Fool’s Gold Standard: The Unvalidated Science of DNA
An Essay
In 2011, researchers obtained DNA evidence from a real criminal case—a gang rape prosecution in Georgia—through a Freedom of Information Act request. The original laboratory had concluded that one of the accused men “could not be excluded” as a contributor to the DNA mixture found on the victim. This conclusion was essential to the prosecution. The man was in prison.
Itiel Dror, a cognitive neuroscientist who studies bias and decision-making, and Greg Hampikian, a geneticist and innocence project advocate, sent the same DNA evidence—the same electropherograms the original analysts had examined—to 17 independent DNA examiners working in accredited governmental laboratories across North America. These weren’t amateurs. They averaged nearly nine years of experience in DNA analysis. The critical difference: they received only the DNA data itself, without knowing about the rape accusation, the cooperating witness, or the prosecution’s theory. They were asked simply to examine the DNA mixture and determine whether the suspect could be excluded.
The results: only 1 out of 17 agreed with the original laboratory’s conclusion.
Four found the result inconclusive. Twelve concluded “exclude”—the opposite of what had been presented in court.
Same evidence. Qualified experts. Working independently. And 12 out of 17 reached a different verdict than the one that helped send a man to prison.
This wasn’t an outlier. A 2013 NIST study sent DNA mixture samples to 108 accredited laboratories. For a three-person mixture, 70% said the suspect “might be in the mix.” Only 6% reached the correct conclusion. The claimed accuracy of forensic DNA testing—99.8%, one in a billion, the gold standard of evidence—collapsed to something closer to guesswork when analysts didn’t know what answer was expected.
The entire field of forensic DNA testing traces back to a single moment in September 1984. Alec Jeffreys—then just Dr. Jeffreys—was working in a Leicester laboratory with DNA samples from a technician’s family. He was studying genetic variation, not forensics. When he looked at the X-ray film from a Southern blot analysis, he saw banding patterns that appeared to combine parental patterns in the child. He later described the moment of insight as taking “about five seconds.” From one family in Leicester, he concluded that every person has a unique DNA “barcode” and that these patterns are inherited predictably. He called it “DNA fingerprinting.”
Within months, the technique was being applied to real cases. By 1987, DNA profiling was being used in American courts. By 1990, the Human Genome Project was launched. The speed of adoption was breathtaking.
But consider what was never established in those early years: large-scale blinded validation. Jeffreys observed patterns in a single family and extrapolated a universal principle. One family in Leicester is not the global population. A pattern observed under specific laboratory conditions is not proof that the pattern will hold under all conditions, with all samples, interpreted by all analysts. Proper validation would have required thousands of samples from diverse populations, analyzed by independent laboratories who didn’t know which samples came from whom.
This validation did not happen. Not before the technique was deployed in courts. Not before people were sentenced to death on DNA evidence.
The forensic establishment’s response to the Dror study and similar findings was not reform. It was resistance. Dr. Dan Krane, a geneticist who has provided expert testimony in over 100 court cases, spent years trying to understand why forensic DNA laboratories resist blind testing. Their argument, he noted, follows a predictable pattern: the work is critically important, lives are on the line, they need access to all available information—including the suspect’s DNA profile—to get the right answer.
Krane’s response: “It seems to me it’s very similar to a student telling a teacher, ‘I really want to do well on the test you’re about to give me, and I’m confident I’ll do much better if you give me the answer key before you give me the test.’”
In 2008, Krane and colleagues published a letter in the Journal of Forensic Sciences proposing “sequential unmasking”—analysts would interpret crime scene samples before seeing the suspect’s profile. This is the minimum standard any legitimate science should meet. The proposal was ignored. Forensic DNA laboratories continue operating exactly as before.
Why would laboratories resist such an obviously sensible reform?
