5910 and marks it exempt. Regulatory reads that as "largely retiring" the 866.9(a) limb and making the (c)(1) limb "look favourable" but not dispositive: SCU is defined around NGS (next-generation sequencing) IVD test systems, and exemption is always "subject to the limitations", device by device. REQ-0039 Is an FDA pre-submission, the commission IR-13 claims 52 and 53 came back unconfirmed raised, came back with a sequencing answer: the right filing is a 513(g), not a Q-sub. Check claims 131-132 first, pick the shape, then file on the pivoted sentence.
• Who depended on it: A4 A quality control material whose entire…, A5 21 CFR 866.5910's generic type (Regulatory C3, C4). Since P2 clinical QC material at launch, the launch. At P0 the original shape these rows decided the second SKU's economics. Now they decide whether first revenue needs a 510(k).
• What it costs: if either limb goes against the product, a 510(k) is cleared before first distribution. Claims 20 and 50 budget no premarket submission and no FDA review calendar, and claim 20unconfirmed is already read by Operations as a development budget only.
• Outcome: commission
• Because: unconfirmed on an interpretive question that FDA answers on request. Two of Regulatory's four sequencing steps are done: claims 131-132 were checked, and P2 clinical QC material at launch picked the shape. The next step is the filing. REQ-0081 Draft the 21 CFR 513 (front end → regulatory) asks Regulatory to draft the 513(g) on the P2 clinical QC material at launch intended-use sentence, name both limbs, and state what the company must supply and the expected response time. Filing it is a human decision. The draft is the artifact.

• The statement: claim 132unconfirmed: Illumina lists TruSight Oncology DNA/RNA controls under SCU, and SensID GmbH lists Idylla oncology reference sets under NZB, in FDA's registration and listing database.
• What the check found: no connector reads FDA registration and listing, because no subcommand exists. The classification records carry registration numbers without firm names, proprietary names or listing dates (SCU: one, 3003218906; NZB: six). Only the two codes' existence under 866.5910 was confirmed. This is a structural gap, and Regulatory anticipated it when binding C7 ("a Verifier may hold it at unconfirmed on evidence-class grounds, as it did claims 64-65").
• Who depended on it: A37 In FDA's establishment registration and…, with Regulatory (C7) behind it. What it bears on is the 866.9(a) limb (A5 21 CFR 866.5910's generic type), as the listing-level evidence that a reference-material maker, not only an IVD kit maker, sits under 866.5910 for oncology.
• What it costs: that second, more on-point strand of precedent for A5 21 CFR 866.5910's generic type. Nothing else.
• Outcome: absorb
• Because: what depends on it survives without it. A5 21 CFR 866.5910's generic type is carried by the verified classification record itself (claim 131verified), which Regulatory already reads as having "largely retired" that limb. Whatever doubt remains goes into the 513(g) (IR-26 claims 52 and 53, REQ-0081 Draft the 21 CFR 513), which settles A5 21 CFR 866.5910's generic type directly whatever 132 says. No decision in this loop waits on 132. Commissioning it would buy one of two things. A human screenshot of the listing would be held at unconfirmed on evidence class exactly as claims 64-65 were. A new registration-and-listing connector is an engine/ change outside this department's remit, and it is worth raising there on its own merits (Regulatory has now twice needed that endpoint), but not as this opportunity's commission. The row stays live and the chair's condition stands. What the absorb says is that the P2 clinical QC material at launch shape does not wait on it.

• The statement: claim 136unconfirmed: LCD L38779 gives criteria-gated coverage of tumor-informed MRD testing. claim 137unconfirmed: its analytical-validity criteria require LoD shown with known-VAF material without naming a source.
• What the check found: coverage 0340U confirmed that the five billing articles exist and that L38779 is A58376's related LCD. The cache has no coverage-stance column and no LCD prose, and ncd does not read LCDs. The text that would settle both claims has never been read by anything here.
• Who depended on it: A41 Active MolDX LCD L38779 (Market Access C4), A42 MolDX LCD L38779 conditions coverage of a… (Market Access C6). Since P2 clinical QC material at launch, A42 MolDX LCD L38779 conditions coverage of a… is what Market Access calls the central condition, because a clinical QC material sold to MolDX-billing labs is what a MolDX analytical-validity requirement would pull through.
• What it costs: A42 MolDX LCD L38779 conditions coverage of a… decides whether a payer-side demand driver exists at all for P2 clinical QC material at launch's product, alongside CLIA's (claim 7verified, which names no material). A41 Active MolDX LCD L38779 decides whether the downstream test is covered or only priced and paid.
• Outcome: commission
• Because: unconfirmed because a free public document has not been read. REQ-0072 Retrieve active MolDX LCD L38779 (market-access → research, open) is exactly that read, keyed on the LCD ID, and it settles both claims and claim 126unconfirmed. Market Access calls it the last open link in its field. No new row is needed.

