Missing the Origin: What the WHO/SAGO Report Left Out

A Forensic Dissection of the WHO SAGO Report on SARS-CoV-2

The Scientific Advisory Group for the Origins of Novel Pathogens (SAGO), under the auspices of the World Health Organization, released its final report on June 27, 2025. Framed as an "independent assessment," it purported to synthesize all available evidence concerning the origins of SARS-CoV-2. The report is structured around the WHO's own Global Framework for investigating emerging pathogens and evaluates competing hypotheses: zoonotic spillover via intermediate host, cold chain introduction through imported frozen goods, accidental laboratory escape, and deliberate laboratory manipulation followed by a containment breach. SAGO claims to find the strongest—though not conclusive—support for a zoonotic origin, citing the presence of susceptible animal species at the Huanan Seafood Market (HSM), the spatial distribution of early cases around the market, and the lack of direct genetic precursors in known laboratory constructs. The report asserts that while hypotheses involving laboratory accidents or manipulation cannot be ruled out, no credible scientific evidence currently favors them over natural origins.

That is their story. Now here is what they left out.

Missing Primary Evidence from Chinese Institutions

They omitted the forensic trail of the reference genome MN908947.1—the earliest full genome from Wuhan, uploaded on January 5, 2020. That record, originally public on GenBank, underwent silent revision. Neither the sequence history nor version control annotations explain the changes made in the crucial days surrounding the WHO's declaration of a global health emergency. More damning, the authors ignore that no raw sequencing reads (e.g., FASTQ, .ab1 chromatograms) from the Wuhan-Hu-1 isolate were ever made publicly available—violating standard practices for forensic virology. This renders the entire genomic scaffold used for vaccine design and viral comparison unverifiable.

Early Case Mapping Centers on WIV, Not the Wet Market

SAGO anchors its zoonotic conclusion on spatial analyses characterizing the Huanan Seafood Market as the “epicenter.” Yet a closer, more nuanced geographic review reveals a conflation of case proximity, unsupported by rigorous source analysis. The clustering of early COVID‑19 cases hovered in central Wuhan—an area encompassing both the market and the WIV, muddying clear attribution to either as the origin.

A kernel‑density mapping of 155 confirmed December 2019 cases demonstrates that although many residences fell within the market vicinity, the highest 1% density contour spans a larger central swath of Wuhan, not isolated to the market locus. This statistical footprint also envelopes the district where WIV resides, raising the question: did SAGO statistically privilege one proxy location over another? Another, more granular analysis of the positive test swipes at the Wuhan market found the highest concentration of positives swipes at the outskirts of the market near the public results. Without rigorous spatial regression or centroid shift analysis, attributing causation to the wet market remains an unsupported inference.

Critics note inherent sampling biases. Patients clustered around the market might reflect early testing bias—cases near the market were more likely to trigger alarms, thereby more likely to be sequenced and mapped . This introduces “proximity ascertainment bias,” yet no attempt appears in the report to correct for it. SAGO cites the WHO mission’s map layering of case addresses and environmental sampling—but neglects how differing spatial resolution or recall bias might tilt origin attribution.

Contrasting spatial models further fragment the notion of an unambiguous market epicenter. Debates over centroid vs. mode interpretations emerged: Worobey’s Science report positioned both outbreak lineages A and B around the market. Yet subsequent rejoinders by Stoyan & Chiu argued this inference lacks robustness, cautioning that statistical tests anchor on often-unjustified assumptions about case dispersion.

Crucially, none of these spatial reconstructions conduct a counterfactual proximity analysis centered on the WIV. 2025 summaries emphasize the institute’s mere physical presence, failing to overlay early clusters, which in part lie between the market and WIV assure-test.com. This omission of forensic spatial contrast suggests SAGO consciously narrowed its geographic frame to reinforce the market-centric narrative—without testing alternatives.

At minimum, the geographic overlap of early cases around both the market and WIV warrants granular, multi-centric spatial regression to assess statistical significance. Uniformly rejecting WIV’s possible centrality by invoking market proximity—while ignoring bias or counter-analyses—reflects selective inference.

Reports of Flu-Like Illness Among WIV Staff in Late 2019

SAGO’s report skirts around allegations that several Wuhan Institute of Virology (WIV) staff fell ill with flu-like symptoms in November 2019—well before the first officially confirmed COVID‑19 cases. Yet intelligence, including that reported by Reuters and U.S. outlets, indicates that three WIV researchers sought medical attention with a respiratory illness closely resembling COVID‑19 symptoms in autumn‑winter 2019.

That illness cluster matters because pneumonia-like symptoms in November 2019 predate the Wuhan Municipal Health Commission’s acknowledgment of any such cases. Reuters reports are unambiguous: the hospital visits occurred “a month before China reported the first cases of COVID‑19” reuters.com. Livescience echoes that intelligence documentation indicates symptoms “consistent with both COVID‑19 and common seasonal illness”. These illnesses fall exactly in the window when zoonotic transmission had not yet been demonstrated by any market-associated sampling.

