topup-gaps

Orientation — step-1 top-up searcher (3 consolidator-flagged coverage gaps)

Not a graph node. Read consolidated-overview.md + early-case-epidemiology.md, market-environmental.md, wildlife-host-serology.md before searching, to avoid re-minting anything in the existing 86-node pool. Budget: ~5 source notes, spent 3:1:0 across the three gaps (below) plus one shared quantitative primary, per the priority order in the brief.

Gap 1 (PRIORITY) — base-rate/ascertainment primary: does a novel pathogen’s first-detected cluster center on a market/high-surveillance venue regardless of true origin?

Filled — 4 sources minted, best-first:

1. S-87 - Lloyd-Smith et al. 2005 — Superspreading and the effect of individual variation on disease emergence — the quantitative foundation: fits contact-tracing data from 8 diseases showing transmission is dominated by rare, highly-skewed superspreading events, i.e. a pathogen’s first visible cluster size/location is set by which contact network a chance superspreading event hits, not by where true spillover occurred.
2. S-88 - International Commission 1978 — Ebola haemorrhagic fever in Zaire, 1976 — the cleanest historical precedent for a non-market pathogen: the first internationally recognized Ebola outbreak is entirely a hospital-needle-reuse amplification chain; the report itself cannot identify the antecedent zoonotic-spillover contact from the outbreak record, and documents asymptomatic seropositive villagers with no hospital/case link (undetected endemic circulation the hospital cluster doesn’t capture).
3. S-89 - Cowling et al. 2015 — Preliminary epidemiologic assessment of the MERS-CoV outbreak in South Korea — same viral family as SARS-CoV-2: a returning traveller with presumed (never pinned-down) Middle-East camel exposure seeded a 186-case, 16-hospital nosocomial chain in Korea; the outbreak’s recorded geography is entirely Korean hospitals, wholly disconnected from the true zoonotic-exposure geography.
4. S-90 - Faria et al. 2014 — The early spread and epidemic ignition of HIV-1 in human populations — the canonical long-range example: phylogeographic dating puts HIV-1’s founding spillover at ~1920s Kinshasa, ~40 years and a different continent-region before the first clinically-recognized cluster (US cities, 1981), the latter an artifact of which population clinicians/surveillance happened to be watching.

Judgment: this is a real gap that generalizes past a single case study — no A/B/F node in the existing pool supplies the null/base-rate (the SARS-1/MERS-camel precedent nodes already in the pool, S-41/S-44/S-45, show market/farm amplification of a true zoonosis, i.e. the venue was causally on the origin pathway, which is the opposite structure from what this gap needed: venue prominent in the record but disconnected from true origin). I did not find a single dedicated meta-analysis or simulation study titled around “ascertainment bias in outbreak-origin siting” as a general phenomenon across many pathogens — the claim’s empirical support is a precedent-by-precedent case, not one primary dataset, so I minted the strongest individual precedents/primaries rather than one review.

Gap 2 — China/Chinese-language rebuttal or reanalysis responding to Bloom (2021 deleted sequences) or Nickels/the raccoon-dog dispute

Filled, partially — 1 source minted:

1. S-91 - Chinese Embassy (US) 2021 — official rebuttal of Bloom’s deleted-sequences allegation — a formal, written PRC-government position document (compiled “Malicious Slander” rebuttal list) that directly names and rebuts Bloom’s 2021 claim, restating in writing the account Vice-Minister Zeng Yixin gave orally at the 22-Jul-2021 SCIO press conference: the sequence withdrawal originated with the Wuhan University data-generators themselves after a journal (Small) editing error, not a coordinated cover-up.

This addresses the Bloom/deleted-sequences half of the gap but not the raccoon-dog/Nickels half, and it is a diplomatic rebuttal, not an academic reanalysis — record as partially structural: I could not find any Chinese-academic (as opposed to government/diplomatic) paper, preprint, or letter engaging point-by-point with Bloom’s metagenomic-association critique or with Crits-Christoph’s raccoon-dog reanalysis / the Nickels retraction request. What exists on the raccoon-dog side is (a) China CDC’s own updated ChinaXiv preprint of their original sampling paper (already the S-15→S-16 version chain in the pool — not a new rebuttal artifact, it just folds in an acknowledgment that raccoon-dog material was present but “more often in SARS-CoV-2-negative samples”), and (b) informal, non-citable statements: an anonymous Wuhan-University-researcher interview (Pekingnology, 24-Jul-2021) and unattributed emailed comments from former China CDC director George Gao to a journalist (“they should do decent science”) reported secondhand by USRTK — neither is a stable, citable primary artifact.