• The statement: claim 140unconfirmed: a fully loaded per-panel cost below $1,919.59 at a named volume. claim 142unconfirmed: ISO 17034 initial assessment cannot precede first-lot production. claim 143unconfirmed: the spike source material is obtainable on terms that permit commercial use.
• What the check found:
  • claim 140unconfirmed: the build-up does not exist. Its comparator, claim 117unconfirmed, is itself an unconfirmed web price.
  • claim 142unconfirmed: ISO 17034 is a copyrighted standard that no connector reads, and its timing input (claim 82unconfirmed, A2LA 3-6 months) is unconfirmed.
  • claim 143unconfirmed: the claim names no supplier, and no connector reaches contract terms.
  • All three are gaps in the check.
• Who depended on it: A45 A bottom-up cost-of-goods build-up for one… (Operations C11, with Commercial's C8 / A31 The gross margin earned on the panel reading its output), A47 ISO 17034 requires a reference material… (Operations C13), A48 The variant-bearing source material from… (Operations C14). Through A45 A bottom-up cost-of-goods build-up for one… it reaches A31 The gross margin earned on the panel, the margin kill condition.
• What it costs: A45 A bottom-up cost-of-goods build-up for one… is the floor test: can the plasma-format product be made for less than the buffer-format incumbent sells for. A47 ISO 17034 requires a reference material… is claim 50unconfirmed's calendar. A48 The variant-bearing source material from… decides whether the spike can lawfully go into a product at all.
• Outcome: commission
• Because: each has a named settling artifact and none had a ledger row. REQ-0083 Produce the cost-of-goods build-up your… (front end → operations-npi) asks for the C11 build-up in its P2 clinical QC material at launch (QMS) variant. It uses Commercial's declared volumes (50/5/500 panels, 2,000 vials), depends on REQ-0075 What is one unit of this panel for the unit definition, and is the natural companion to REQ-0070 If idea-003 launches as a clinical…, which Operations accepted and will answer with the same build-up. REQ-0084 Obtain a written answer from an ISO 17034… (front end → research) asks an accrediting body in writing about sequencing and fees. It settles claim 142unconfirmed, and it also carries claim 33unconfirmed (A12 ISO 17034 reference-material-producer…), whose IR-5 claim 33 came back unconfirmed commission never had a row. REQ-0085 Name the variant-bearing source material… (front end → operations-npi) asks for the spike supplier and its terms, including whether deep sequencing is permitted (A43 At this panel's lowest assigned level). It mirrors Legal's REQ-0078 Name the pooled human donor plasma supplier on the plasma side.

• The statement: claim 141unconfirmed: published healthy-donor cfDNA central values are ≤10 ng/mL, so a 10-30 ng extraction needs at least 1-3 mL of native plasma per replicate.
• What the check found: the Verifier ran literature and literature-detail 27834061 and recorded counter-evidence, not support. That study (Ann Lab Med 2017, duplex real-time PCR, 5,442 healthy adults) reports median plasma DNA of 20.3 ng/mL in males and 16.1 ng/mL in females, both above 10. It was recorded as counter-evidence rather than a refutation, because qPCR on plasma and post-extraction yield may differ by method. No scan returned a supporting measurement.
• Who depended on it: A46 Published measurements of total cell-free…, with Operations (C12) behind it. It feeds the plasma-volume line of A45 A bottom-up cost-of-goods build-up for one…'s build-up.
• What it costs: little, and in the good direction. The row's polarity is inverted. If the claim is false, the plasma volume per panel roughly halves (the Verifier's own figure is ~0.5-1.9 mL per replicate at 16-20 ng/mL) and Operations' named cost driver shrinks. Operations said so when it bound the condition ("If false... this burden shrinks and I would say so").
• Outcome: absorb
• Because: A46 Published measurements of total cell-free… is the only dependant, and what A46 Published measurements of total cell-free… feeds survives either answer. The build-up commissioned in IR-29 claims 140 (REQ-0083 Produce the cost-of-goods build-up your…) has to cost plasma against the actual pool's measured cfDNA concentration, which the unit definition in REQ-0075 What is one unit of this panel must state anyway. So nothing in this loop waits on the general literature proposition. If it is true, the cost lands in the build-up. If it is false, the build-up is cheaper. Commissioning a literature review to settle a number that the build-up will measure directly would buy nothing.

• The statement: claim 144unconfirmed: the CAP MOL checklist requires third-party traceable material for LoD. claim 145unconfirmed, claim 146unconfirmed, claim 147unconfirmed: claim-element readings of US11215622B2, US9499870B2 and US10577655B2. claim 148unconfirmed: donor-plasma supply and consent terms permit resale and genetic characterization.
• What the check found: every claim is a read of a document no connector reaches. The CAP checklist is licensed. The patent connector returns title and dates only; the kind-coded IDs returned no record and the bare numbers resolved to bibliographic data. Supplier terms are contracts. claim 146unconfirmed's wording and claim 105unconfirmed's cannot both be accurate transcriptions of the same claim 1unconfirmed, which the Verifier flagged.
• Who depended on it: A49 The current CAP Molecular Pathology… (Legal C2), A50 Every independent claim of US11215622B2-A52 No independent claim of US10577655B2 (Legal C4-C6), A53 The pooled human donor plasma used as this… (Legal C7). A52 No independent claim of US10577655B2 is the one Legal names as able to move it to blocks on any MNase-nucleosomal spike route.
• What it costs: A49 The current CAP Molecular Pathology… decides whether the documentary moat is a requirement or a convenience. A50 Every independent claim of US11215622B2-A52 No independent claim of US10577655B2 decide which spike routes are free to build and which need a licence. A53 The pooled human donor plasma used as this… decides whether the matrix can be sold and characterized.
• Outcome: commission
• Because: every one already has a vehicle whose question carries it. REQ-0028 Read ISO 17034 (open) carries claim 144unconfirmed; its residue is exactly the current MOL text. REQ-0077 Attach from USPTO full text (open) carries claims 145-147 and the 105/146 conflict. REQ-0078 Name the pooled human donor plasma supplier (open) carries claim 148unconfirmed. No new row is raised. The one to watch is REQ-0077 Attach from USPTO full text: per Legal, the granted claims of '655 are the single document that could reopen a block on this file.