Jamie Andrews, who has spent years conducting control experiments on PCR testing and genetic sequencing through his Virology Controls Studies Project, suggests an uncomfortable answer: because they know what would happen if they actually ran blind tests. The results wouldn’t look like 99.99% accuracy. They’d look like the Dror study. They’d look like the NIST study. They’d look like guesswork.
The question this raises is not whether a few laboratories have quality control problems. The question is whether the entire foundation of DNA science—the theory that a molecule called DNA exists in a specific double-helix structure, carries genetic information in a readable sequence, and can be reliably identified and matched—has ever been properly validated.
The answer, when you examine the evidence, is troubling.
If the underlying molecule and its interpretive framework were as robust as claimed, blind testing would converge. Seventeen qualified analysts examining the same data would reach the same conclusion, or something close to it. They didn’t. Twelve reached the opposite conclusion. This isn’t laboratory sloppiness or individual incompetence—these were accredited facilities staffed by experienced professionals. Systematic failure under controlled conditions doesn’t just indict individual laboratories. It justifies re-examining the foundational assumptions those laboratories rely on.
Related on Genetics
Support Independent Research
This work remains free because paid subscribers make it possible. If you find value here, consider joining them.
What paid subscribers get: Access to the Deep Dive Audio Library — 170+ in-depth discussions (30-50 min each) exploring the books behind these essays. New discussions added weekly. That’s 100+ hours of content for less than the price of a single audiobook.
[Upgrade to Paid – $5/month or $50/year]
Get in touch Essay ideas, stories, or expertise to share: unbekoming@outlook.com
The Observation Problem
A nucleotide—the basic unit of DNA—is approximately 340 picometers in length. This is roughly 100 times smaller than a claimed virus particle. No microscope can image it directly. The human genome supposedly stretches 2.2 meters per cell, yet after 70 years of technological advancement, no photograph exists that matches the iconic double helix of textbook illustrations.
Every biology textbook contains elegant artistic renderings of the double helix—the twisted ladder, the base pair rungs, the sugar-phosphate backbone. These images are so familiar they feel like settled fact. But none of them are photographs. They are models, interpretations, artistic reconstructions based on indirect evidence. The canonical structure has never been directly imaged in a biological sample.
The foundational evidence for DNA’s double helix structure is Photo 51, an X-ray diffraction pattern produced by Rosalind Franklin in 1952. Watson and Crick used this image to develop their model of DNA structure. The photo required 62 hours of continuous X-ray exposure to produce. The sample was saturated with water to form a gel, with specific hydration levels maintained through hydrogen gas pumping during the extended imaging process.
Here’s what’s rarely mentioned: undergraduate researchers have produced identical X-ray diffraction patterns from ballpoint pen springs after subjecting them to similar processes. The famous pattern that launched molecular biology may not be unique to DNA or biological materials at all. Any helical structure, under the right conditions, produces the same diffraction pattern. The X diffraction pattern indicates helical structure—it doesn’t identify what the helix is made of.
The 2012 attempts at “direct imaging” of DNA—presented as breakthrough visualization—produced unclear, grainy structures that bear little resemblance to the iconic double helix. The images were technically impressive but scientifically ambiguous: they showed something, but not the clean helical ladder the theory predicts. After seven decades of claiming certainty about DNA’s structure, the inability to produce clear, convincing photographs should raise serious questions about what has actually been observed.
Watson and Crick’s original paper repeatedly used words like “suggested,” “assumed,” and “believed.” They admitted their structure needed to be “checked against more exact results” and that previously published X-ray data was “insufficient for a rigorous test.” The structure was largely theoretical, based on mathematical interpretations of scattered X-ray patterns rather than direct observation. Franklin and Gosling themselves stated their X-ray data alone couldn’t prove DNA was helical.
The base pairs—adenine, thymine, guanine, cytosine—that supposedly form the rungs of the double helix were identified by Albrecht Kossel between 1885 and 1901. He isolated these components from various animal sources—ox pancreas, calf thymus, fish sperm—using different methods for different components. His work earned him the 1910 Nobel Prize. But his research papers describing the isolation methodology are not freely available, many remain untranslated from German, and there’s little evidence of his experiments being independently reproduced.