Gap 3 — post-2021 farmed-wildlife (raccoon dog/mink/civet) serosurvey responding to the 2022–23 market-metagenomics controversy

Structural — none found, not minted. Searched (English): “raccoon dog farm serosurvey China 2023/2024 SARS-CoV-2 antibody,” “raccoon dog farms serological survey SARS-CoV-2 China wildlife trade,” “China fur farm raccoon dog mink civet SARS-CoV-2 surveillance survey,” “wildlife farm workers raccoon dog mink SARS-CoV-2 seroprevalence China post-pandemic.” Found only: a companion-animal (dog/cat) serosurvey (unrelated species class), and multiple 2022-23 reviews (EFSA 2023, Springer Archives of Virology 2022) that explicitly flag this exact absence — recommending farmed raccoon-dog/mink/civet serosurveillance as future work and describing Southeast-Asian/southern-Chinese wildlife-trade serological sampling as “highly undersampled or underreported” as of their writing. This matches slice F’s own orientation note, which independently flagged the identical gap. Two independent searchers now confirm no such study exists as of 2026-07-23; treat as a genuine, currently-unfillable pool gap rather than a search failure.

search_scope

WebSearch (English + targeted Chinese-language queries) on: ascertainment/detection/surveillance bias in emerging-pathogen outbreak siting; historical precedent searches by name once identified (HIV/Los Angeles 1981, Ebola/Yambuku 1976, MERS/Korea 2015); Lloyd-Smith superspreading; China CDC / Chinese-language responses to Bloom and to the raccoon-dog dispute (中国疾控中心, 石正丽, 中华预防医学杂志 queries); farmed-wildlife serosurvey searches (raccoon dog/mink/civet, English only, Chinese-language variants not separately run for gap 3 given convergent structural signal from two independent slices). WebFetch used on USRTK, Pekingnology, and PRC Embassy pages to verify exact primary-vs-secondary status and pull direct quotes before minting or excluding. Citation counts and canonical bibliographic detail (author lists, venues, DOIs) pulled via the Semantic Scholar Graph API (api.semanticscholar.org/graph/v1/paper/...).

exclusions

1. China CDC’s ChinaXiv-hosted update of their own market-environmental sampling paper (posted ~March 2023, between the Research Square preprint and the Nature publication) — not minted: this is a third version-stage of the already-pooled S-15→S-16 chain (same authors, same underlying campaign), not an independent rebuttal artifact, even though it is the point where the paper’s text first acknowledges raccoon-dog material.
2. Anonymous Wuhan University researcher interview (Pekingnology/Zichen Wang, 24 Jul 2021) and George Gao’s emailed comments to USRTK — read, not minted: informal, unattributed or single-outlet-reported statements, not stable citable primary documents (no DOI/institutional page of record).
3. 22 Jul 2021 SCIO press-conference transcript (Zeng Yixin et al.) — read about via secondary sources (People’s Daily/Xinhua, PRC consulate mirrors) but the primary transcript page (scio.gov.cn) could not be fetched (TLS certificate error); used the Nov-2021 PRC Embassy document instead, which restates the same content in stable written form and explicitly cross-references Bloom by name — treated as the citable representative of this government position rather than double-minting both.
4. “Delayed correct diagnoses in emerging disease outbreaks” (Clin Microbiol Infect, 2025) — found via search for gap 1, looked promising (historical-pattern analysis of outbreak-diagnosis delay) but paywalled (403 on ScienceDirect/CMI direct); could not confirm from the abstract alone whether it is a systematic primary analysis (vs. narrative review with anecdotal historical examples) or verify it adds anything beyond the cases already covered by S-87/S-88/S-89/S-90 — excluded for lack of full-text access rather than a substantive judgment against it; worth a follow-up if step 2 has better paywall access.
5. Oh et al. 2015, J Korean Med Sci — “MERS-CoV Superspreading Event Involving 81 Persons, Korea 2015” (single-hospital ER superspreading case study, a finer-grained companion to S-89’s whole-outbreak assessment) — read, not minted for budget; S-89 already covers the same outbreak and mechanism at the whole-outbreak level.
6. Everything already covered by the three read slice notes (market-environmental, early-case-epidemiology, wildlife-host-serology) and the consolidated overview — not re-searched or re-minted; confirmed via those notes’ own “Sources by topic” and “exclusions” sections before starting new searches.

Slice shape

A narrow, three-item top-up slice rather than a full domain slice. ~15 sources read closely (mostly secondary/news coverage used to locate and verify primaries), 5 minted. Gap 1 (priority) absorbed 4 of the 5 notes and is now well-supported by precedent, though it remains a “case-by-case precedent” rather than a single meta-analytic primary — flagging for step 2/3 that these four should probably be read as a shared thematic cluster (illustrating one general phenomenon) rather than four independent lines converging on a number. Gap 2 is half-filled (Bloom side; government-only, not academic) and half-structural (raccoon-dog/Nickels side). Gap 3 is fully structural, independently confirmed by two searchers.