• The statement:
  • claim 149unconfirmed: $1,500-$3,000 net per panel and $400-$900 per control vial.
  • claim 150unconfirmed: 2-400 panels/yr.
  • claim 151unconfirmed: 500-5,000 control vials/yr for run-by-run monitoring.
  • claim 152unconfirmed: SeraCare's direct price is within 25% of $1,919.59.
  • claim 153unconfirmed: developers pay ≥25% more for native plasma with stated uncertainty.
  • claim 154unconfirmed: ≥10 US MRD developers do not report patient results from a CLIA lab.
• What the check found: adoption and market returned unverifiable by design on 149-151 and 153, and the Verifier substituted no estimate. The arithmetic of 150 and 151 was checked and is consistent (2.4-360 and 750-4,500), but every input is unconfirmed or unsourced. claim 152unconfirmed needs a quote, and its comparators (claims 117, 123, 124) are unconfirmed web prices. claim 154unconfirmed needs an enumeration no connector performs.
• Who depended on it: A54 US tumor-informed ctDNA MRD assay…-A59 At least 10 US organisations currently…, with Commercial (C11-C16) behind them. Through A54 US tumor-informed ctDNA MRD assay…/A55 Annual US demand for this candidate's…, Operations' build-up (A45 A bottom-up cost-of-goods build-up for one…) is tested against these. A56 If this candidate's daily-use single-level… is the line P2 clinical QC material at launch was chosen to keep.
• What it costs: the whole demand side of the business case as it now stands. Commercial moves to blocks on the pricing thesis if 153 comes back "no premium" alongside a low direct incumbent price (152). If 151 fails, IR-23 Regulatory blocks the P0 launch framing's main reason for choosing P2 clinical QC material at launch over the development-phase-only shape goes.
• Outcome: commission
• Because: permanent unverifiable from adoption/market is the protocol's cleanest commission case. The instrument that closes it is primary research, not a connector. Checked against the IR-21 claims 47 and 48 came back unconfirmed/idea-001 IR-6 claim 20 came back unconfirmed test (does an existing request's question carry the claim?):
  • REQ-0038 Run a short structured survey of 10-15 MRD…'s question asks about a premium "for a panel characterized into the sub-0.01% VAF range". Since P1 derived-value bottom levels that is not what a premium would be for, and it asks nothing about runs per week or control vials. It does not carry 149, 151 or 153. REQ-0086 Extend the REQ-0038 survey instrument… (front end → research) extends the instrument before it is fielded, with the paired price question, net price per configuration, runs × vials, and panels per year, so the same fifteen people are asked once.
  • REQ-0087 Obtain a written quote from SeraCare/LGC… (front end → research) asks for the SeraCare direct quote for claim 152unconfirmed.
  • claim 150unconfirmed's organisation count is REQ-0037 Obtain a licensed lab-economics report on… (open).
  • claim 154unconfirmed is REQ-0080 Enumerate US organisations currently… (open), with REQ-0079 Under the development-phase-only RUO shape… (open) deciding whether the hybrid developers count. As IR-21 claims 47 and 48 came back unconfirmed recorded, a favourable survey makes forecasts better-founded and leaves them unconfirmed. Nobody should read a clean survey as a verified market.