More importantly: these base pairs remained invisible under both X-ray crystallography and electron microscopy. Their existence was assumed to explain the diffraction pattern, then the pattern was used as evidence for their existence. The circularity is complete. The structure is inferred from mathematical models applied to diffraction patterns, not from direct observation of the molecule in its natural state.
The Extraction Problem
DNA was first isolated in 1869 by Friedrich Miescher, who collected white blood cells from pus on surgical bandages and subjected them to a series of chemical treatments: soaking in sodium sulfate solution, washing with hydrochloric acid, shaking in ether, treating with sodium carbonate followed by acidic solution. The resulting precipitate—which dissolved in alkali and reformed in acid—he named “nuclein.”
This method hasn’t fundamentally changed in 156 years.
Modern DNA extraction kits use the same principles with updated terminology: “lysis buffer” instead of acid, “binding buffer” instead of alkali, centrifugation instead of manual filtering. The basic approach remains mixing biological material with harsh chemicals and separating components through spinning. True isolation—the kind analytical chemists would require for characterizing a new compound—would involve physically separating a substance while observing it, demonstrating it exists in the original tissue, and showing the extraction process doesn’t create or alter what’s being studied. What biochemists call “DNA isolation” is observing chemical reactions and byproducts.
Jamie Andrews has conducted extraction experiments that challenge the standard narrative. When he performed the classic strawberry DNA extraction—the one done in high school biology classes worldwide—he found the “DNA goo” precipitates whether or not you add dish soap. He tried varying the soap concentration; the amount of precipitate remained roughly the same. The soap, supposedly essential for breaking cell membranes to release DNA, appeared irrelevant to the outcome.
He tried the extraction on protein powder. Same stringy precipitate.
His hypothesis: the material being extracted is not DNA but collagen—the most abundant protein in the body—or lignin in the case of plants. Both are soluble in the conditions used for DNA extraction. Both form similar precipitates. The fibrous protein structures in cells, called fibrils, join together to form collagen. Under extraction protocols, this protein precipitates and is labeled “DNA.”
This would explain why extraction works regardless of whether the claimed mechanisms are present. It would also explain why a substance supposedly too small to see produces visible, stringy goo that high schoolers can wrap around glass rods. As Andrews puts it: it’s like claiming to isolate a frog from pond water by adding chemicals until something precipitates, then calling the precipitate “frog.”
The Dissolution Problem
Both PCR and genetic sequencing take place in liquid solution. The reagents are powdered chemicals dissolved in water or buffer. Look at any photograph of PCR tubes before testing: clear, colorless liquid.
Andrews identifies this as a fundamental logical problem: how can you “read” a nucleotide sequence from left to right, like a physical string, when the physical string has been dissolved into solution?
The rescue device offered is that the “molecular structure remains intact” even when dissolved. The double helix persists in solution, they claim, maintaining its physical arrangement even though you can’t see it.
This claim has never been directly verified. No one has performed an A/B comparison showing that an intact physical structure exists in solution versus not. The verification methods used—gel electrophoresis, mass spectrometry—don’t image structure. They measure electrical properties.
The entire premise of PCR is that dissolved primers find dissolved templates in solution, enzymes recognize specific sequences, and nucleotides attach in the correct order to replicate a target region. All of this happens in liquid, with chemicals that have been dissolved, and is verified by methods that measure charge rather than structure.
When Andrews asked AI systems about this directly, the response was revealing: yes, the DNA is completely dissolved into solution for PCR and sequencing. The claim that molecular structure persists is theoretical—inferred from the results of tests that assume the structure persists.
The Verification Problem
Every method used to study DNA ultimately measures electrical charge.
