fox files

One-third of the so-called SARS-CoV-2 spike protein was assembled from human genetic material, not viral RNA—raising urgent questions about the origins of the sequence used in mRNA vaccines.

3D print of a spike protein on the surface of SARS-CoV-2—also known as 2019-nCoV, the virus that causes COVID-19. Spike proteins cover the surface of SARS-CoV-2 and enable the virus to enter and infect human cells. For more information, visit the NIH 3D Print Exchange at 3dprint.nih.gov. Credit: NIH/Wikimedia Commons under the Creative Commons Attribution 2.0 Generic license. Image saturation, vibrace, color, and temperature have been altered in Canva Pro.

The most critical sequence in modern medical history—the “spike protein” of SARS-CoV-2—may never have existed in nature at all.

As a new research paper explains, A 32% Human-Derived Mosaic in the In Silico-Assembled SARS-CoV-2 Spike Protein: Accidental Contaminant Misincorporation or Intentional Functional Chimeric Design?, the Wuhan “spike” was never isolated from a virus.

It was digitally stitched together—in silico (in a computer)—from fragments of RNA found in the lung fluid of one patient in China by Chinese researchers in early January 2020.

The genetic information was rapidly disseminated through databases such as GenBank and GISAID.

The foundational publication by Wu et al. in Nature in February 2020 represented the first peer-reviewed article presenting the full genome sequence of the novel coronavirus (SARS-CoV-2), including its spike protein sequence.

That synthetic model then became the blueprint for the Pfizer and Moderna mRNA vaccines injected into over five billion people worldwide.

This raises a critical question: if the Wuhan team’s “virus” was assembled from a sick man’s lung fluid, why does its defining spike protein contain extensive human genetic material—was this simply contamination from the patient’s own RNA, or evidence that the sequence was artificially constructed using human genes?

To identify the human components of that digital construct, I used the NCBI BLASTp tool—a government-hosted bioinformatics search engine that compares protein sequences against the entire global database of known organisms—to systematically test the Wuhan spike protein for matches to human proteins, revealing extensive alignments to human endogenous retroviruses and cellular genes that are absent in any bat or pangolin coronavirus.

The full research article, along with all six NCBI BLASTp run raw data files and their reproducible Request IDs, has been publicly archived on Zenodo (DOI 10.5281/zenodo.17583428), ensuring complete transparency and independent verification of every alignment reported.

You can also read and download the research article below:

	A 32% Human Derived Mosaic In The In Sil…724KB ∙ PDF file
Download

---

Key Findings

• 32% of the spike protein—416 amino acids—matches human genetic material.
  The overlaps include sequences from human endogenous retroviruses (HERV-K, HERV-H, HERV-W) and cellular proteins linked to immune modulation, fusion, and intracellular trafficking.
• No comparable overlaps exist in bat or pangolin coronaviruses.
  These human alignments appear only in the SARS-CoV-2 spike, not in its supposed animal precursors.
• Six independent NCBI BLASTp runs confirm the findings.
  Each run produced reproducible, statistically significant human alignments—with probabilities of random occurrence as low as one in 10²⁰.
• Critical overlaps occur in known functional domains:
  • HERV-K envelope homology in the S2 fusion region, which controls cell-to-cell syncytia.
  • HERV-H alignment at the furin cleavage site, already linked to a patented human gene (MSH3).
  • HERV-W (MSRV) match in the N-terminal domain, associated with neuroinflammation.
  • Additional matches with human lysosomal, mitochondrial, and zinc-finger proteins that govern energy metabolism and DNA regulation.

Why It Matters

If one-third of the spike’s code came from human sources, two explanations remain:

1. Accidental misassembly—contamination from human RNA in the original Wuhan sample (BALF), which was never purified before computational assembly; or
2. Intentional inclusion—deliberate use of human sequences to enhance infectivity, persistence, or immune modulation.

Both possibilities challenge the official story that the SARS-CoV-2 genome was a “naturally emerging” virus.

Independent Validation

Multiple clinical studies now confirm that these same HERV sequences are biologically active in COVID-19 patients:

• Petrone et al., 2023: HERV-K and HERV-W upregulated in nasal mucosa; expression levels predict hospitalization.
• Temerozo et al., 2022: HERV-K found in lung aspirates of deceased ICU patients.
• Guo et al., 2022: HERV activation triggers interferon and inflammation via cGAS-STING.
• Balestrieri et al., 2023: HERV-W linked to pediatric MIS-C and Kawasaki-like syndromes.
• Wang et al., 2023: HERV-K expression correlates with pulmonary hypertension.
• Wu et al., 2025: SARS-CoV-2 directly transactivates HERV-K, producing retrovirus-like particles that drive senescence and neurodegeneration.

Together, these findings corroborate that the “human” portions of the spike aren’t computational noise—they’re functionally active biological components.

Bottom Line

The official reference spike (YP_009724390.1)—used worldwide in vaccine development—is a digital, computationally assembled sequence derived from patient RNA rather than from a purified viral protein.

This sequence contains hundreds of human gene fragments precisely placed within key functional domains impacting viral fusion, immune evasion, and inflammation pathways.

While standard re-assembly of the original raw sequencing data (SRR10971381) excluding human reads has been performed, no published study has conducted an independent, viral-only de novo assembly focusing specifically on the spike region to test for potential human contaminant incorporation.

Thus, the origin of this 32% human mosaic—whether due to accidental laboratory contamination or intentional chimeric engineering—remains unresolved and requires targeted re-analysis.