• The statement:
  • claim 156unconfirmed: use "during assay validation" falls within 866.5910(a)'s identification and is not a different intended use under 866.9(a).
  • claim 157unconfirmed: a native-cfDNA donor-plasma QC material has "existing or reasonably foreseeable characteristics" of 866.5910 devices under 866.9's introductory text.
• What the check found:
  • regulation 866.5910 and regulation 866.9 returned verified (eCFR issue 2026-09-22) for the text. 866.5910(a) reads "intended to help monitor reliability of a test system by detecting analytical deviations such as those that may arise from reagent or instrument variation in genetic testing". 866.9(a) requires a 510(k) where the device "is intended for a use different from the intended use of a legally marketed device in that generic type". Neither text mentions assay validation.
  • For claim 157unconfirmed, 866.5910(a) reads "This type of device includes recombinant, synthetic, and cell line-based DNA controls". The QIG, NZB and SCU definitions also name only synthetic and cell-line controls. 866.9's introduction limits the exemption to devices with "existing or reasonably foreseeable characteristics of commercially distributed devices within that generic type".
  • Both are FDA intended-use judgments that CFR text cannot settle. Neither is contradicted.
• Who depended on it: A61 Use of a quality control material 'during… (Regulatory C9), A62 A quality control material built on pooled… (Regulatory C10), and through them the P2 clinical QC material at launch launch, as with A4 A quality control material whose entire…/A5 21 CFR 866.5910's generic type. Regulatory names C9 as one of three conditions that would move it to blocks.
• What it costs:
  • If 156 fails: a 510(k) or De Novo (De Novo classification request) before first revenue on the (a) limb, or a rewording that removes the multi-level panel's labelled validation use.
  • If 157 fails: a 510(k) before revenue, or the buffer/synthetic format, which gives up the matrix differentiator.
• Outcome: commission
• Because: these are unconfirmed on interpretive questions that FDA answers on request. The vehicle exists: Regulatory's 513(g) draft (REQ-0081 Draft the 21 CFR 513, answered). Its question (d) is claim 157unconfirmed's limb, and its question (a), asked of the sentence as filed, is claim 156unconfirmed's. Filing is a human act and a spend: about $8,596 under the FY2027 fee, or $4,298 with a Small Business Determination obtained before filing, plus a 60-day statutory period. It is Tier 3 and no department here may take it. Of Regulatory's sequence, step 1 is done: the Verifier has checked claims 155-159. Step 2 is REQ-0089 Before the 513, addressed to this department, and **it is answered here: the 513(g) asks FDA about the P2 clinical QC material at launch sentence as it stands, validation clause included.** Three reasons:
  1. Rewording the sentence changes intended_use:, and only a pivot may do that. No failed statement forces one. claim 156unconfirmed is unconfirmed, and the same filing settles it at no extra fee.
  2. The variant is not a free edit. It swaps "validation" for "verification of analytical performance" and adds "not for establishing an assay's performance claims". That takes away the multi-level panel's labelled validation use, which is where Clinical places the sub-0.01% levels ("they belong to the validation panel"). It is a shape change on the panel.
  3. If FDA answers against claim 156unconfirmed, the variant is the pre-named next shape (see "Pivot options"). The cost of having asked is one re-filing, because a 513(g) answer lapses when labelling changes: one fee and about 60-90 days. One qualification, raised this turn rather than left implicit. REQ-0104 P2's intended-use sentence names… asks Regulatory whether the precision-monitoring restriction the label must carry (claim 161unconfirmed) fits the P2 clinical QC material at launch sentence. REQ-0090 Under the 866.5910 special control (open) asks whether the derived-value and LDT-site questions belong in the same filing. The filing should wait for both answers so that one filing carries everything. If REQ-0104 P2's intended-use sentence names… says the sentence must change, that change comes through this loop as a review, not as a quiet edit.

• The statement: claim 155unconfirmed: Maine Molecular lists Xpert BCR-ABL1 control panels under QIG (866.5910, Class II, exempt), with no 510(k), for a CML treatment-monitoring assay (K190076 (an FDA 510(k) clearance number)).
• What the check found:
  • QIG is class 2, "Quality Control Material, Genetics, Rna", regulation 866.5910, submission type 4 (exempt), read from the raw classification record.
  • The record's registration number 3005959679 is not resolved to a firm by any connector.
  • A full-window 510(k) scan for QIG returned zero records. This is a ScanResult, a lead rather than a tag.
  • 510k "BCR-ABL" --applicant "Maine Molecular" found no clearance, which supports the claim's negative. The firm does hold K221253 for other controls, so it does file 510(k)s.
  • The listing itself, which the claim's conclusion rests on, was not checked, because no connector reads device/registrationlisting.json. Regulatory says its own raw query returned the listing. A chair's query is not a tag.
• Who depended on it: A60 In FDA's establishment registration and…, with Regulatory (C8) behind it. What it bears on is A4 A quality control material whose entire… (claim 52unconfirmed), as precedent cited in the 513(g)'s question (b).
• What it costs: the closest analogue on file to an MRD QC material, meaning an oncology-monitoring control distributed as 866.5910-exempt. Nothing else.
• Outcome: absorb
• Because: A4 A quality control material whose entire… is settled by the 513(g) and by nothing else, whatever claim 155unconfirmed says. Regulatory's own binding says C8 "narrows the question but cannot settle it", and the draft cites the precedent either way to an agency that holds its own listings. The same reasoning as IR-27 claim 132 unconfirmed applies to the alternative. A human-attached listing would most likely be held at unconfirmed on evidence class, as claims 64-65 were, and would buy no decision. The row stays live and the chair's condition stands. **The pattern is recorded, because it is now three claims long.** Claims 132, 155 and 171, plus Commercial's REQ-0099 From the manufacturer's labelling or…, all need FDA registration and listing, and no connector reads it. That is a case for a typed connector in connectors/ on its own merits. It is recorded here for whoever owns that folder, not raised as this opportunity's commission. REQ-0100 For each FDA-registered establishment that… (IR-36 claims 163 and 171 unconfirmed) will resolve the 866.5910 registration numbers to firms for claim 171unconfirmed. It may incidentally surface this listing, and that would not change this absorb.