Gel electrophoresis—the workhorse of molecular biology—is, at its core, battery terminals in jelly. DNA fragments are loaded into a gel matrix. An electrical current is applied. Fragments migrate toward the positive terminal (because DNA is negatively charged) at rates determined by their size—larger fragments experience more friction and move slower. The resulting bands are visualized with dyes and interpreted as evidence of specific DNA sequences.
Andrews has documented how a 1% change in gel concentration produces wildly different banding patterns from identical samples. The separation is based on friction through a matrix, not identification of sequence. The bands tell you something moved at a particular rate under particular conditions. They don’t tell you what that something is.
The circularity: gel electrophoresis is used to verify that DNA extraction worked. DNA extraction is verified by gel electrophoresis. The existence of DNA is assumed in interpreting the bands; the bands are used as evidence for DNA’s existence.
This pattern repeats across molecular biology. NMR spectroscopy measures how nuclei respond to magnetic fields. Mass spectrometry measures mass-to-charge ratios. Nanopore sequencing measures changes in electrical current as molecules pass through a pore. Every method detects electrical properties, then interprets those properties through the lens of DNA theory.
When Andrews tested household items with PCR—over 200 home tests on foods and other materials—he found that items high in ionic components produced positive results. PCR, he concluded, detects electrical charge, not specific genetic sequences. The primers and polymerases are supposed to be exquisitely specific, recognizing exact nucleotide sequences in a sea of other molecules. But the chemistry tells a different story.
The Taq polymerase enzyme claimed to copy DNA is supposedly not specific to the primer sequence—it has too many claimed functions to be chemically specific. It doesn’t chemically bind to primers; it merely “interacts” based on electrostatic charge. The primer-template hybrid containing the specific region to be amplified is chemically identical to any other template-template hybrid in the sample. The specificity is theoretical, inferred from the results of a process that assumes specificity.
Early 2020 COVID primer sets sent to US laboratories produced positives in negative controls using only nuclease-free water. The test found what it was looking for even when nothing was there—because the test doesn’t find sequences. It detects electrical activity and calls it genetic identification.
Andrews contracted independent laboratories to run virus isolation protocols on uninfected cell cultures. Over 150 electron microscopy images from live sessions found particles matching “SARS-CoV-2,” “Measles,” and “HIV” by exact size, shape, coating, and visible inclusions—in cultures with no virus present. The particles were cellular debris from cells starving in reduced nutrient medium. The tests that were supposed to identify specific biological entities identified whatever produced the right electrical or physical characteristics.
The Assembly Problem
The Human Genome Project was announced as complete in 2001, with a “finished” sequence declared in 2003. This became the foundation for decades of genetic research—the reference against which all other human DNA was compared. Researchers claimed to have read the book of life.
In 2023, the BBC revealed what genetics insiders already knew: no complete human genome had ever been sequenced. The 2001-2003 announcements were based on composite DNA from multiple individuals and contained 8-10% gaps. Significant portions of the genome—the parts that were hardest to sequence—remained unread. The “complete” sequence was neither complete nor from a single human.
The methodology for genome sequencing—human or viral—involves taking samples, fragmenting them into small pieces, and using computer algorithms to reassemble the pieces into a presumed whole. No genome has ever been read end-to-end like a book. The “sequence” is a computational reconstruction based on assumptions about how the pieces should fit together.
For Illumina sequencing—the dominant technology—samples are fragmented into reads of approximately 150 base pairs. These short fragments are then assembled into longer continuous sequences (contigs) using overlap-detection algorithms. The computer looks for matching segments (k-mers) at the ends of reads and stitches them together.
The problem, as Andrews documented, is mathematical certainty of fabrication. When reads don’t naturally overlap—which is inevitable in any complex sample—the algorithms force overlaps anyway. The De Bruijn graph method used by assembly programs will connect reads that share k-mer sequences even when those reads come from entirely different sources. Forcing non-overlapping reads to overlap, Andrews demonstrated through dialogue with AI systems, results in “fabricated and misaligned contigs.” The outcome is not discovered; it is manufactured.