Influenza season was peaking in Wuhan at that time, which complicates diagnostic interpretation. A U.S. Office of the Director of National Intelligence (ODNI) memo remarked that symptoms included allergies and colds, and that no hospitalizations occurred for COVID‑like illness, yet at least one lab employee was treated for a non-respiratory condition. Whether these cases represent COVID‑19 or not remains unverified—but the mere existence of symptomatic WIV staff demands rigorous forensic inquiry, not superficial dismissal.

SAGO, however, does not perform that inquiry. The SAGO report invokes these cases only to at time favor a stalemate, at other time echo Chinese denials, without querying hospital records, staff medical logs, or key metadata like onset date, symptomatology, or diagnostic testing. They rely instead on generalized statements of “no evidence” rather than actively seeking or scrutinizing it. The absence of inquiry thus becomes de facto suppression.

SAGO itself does not have subpoena power, data collection authority, or enforcement capacity to demand or retrieve protected medical records from the Wuhan Institute of Virology (WIV), the Chinese government, or any sovereign institution.

However, what SAGO could have done—and failed to do—is critically important:

1. They could have formally requested access.

There is no evidence in the final report that SAGO even submitted formal written requests through WHO diplomatic channels for:

• De-identified staff health records from WIV;

• Logs of clinic visits or internal quarantine measures from November–December 2019;

• Lists of WIV staff tested (if at all) using PCR or serology, and what the results were.

Even if those requests were denied, documentation of the attempt would demonstrate that the investigation reached for forensic standards rather than accepting denial by default.

2. They could have subpoenaed testimony through host-country collaboration.

While SAGO cannot independently issue subpoenas, the WHO could negotiate with China to secure anonymized testimony or written responses from WIV staff. These could be reviewed without violating medical privacy laws. SAGO did not pursue this route.

3. They could have constructed counterfactual timelines.

Even without direct access to WIV clinical logs, SAGO could have cross-validated local hospital admission logs in Wuhan for upticks in “unspecified viral pneumonia,” especially among research personnel. They did not request this data.

4. They could have used third-party sources.

Multiple intelligence agencies (e.g. U.S., UK, EU) have confirmed the existence of these early illness reports. SAGO ignored these, did not request declassified intelligence summaries, and failed to invite whistleblower testimony from Chinese or international researchers.

• Redacted medical records or de-identified case logs for the three affected WIV staff.

• Any PCR or serologic tests performed in late 2019.

• Documentation on staff illness patterns, symptom descriptions, and treatment details in November–December 2019.

Without those data, SAGO treats potential early cases as inconvenient rumors, not credible indicators.

In their failure to mount even the minimal epidemiological detective work, SAGO amplifies standard phrases: “no hospitalizations for COVID‑like illness,” “we are not aware of diagnostic confirmation”—but they never say they sought records from hospital systems or WIV’s infirmary logs. That void renders their spatial core (wet-market origin) conceptual, not forensic. Meanwhile, compelling intersecting clues—a cluster of symptomatic lab workers and the absence of early market cases—remain unexamined.

Absence of Forensic Genomic Tools

They failed to apply basic forensic genomic tools: no codon usage bias analysis to assess the probability of CGG-CGG codons at the PRRA insertion; no restriction enzyme site mapping to determine if standard lab cloning scars are present; no RNA secondary structure modeling, which would reveal instability profiles typical of synthetic constructs; no G-quadruplex inspection—despite these motifs being hotspots for template switching during recombination events. They did not compare SARS-CoV-2 inserts to known lab constructs beyond vague references to bat CoVs. They did not attempt to replicate the patented insert overlaps—such as the 19-nucleotide perfect match between the FCS region and a codon-optimized MSH3 sequence in a 2017 Moderna patent.

Metadata Amnesia and Submission Trace Failure

They did not trace the metadata behind sequence submissions. Who uploaded the Wuhan-Hu-1 genome? What machine was used? What protocols? What institute? Were there any intermediaries? What IP addresses were used? Which submission system? This information exists within GenBank, GISAID, and internal audit logs. SAGO did not request it, did not mention it, and did not attempt to reconstruct the timeline. In a digital age, where every submission leaves a timestamp and machine signature, this omission is intentional.

In June 2020, when several early Wuhan SARS‑CoV‑2 genome sequences were taken down from the NIH Sequence Read Archive (SRA). The NIH confirmed the withdrawal was made at the request of the investigators who originally submitted the data — Chinese scientists — and the deletion was later described as an “error” (a copy-editing issue), with the sequences restored in July 2021.

The report ignores this lead on potential cover-up and obfuscation.