• The statement:
  • claim 158unconfirmed: the 866.5910 special control (the 2007 CF QC guidance) recommends matrix and extraction studies, per-variant value assignment and traceability, multi-lot stability, 3-site/3-lot reproducibility, testing on cleared assays or an accepted reference method, and Part 610 labelling.
  • claim 164unconfirmed: BLOODPAC's MRD analytical-validation consensus requires a contrived-sample functional characterization before contrived samples are used in validation.
  • claim 165unconfirmed: MolDX's TA FAQs accept purchased or contrived material for LoD and name no required source or standard.
• What the check found:
  • claim 158unconfirmed: only the title is confirmed, verbatim from 866.5910(b), and 72 FR 1176 (2007-01-10) fits the date. None of the listed recommendations was checked, because FDA guidance is outside every connector and the Verifier did not substitute a fetch no connector points to.
  • claim 164unconfirmed: the paper exists as described (PMID (PubMed identifier) 41533993, JCO Precision Oncology 2026, via literature-detail). The abstract says nothing about contrived samples or a CSFC, and full text is out of reach.
  • claim 165unconfirmed: no connector reads MAC (Medicare administrative contractor) programme documentation. The claim is also a negative about a document's contents. Market Access's own fetch of the FAQs PDF returned HTTP 503 on 2026-09-26.
• Who depended on it:
  • A63 FDA's special control for 21 CFR 866.5910: Regulatory C11. Operations' P2 clinical QC material at launch serial calendar (its REQ-0070 If idea-003 launches as a clinical… answer and REQ-0096 Under 21 CFR 866.5910 as a 510), Clinical's reproducibility and matrix-effect design (REQ-0088 The 866.5910 special control answer), and Technology's per-variant traceability all read it.
  • A69 The BLOODPAC MRD Analytical Validation…: Clinical C11, and through it A8 Commutability against clinical patient….
  • A70 Palmetto GBA's MolDX Technical Assessment…: Market Access C7, and through it A42 MolDX LCD L38779 conditions coverage of a….
• What it costs:
  • claim 158unconfirmed is the study list every chair's P2 clinical QC material at launch cost and calendar is built on, and nobody has read it from the document.
  • claim 164unconfirmed decides whether Clinical's commutability study is a precondition of the validation use or optional support.
  • claim 165unconfirmed decides whether the payer-side driver is permissive or mandatory, which sets the sign of P2 clinical QC material at launch's demand case in Market Access's field.
• Outcome: commission
• Because: the cleanest unconfirmed there is: three free documents nobody has read. REQ-0024 Does FDA guidance "Distribution of In Vitro… is the precedent. It retrieved the 2013 RUO guidance, and the Verifier confirmed the quoted content (claims 64-65).
  • REQ-0102 Retrieve FDA's 'Class II Special Controls… (front end → research) retrieves the 2007 guidance and quotes each recommendation verbatim. That includes any statement on whether the studies must precede distribution, which is the text Regulatory needs to answer REQ-0096 Under 21 CFR 866.5910 as a 510.
  • REQ-0103 Retrieve the full text (front end → research) retrieves the BLOODPAC full text.
  • REQ-0093 Retrieve Palmetto GBA's MolDX Technical… (market-access → research, open) already carries claim 165unconfirmed. No new row for it. Said plainly, as IR-23 Regulatory blocks the P0 launch framing said of claims 64-65: a retrieved and quoted guidance may still be held at unconfirmed on evidence class. Retrieval is worth it anyway. What the chairs' calendars need is the content, and a quotation the Verifier has confirmed is what they can be built on. These are the cheapest open items on the file: minutes of reading each.

• The statement:
  • claim 163unconfirmed: as of 2026-09-26 no tumor-informed ctDNA MRD assay for solid tumors holds 510(k), De Novo or PMA (premarket approval) authorization.
  • claim 171unconfirmed: at least one establishment listing under 866.5910 offers contract manufacture of nucleic-acid QC materials.
• What the check found:
  • claim 163unconfirmed, 510(k): openFDA device-name searches for "Signatera", "residual disease" and "circulating tumor DNA" returned nothing. "RaDaR" matched only an unrelated vascular device (K142122).
  • claim 163unconfirmed, De Novo: the cached listing (494 records) has no match.
  • claim 163unconfirmed, PMA: no PMA connector exists, so that limb was not checked at all. Name searches cannot prove absence, and Clinical notes that this pass's name-only searches missed clonoSEQ.
  • claim 171unconfirmed: the classification records carry seven registration numbers across SCU, NZB and QIG. No connector resolves them to firms, and a contract-manufacture offer is a commercial fact no primary source here answers.
• Who depended on it:
  • A68 As of 2026-09-26 no tumor-informed…: Clinical C10 (Market Access reads it).
  • A76 At least one establishment registered with…: Operations C15.
• What it costs:
  • claim 163unconfirmed decides the special control's reproducibility design: whether it can run on an authorized assay, or must rest on LDT sites plus a reference method (REQ-0090 Under the 866.5910 special control).
  • claim 171unconfirmed decides whether the longest fixed item P2 clinical QC material at launch adds, the QMS and design history file, can be rented. It is Operations' stated route away from blocks on claim 50unconfirmed.
• Outcome: commission
• Because: gaps in the check, both settleable by a desk search a human can attach. REQ-0101 Search FDA's PMA (front end → research) searches the three FDA databases by applicant and product code rather than device-name substring, and reports the query set if the answer is none. Whether a documented negative search is enough to tag a negative is the Verifier's call. REQ-0100 For each FDA-registered establishment that… (front end → research) resolves the 866.5910 registration numbers to firms and attaches any published contract-manufacturing offer. The missing PMA endpoint is the same kind of connector gap as IR-34 claim 155 unconfirmed's, and is recorded the same way.