When researchers ran de novo assemblies of SARS-CoV-2 using different software programs, the results varied dramatically. One study using 6,648 assemblies across eight different programs found that in most cases, they couldn’t replicate the “known” genome. The total fraction of the genome found varied by tenfold depending on which assembly program was used. The number of base pairs in contiguous alignment ranged from 29,000 down to 1,100. These aren’t minor technical variations; they’re fundamentally different outputs from the same input data.
The ARTIC protocol for sequencing SARS-CoV-2—the standard method used worldwide during the pandemic—contains a primer set of more than 100 different synthesized oligonucleotides covering the whole genome. These are specific sequences, deliberately added to the sample before sequencing. The sample is also amplified by PCR using specific primers before sequencing—another round of adding target sequences. Then additional P5/P7 primers are added to attach fragments to the flow cell. By the time sequencing begins, the sample has been spiked multiple times with the sequences being sought.
Andrews’ observation cuts to the heart of it: “Is it any surprise that you find the thing that you put in there?”
The sequences are not discovered. They are, to a significant degree, created by the process of looking for them. The assembly programs are designed to produce a continuous sequence even when the underlying data doesn’t support one. The primers ensure the “right” sequences are amplified. The databases against which results are compared were themselves built using the same circular methods.
This is why Signer’s DNA extraction method—which produced stable, high-quality DNA that could be stored dry—was never adopted despite apparently superior results. His methodology didn’t fit the pipeline. The commercial extraction kits, the sequencing protocols, the assembly software, and the reference databases all assume the same methodology and produce results that validate each other. A different approach might produce different results, and different results would threaten the entire edifice.
The Prediction Problem
If DNA theory were correct, genes should predict disease. This was the promise of the Human Genome Project and the billions invested in Genome-Wide Association Studies (GWAs): find the genetic variants that cause illness, then develop targeted treatments. The genome was supposed to be the “language of life”—decode it, and medicine would be transformed.
Over 700 GWA studies have been completed, covering approximately 80 different diseases including cancers, heart disease, stroke, diabetes, and mental illnesses. The results are consistent and devastating: genes contribute at most 5-10% to common disease risk. The genetic variation confidently expected by medical geneticists simply cannot be found.
Francis Collins, former director of the Human Genome Project and the National Institutes of Health, had his own genome scanned using the most advanced technology available. For most common diseases—stroke, cancer, heart disease, dementia—his risk was completely average. The outstanding finding: a 6% increase in type 2 diabetes risk compared to the population average of 23%. After decades of research and billions of dollars, genetic scanning revealed nothing about his health prospects that wasn’t already known from basic demographic statistics. Collins was so underwhelmed by these revelations that he wrote a book encouraging everyone to get scanned anyway.
The few genetic effects discovered are scattered across large numbers of genes, each with minute individual effects. Human populations contain at least 40 distinct genes associated with type 1 diabetes, 27 associated with prostate cancer, 32 associated with Crohn’s disease. Even if someone inherited every known “bad” genetic variant for a disease—statistically highly unlikely—their risk would barely differ from the population average. The hoped-for outcome—finding genes that cause individual risk to deviate significantly from average—never materialized.
Rather than accept these results, geneticists invented “missing heritability.” The genes must be hiding somewhere: in rare variants with large effects, in epigenetics, in mitochondrial DNA, in complex genetic architectures, in copy number variants. A 2009 Nature paper titled “Finding the Missing Heritability of Complex Diseases,” authored by 27 senior scientists including Collins, formalized this special pleading. The paper should be understood not as scientific contribution but as an effort to conceal the gaping hole in medical genetics.
Each proposed hiding place, when investigated, fails to contain the missing genes. Copy number variants have been “largely ruled out” by subsequent research. Rare variants with large effects would have left historical evidence—family lines devastated by specific diseases—yet no such evidence exists. Epigenetics and mitochondrial DNA would require overturning established genetic principles, not extending them. The hiding places keep moving because there’s nothing to find.