Forensic Audit Denied: Deleted Records, Missing BLAST Hits

They declined to perform a forensic audit of early BLAST hits using archived versions of GenBank and mirrored databases, which would have allowed recovery of deleted alignment signals. They did not review which sequences were replaced or removed from NCBI’s SRA in the first quarter of 2020, nor who ordered their deletion. They cite no internal email from NIH, no FOIA record from EcoHealth Alliance, no record of Baric’s infectious clone work with SHC014 and WIV1 under NIH grant AI110964.

Disregard of Government Intelligence and FOIA Evidence

They ignored the U.S. intelligence community’s own divergence. The Office of the Director of National Intelligence (ODNI) has reported that at least two agencies hold moderate confidence in a lab origin. The FBI said so explicitly. SAGO summarizes this divergence, then discards it with a shrug. They ignore Fauci’s email correspondence with Andersen, Holmes, Garry, and Daszak. They do not quote the now-public line: "destroy the world with sequence data—yay or nay?"

This quote surfaced during a January 31–February 1, 2020 conference call among top virologists—Kristian Andersen, Eddie Holmes, and Michael Farzan—shortly before their influential Proximal Origin correspondence. This moment was documented in transcripts from congressional oversight, which quote Andersen asking:

“Destroy the world based on sequence data. Yay or nay?”

Rejection of DEFUSE Grant Evidence and Construct Design

They do not mention the DEFUSE grant, which proposed insertion of novel furin cleavage sites into bat coronaviruses using Baric's reverse genetics platform. That proposal was rejected by DARPA but details are now public. It maps exactly to what emerged in Wuhan. Baric and others published papers years ago describing changing sequences in the lab to make them more capable of infecting humanized animals or cell lines. SAGO says nothing.

Disqualification of Early International Positives Without Replication

They rejected early positive serology and PCR from France, Italy, and Brazil without replicating the studies using the same assays. They blame methodological differences and sample degradation—but offer no controlled counter-study. They offer no sequencing of archived samples, no metagenomic recovery, no PCR revalidation. Their rejection is procedural and rhetorical, not empirical.

Internal Dissent: Withheld Authorship as Protest

They do not fully account for the three SAGO members who refused authorship. Yungui Yang, Deputy Director of the Beijing Institute of Genomics, declined to put his name on the final document. So did Vladimir Dedkov of the Russian Federation and Sowath Ly of Cambodia. One resigned.

WHO describes this action tersely, offering no rationale for their withdrawal. The fact that multiple senior members with deep regional and technical insight chose to dissociate from the report signals significant internal resistance to its final framing—whether on epistemic realignment, political pressure, or methodological dispute. The report’s silence on their motivations removes critical transparency and obscures potential conflicts or dissent.

Their silence speaks to internal resistance.

Cold Chain Hypothesis Rejected Without Empirical Testing

Cold Chain Hypothesis Dismissed Without Experimental Rebuttal

In its final report, the Scientific Advisory Group for the Origins of Novel Pathogens (SAGO) summarily dismisses the cold chain hypothesis—the idea that SARS-CoV-2 may have been introduced into Wuhan via frozen food imports—citing only its presumed improbability. However, no empirical virological studies are referenced that would test this pathway directly. No experiments examining the stability of SARS-CoV-2 on seafood packaging at -20°C or -80°C are cited, even though prior work has shown the virus can remain viable for days on surfaces under refrigeration. Nor does the report attempt to simulate cross-border transmission using logistics datasets from high-salience-market (HSM) supply chains, which might have clarified whether such a pathway was plausible or even likely under real-world conditions. This is not merely an academic oversight—it is a fundamental failure to empirically test and falsify an origin hypothesis that Chinese authorities previously promoted and which some international experts considered technically feasible. Scientific dismissal without experimental negation is not falsification; it is convenience masquerading as consensus.

Conclusion: Literature Summary Masquerading as Investigation

What SAGO has delivered is not an independent forensic investigation. It is a literature review bounded by permissible narratives. They admit their limitations—but never interrogate why those limitations exist. This is not uncertainty. It is denial by omission.

The record must be corrected.

Call to Action: Demand for a Full Forensic Audit

We call for an immediate, open-source forensic audit of all early SARS-CoV-2 genomic data. All withdrawn SRA files, all version-controlled GenBank records, all FASTQ/FAST5/ab1 files from Wuhan-originating sequences must be restored and archived with SHA256 signatures. All early BLASTn alignments, especially those involving vectors like pShuttle-SN, must be re-run using unaltered 2020-era databases. All sequence metadata—machine ID, lab of origin, timestamp—must be made public. Sequence alignments involving the INS1378 region must be recalculated against cloned spike constructs and patent-disclosed sequences.

The origin of this virus remains unresolved not because the evidence is missing, but because the permitted inquiry has been defined in such a way as to avoid it. Science does not require permission to follow the trail. It requires resolve.

And the trail begins again where SAGO chose not to look.