• The statement:
  • claim 166unconfirmed: daily control at 0.03% VAF, ≥30 ng per vial, one gravimetric step from the measured 0.1% level, with a dPCR cross-check agreeing.
  • claim 167unconfirmed: per-pool duplex background at 3 µg is ≤1e-5 per position and ≤1/10 of the floor in aggregate.
  • claim 168unconfirmed: a paired tumor/normal line gives ≥1,200 (≥150 at 0.01%) certified clonal variants that are stable over three lots.
  • claim 169unconfirmed: aggregate error-corrected sequencing agrees with the gravimetric value at 0.01% (and at 0.001% if that level is kept).
  • claim 170unconfirmed: the daily control is within 20% of nominal across three independently sourced lots.
• What the check found: every Verifier note says the same thing. Each is a design or performance condition on a product that has not been built, and no primary source anywhere can answer it. It resolves only when bench data is produced and attached. One arithmetic flag: claim 166unconfirmed's "at least 9,100 GE" rounds the wrong way (30 ng / 3.3 pg = 9,091 GE), so a vial filled at exactly 30 ng fails the claim's own bound. Nothing was found against any of them, and nothing could have been.
• Who depended on it:
  • A71 idea-003's daily-use single-level control…: Technology C9; Commercial's C17 prices it; Clinical's C7/C8 are computed at it.
  • A72 In the unspiked pooled-donor-plasma lot…: Technology C10; Operations C16/C17.
  • A73 idea-003's variant-bearing spike is…: Technology C11; Operations C8; Legal C8.
  • A74 For idea-003's pilot lot: Technology C12; the reference-method question in REQ-0090 Under the 866.5910 special control.
  • A75 Across three consecutive production lots of…: Technology C13; Operations C1/C9; Commercial C17.
• What it costs:
  • 166: the product carrying P2 clinical QC material at launch's revenue line has no confirmed value at its level. The fallback is a measured 0.1% control.
  • 167: certified positions drop out pool by pool.
  • 168: the floor rises.
  • 169: the derived levels have no independent confirmation, and this is Technology's go/no-go for 0.001%.
  • 170: a lab's QC range does not survive a lot change.
• Outcome: commission
• Because: this is IR-25 P1's pilot-lot conditions's shape, and IR-25 P1's pilot-lot conditions's reasoning holds. What has changed since IR-25 P1's pilot-lot conditions is that the programme exists as a specification with a pre-specified pass/fail per claim. REQ-0082 Specify the single pilot-lot bench… was answered in Technology's position: WP0-WP5, about USD 0.5-0.9M and 9-12 months on Technology's judgement, excluding WP4, equipment and the QMS.
  • REQ-0095 Price, as a cost and calendar with named… (accepted by Operations) prices it.
  • REQ-0094 For published analytical validations of… (open) checks the yield and tracked-variant inputs behind claim 166unconfirmed's after-loss arithmetic.
  • REQ-0106 The Verifier flagged claim 166 (front end → technology) asks Technology to restate claim 166unconfirmed's genome-equivalent bound so that a vial built to specification can satisfy it. The front end does not edit a chair's claim. A chair that is wrong in its own field gets a request.
  • claim 170unconfirmed settles only on three production lots after design freeze. Its commission is the three-lot line in REQ-0083 Produce the cost-of-goods build-up your…'s build-up and the sequencing question in REQ-0096 Under 21 CFR 866.5910 as a 510.
  • WP4 (claims 31 and 63), which the pilot excludes, is commissioned as REQ-0109 Source and cost the commutability and… (front end → clinical). Stated as plainly as IR-25 P1's pilot-lot conditions did: no request settles claims 166-170. Running the pilot lot is a human capital decision. What the ledger holds is everything that decision needs. Operations adds one thing the funder must know. WP0 and WP1 cannot start on sellable material until the spike-line licence (REQ-0085 Name the variant-bearing source material…, REQ-0108 Extend REQ-0085's licence enquiry for the…) and a commercial-terms plasma pool (REQ-0078 Name the pooled human donor plasma supplier, REQ-0107 In the same custom-pool procurement…) exist, and neither lead time is inside Technology's band.