The pattern is familiar: when predictions fail, invent rescue devices to explain why the theory is still correct. The genes must exist because twin studies say they should. If GWA studies can’t find them, the methodology must be missing something.
Meanwhile, the environmental evidence is overwhelming. Populations that migrate acquire the diseases of their adopted country within a generation. Genetically unchanged populations can shift from 0% to 80% myopia prevalence in a single generation when they adopt Western lifestyles. Type 2 diabetes risk drops by 89% with moderate lifestyle changes—not smoking, maintaining healthy weight, exercising moderately, avoiding excessive fat. The Seventh Day Adventists—non-smoking, non-drinking vegetarians—live to an average of 88 years, eight years longer than typical Americans.
Richard Lewontin of Harvard argued that heritability is fundamentally meaningless as a concept—gene contributions depend on environment, environmental susceptibility depends on genes, and there can be no universal constant defining their relationship. Martin Bobrow of Cambridge called human heritability “a poisonous concept” and “almost uninterpretable.” Yet the entire case for genetic causation of common diseases rests on twin studies producing heritability estimates—studies that systematically exclude environmental variation between families, communities, and populations from their calculations.
The genes never mattered. The environment always did. But admitting this would redirect research funding, challenge pharmaceutical strategies built on genetic targets, and require politicians to confront industries that profit from making people sick.
The Paternity Problem
Forensic DNA analysis involves crime scenes, degraded samples, and mixtures—acknowledged complexities. Surely paternity testing is different: clean cheek swabs, single sources, straightforward comparison. This, at least, must be reliable.
A 2006 study published in Forensic Science International tested that assumption. Researchers examined paternity testing under “motherless” conditions—when only the child’s and alleged father’s DNA are available, without the mother’s sample for comparison. This scenario is common in real-world testing.
The finding: 95.8% of children could match with at least one unrelated man under standard testing conditions.
The implications are staggering. Paternity tests are used to establish child support obligations, determine inheritance rights, grant or deny immigration status, and shape custody decisions. Lives pivot on results that laboratories present with “99.99% probability.” But when Ge and colleagues examined false positive and false negative rates in familial relationship testing in a 2020 paper in Forensic Science International: Genetics, they found confidence intervals around paternity probabilities are far wider than commonly appreciated.
The tests work backward from an expected result. When analysts know who the alleged father is, they interpret ambiguous data in that direction. The statistical models assume the tested man is either the father or a random unrelated individual—but don’t account for the probability that another untested man could also match.
Karen Keegan needed a kidney transplant. Her family was tested to find potential donors. The DNA tests showed she wasn’t the biological mother of two of her three children—children she had given birth to. Further testing revealed some of her tissue contained two different types of DNA. The type in her blood was different from the DNA two of her children inherited. The explanation offered: chimerism. She had absorbed cells from a twin in the womb, and different parts of her body contained different DNA. She was genetically her children’s aunt.
Chimerism became a rescue device for paternity test failures. When the test produces an impossible result—a mother isn’t the mother of children she bore—the theory expands to accommodate it. The test is never wrong; biology is just more complicated than we thought.
In another case, a suspect matched a DNA database for a sexual assault. He had a watertight alibi: he was in jail at the time of the crime. The explanation: the true perpetrator was chimeric, having received a bone marrow transplant from the incarcerated man years earlier. His semen contained DNA from someone who was locked up when the crime occurred.
How often does chimerism explain away inconvenient results? How often do laboratories invoke it to explain failures? No one knows, because systematic blind validation isn’t performed. Ancestry testing companies have been fooled by dog DNA submitted as human, with multiple labs claiming the sample showed different human ethnic backgrounds. When a news channel sent human samples labeled as dog samples to pet DNA testing companies, not a single lab identified the sample as human—and more than half claimed to identify specific dog breeds.