• The statement:
  • claim 160unconfirmed: the daily control's level gives ≥3 expected tumor molecules per extraction over the customer's tracked variants, and the label states the count and P(0).
  • claim 161unconfirmed: the label claims precision monitoring only where N ≥ 100, and qualitative detection below that.
  • claim 162unconfirmed: each lot ships with tumor-proxy and normal-proxy DNA and a certified variant list, and at least one external MRD lab has designed a panel against it.
• What the check found: these are design conditions on an unbuilt product, and claim 162unconfirmed's second half is a future event. The arithmetic premises were checked and are correct. N ≥ 3 gives P(≥1) = 1 − e^−3 = 95.0%, and a Poisson count of 100 has a CV of 10%.
• Who depended on it:
  • A65 idea-003's daily-use single-level control…: Clinical C7; Technology; Commercial.
  • A66 idea-003's labelling claims monitoring of…: Clinical C8; Commercial.
  • A67 Each lot of idea-003 is supplied with a…: Clinical C9; Technology; Legal C8.
• What it costs:
  • 160: a control that misses more than one in-control run in twenty, which labs learn to ignore.
  • 161: a Class II label that either overclaims precision or diverges from the sentence FDA classifies.
  • 162: a QC material that cannot be run on the only assay class it is sold to control.
• Outcome: commission
• Because: these are gaps, and they resolve by class-1 arithmetic once a specification and a draft label exist. Half of that now exists. Technology specified the level (0.03%, ≥30 ng) and computed N: claim 160unconfirmed holds for ≥16 tracked variants even at 10% yield, and claim 161unconfirmed holds at the daily level only for high-plex assays. What is still open:
  • the yield and variant-count inputs: REQ-0094 For published analytical validations of… (open);
  • whether the restriction fits the P2 clinical QC material at launch sentence: REQ-0104 P2's intended-use sentence names… (front end → regulatory), which must be answered before the 513(g) is filed (IR-33 claims 156 and 157 unconfirmed);
  • claim 162unconfirmed's legal half: REQ-0108 Extend REQ-0085's licence enquiry for the… (IR-39 claims 172 and 176 unconfirmed, claim 176unconfirmed);
  • claim 162unconfirmed's external-lab half, which needs a willing lab: REQ-0105 Before the REQ-0038/REQ-0086/REQ-0098… (front end → research) adds the question to the survey instrument. That also recruits the ≥2 external reproducibility sites the special control recommends, which Clinical flagged nobody had asked for. A survey names a candidate laboratory. It does not settle claim 162unconfirmed, which needs a panel design actually done against a lot. The draft label itself is a design-controls output under part 820 (A64 Product codes SCU), inside the fixed layer Operations itemized, and gets no separate row.

• The statement:
  • claim 172unconfirmed: a US supplier will sell a single custom pooled lot of ≥10 L, tested per 21 CFR 610.40, with donor-level records.
  • claim 176unconfirmed: the tumor/normal line pair's licence and donor consent permit redistributing extracted DNA to customers and disclosing the certified variant list.
• What the check found: claim 172unconfirmed is a supplier-capability fact that needs a quote. 610.40 is real, but it was not re-read because it is not what the claim turns on. It carries the cross-reference flag recorded on A77 At least one US supplier of pooled human…. claim 176unconfirmed concerns lines not yet selected, and it needs the licence, MTA and consent documents. No connector reads either.
• Who depended on it:
  • A77 At least one US supplier of pooled human…: Operations C16.
  • A81 The tumor cell line and matched normal cell…: Legal C8, and through it A67 Each lot of idea-003 is supplied with a… (Clinical C9).
• What it costs:
  • 172: about ten pools a year, each with its own background map. That multiplies the largest fixed cost.
  • 176: the tumor-proxy/normal-proxy kit cannot lawfully ship, and without it the product cannot be run.
• Outcome: commission
• Because: neither question is carried by an existing row. That is the test IR-21 claims 47 and 48 came back unconfirmed and IR-32 claims 149-154 unconfirmed applied. REQ-0078 Name the pooled human donor plasma supplier asks about consent and resale terms, not lot size. REQ-0085 Name the variant-bearing source material… asks about manufacturing use and deep sequencing, not redistribution or disclosure, and Legal said so when binding C8. REQ-0107 In the same custom-pool procurement… and REQ-0108 Extend REQ-0085's licence enquiry for the… (front end → operations-npi) ride on the same two procurement conversations, so each supplier is asked once. Both are human acts, and Operations will probably accept and wait, as it did with REQ-0078 Name the pooled human donor plasma supplier and REQ-0085 Name the variant-bearing source material…. That is the correct state for them.

• The statement:
  • claim 174unconfirmed: no live incumbent patent claims the fragmentomic plasma material, the dilution-chain value assignment, or run-by-run monitoring with a spiked QC material.
  • claim 175unconfirmed: the dilution-chain combination is not disclosed in any single prior-art document.
• What the check found: both are negatives that depend on searches nobody has run. The patent connector checks one named patent's bibliographic dates, never an absence and never an FTO conclusion. No generic search was substituted.
• Who depended on it:
  • A79 No live: Legal C1; Commercial through REQ-0031 Does a freedom-to-operate search establish….
  • A80 The combination idea-003's P1 mechanism…: Legal C3; Commercial's Route D.
• What it costs:
  • 174: a licence or design-around before launch, or a P2 clinical QC material at launch label that directs customers to practise a claimed method.
  • 175: no moat in Legal's field, and nothing for Route D to sell.
• Outcome: commission
• Because: REQ-0027 Run a patent/prior-art landscape search (open since 2026-09-14) and REQ-0097 Supplement REQ-0027's landscape to the… (open) carry both questions exactly. REQ-0097 Supplement REQ-0027's landscape to the… was raised by Legal to widen the landscape to the P1 derived-value bottom levels dilution chain and the P2 clinical QC material at launch monitoring use, which is why these claims were re-bound from 54 and 56. No new row is needed. The search belongs to the research lane. The conclusion belongs to counsel, who is a human, because nothing in this repo produces an FTO opinion. REQ-0031 Does a freedom-to-operate search establish… stays open for the reason Legal gave.

• The statement:
  • claim 173unconfirmed: at 5-20 ng/mL, the ≥3 µg per-pool map consumes ≥150-600 mL of plasma, and a 0.001% floor's 33-100 µg consumes ≥1.65-20 L, comparable to a year's product.
  • claim 178unconfirmed: about $875K of low-end revenue at planning volumes exceeds fully loaded COGS (cost of goods sold) plus the recurring 866.5910 QMS layer.
• What the check found:
  • claim 173unconfirmed: the arithmetic is correct (3 µg / 20 ng/mL = 150 mL … 100 µg / 5 ng/mL = 20 L). But the 5-20 ng/mL range is unsourced in the line (the concentration claim is 141; the Verifier's note names 144). The 3 µg is claim 167unconfirmed, the 33-100 µg is Technology's statement, and the 5-16 L derives from claims 150/151.
  • claim 178unconfirmed: the revenue arithmetic is correct ($75,000 + $800,000). The prices (claim 149unconfirmed) and volumes (claims 150/151) are unconfirmed. The cost side has no figure and no source, and market is unverifiable by design.
• Who depended on it:
  • A78 At healthy-donor plasma cfDNA…: Operations C17, bearing on the 0.01% floor option and on A45 A bottom-up cost-of-goods build-up for one…'s plasma line.
  • A83 At Commercial's planning volumes: Commercial C18, which moves to blocks on the P2 clinical QC material at launch revenue model if it fails; A45 A bottom-up cost-of-goods build-up for one… and A64 Product codes SCU feed it.
• What it costs: claim 178unconfirmed is the P2 clinical QC material at launch business test: IR-22 claim 49 came back unconfirmed's margin question (A31 The gross margin earned on the panel) with the QMS layer and a number in it. claim 173unconfirmed has inverted polarity. It decides how much keeping the 0.001% level costs in plasma.
• Outcome: commission
• Because: neither has an input of its own to check. Each settles when its inputs do, and every input already has a vehicle:
  • cost side: REQ-0083 Produce the cost-of-goods build-up your… (accepted; the quantity side is written, the price side waits on quotes), with REQ-0107 In the same custom-pool procurement… (pool), REQ-0085 Name the variant-bearing source material…/REQ-0108 Extend REQ-0085's licence enquiry for the… (licence) and REQ-0100 For each FDA-registered establishment that… (a contract manufacturer's rate, if one exists);
  • prices and volumes: REQ-0086 Extend the REQ-0038 survey instrument…, REQ-0098 Before the REQ-0038/REQ-0086 survey is… and REQ-0037 Obtain a licensed lab-economics report on…;
  • the 3 µg and the 0.001% depth: the pilot's WP1 (REQ-0095 Price, as a cost and calendar with named…);
  • cfDNA concentration: measured per pool (IR-30 claim 141 unconfirmed). No new row. Absorb was considered for 173 on IR-30 claim 141 unconfirmed's reasoning and not taken. claim 141unconfirmed's consequence disappeared into a build-up that measures concentration directly. claim 173unconfirmed's consequence is whether the current shape keeps its 0.001% level, which is a cost the build-up carries only if it is told the map depth. It stays commissioned through REQ-0083 Produce the cost-of-goods build-up your…, whose quantity side is Operations' C17.

• The statement: claim 177unconfirmed: at $400-$900 per vial, at least one US MRD lab in ten would use idea-003's 0.03% daily control as its run-by-run control, instead of in-house material or the Seraseq 0.05% vial diluted into its own plasma at about $80 a run.
• What the check found: adoption returned unverifiable by design, and the Verifier supplied no synthetic stakeholder view. The cross-references were checked and are consistent: claim 111unconfirmed is the buffer-format mix, 117 its price, and 166 the 0.03% specification.
• Who depended on it: A82 At a net price of $400-$900 per single-use…, with Commercial (C17) behind it. Through it, A56 If this candidate's daily-use single-level… (claim 151unconfirmed's capture-share floor), which was IR-23 Regulatory blocks the P0 launch framing's main reason for choosing P2 clinical QC material at launch.
• What it costs: below one lab in ten at $400, the line P2 clinical QC material at launch was chosen to keep fails on differentiation even though the regulatory objection is gone. Commercial moves to blocks on the P2 clinical QC material at launch revenue model, and the development-phase-only option's first trigger fires.
• Outcome: commission
• Because: permanently unverifiable from adoption is the protocol's cleanest commission case (IR-21 claims 47 and 48 came back unconfirmed, IR-32 claims 149-154 unconfirmed). REQ-0098 Before the REQ-0038/REQ-0086 survey is… (commercial → research, open) puts the switching question at three price points into the survey instrument. REQ-0099 From the manufacturer's labelling or… (open) checks whether the incumbent lacks the listed-QC attribute, one of the four things the price gap would be paid for. No new row. As IR-21 claims 47 and 48 came back unconfirmed said: a favourable survey makes the forecast better-founded and leaves claim 177unconfirmed unconfirmed. A practical note for whoever fields it: one instrument of about fifteen respondents now carries REQ-0038 Run a short structured survey of 10-15 MRD…, REQ-0086 Extend the REQ-0038 survey instrument…, REQ-0098 Before the REQ-0038/REQ-0086 survey is… and REQ-0105 Before the REQ-0038/REQ-0086/REQ-0098…. It should be fielded once and should record which question each claim hangs on. Otherwise a long survey will come back with answers nobody can map to a claim.