The accuracy figures—99.99%, 99.9999%—are calculated from theoretical models and database comparisons, not from blind empirical testing. When empirical testing occurs, the results look like the Dror study: 1 in 17. When the answer key is removed, the science disappears.
How would you explain this to a 6-year-old?
Scientists say there’s a tiny string inside you called DNA. It’s like a recipe book that tells your body how to build itself. But here’s the problem:
No one has ever actually seen this string. They show you pretty pictures in books, but those are drawings, not photographs. The real thing is too small to see.
To read the recipe, they first have to mush everything up and dissolve it in liquid—like dissolving sugar in water. Then they say they can still read the string, even though it disappeared.
How do they check if they read it right? They use a machine that measures electricity. The machine beeps, and they say “that’s the recipe.” But when scientists tried reading the same recipe without being told what answer to expect, almost none of them agreed on what it said.
And when scientists said the recipes would tell us who gets sick, they were wrong. They checked thousands of recipes and found almost nothing. Instead of admitting the recipes don’t work the way they promised, they said the important parts must be hiding somewhere they haven’t looked yet.
So: they can’t see it, they dissolve it, they measure electricity and call it reading, they disagree when tested fairly, and it doesn’t predict what they said it would.
That’s DNA science.
What Does This Mean?
The forensic science educator’s classroom and the prosecutor’s courtroom operate in different universes. Students learning DNA analysis are taught about limitations, error rates, and the prosecutor’s fallacy—the confusion between “the probability this DNA came from a random person” and “the probability the defendant is innocent.” They work through exercises showing how statistical certainty evaporates with degraded samples. A mainstream educational guide from miniPCR bio warns students explicitly: given a random match probability of 1 in 5 trillion, claiming there’s “a 1 in 5 trillion chance the defendant is innocent” is an error in reasoning called the prosecutor’s fallacy.
Juries hear none of this. They hear “one in a billion” and “99.99% certain.” The nuance that educators consider essential for understanding DNA evidence disappears precisely when the stakes are highest—when someone’s freedom or life hangs in the balance.
If the foundations of DNA theory are unsound—if DNA has never been properly observed, if extraction produces artifacts rather than isolating a natural substance, if the tests measure charge rather than sequence, if genomes are computational reconstructions rather than discovered realities, if the predictions consistently fail—then the entire edifice built on that foundation is unstable.
This extends beyond forensics and paternity testing. Modern virology depends entirely on genetics. Viral genomes are sequenced using the same methodology as human genomes—fragmenting samples, computationally assembling the pieces, adding primers for the sequences being sought. The criticisms Andrews levels at genetic sequencing apply equally to viral sequencing. If genetic sequencing doesn’t find what it claims to find, neither does viral sequencing.
Andrews arrived at this investigation from virology. His Virology Controls Studies Project ran control experiments on virus isolation protocols, finding that particles matching “SARS-CoV-2,” “Measles,” and “HIV” by size, shape, and coating appeared in cell cultures with no virus present. The particles were cellular debris from cells starving in reduced nutrient medium. PCR tests detected positives in negative controls. The entire methodology was circular.
When he traced virology’s foundations, he found they rested on genetics. And when he examined genetics, he found the same methodological failures: circular reasoning, charge-based verification, computational assembly, rescue devices when predictions fail, and active resistance to blind validation. The pattern is identical. The problems are identical. The institutional responses—denial, special pleading, refusal to test blind—are identical.
Without DNA theory as currently portrayed, you lose genetics. Without genetics, modern virology collapses. Without virology, the entire justification for mass vaccination programs—built on the claim that specific viral genetic sequences cause specific diseases and that injecting other genetic sequences confers protection—crumbles.
This is why the resistance to blind testing makes sense. This is why the forensic establishment ignores proposals for sequential unmasking. This is why geneticists invent “missing heritability” rather than accept their results. This is why Alec Jeffreys’ five-second observation in a Leicester laboratory in 1984—seeing a pattern in one family’s DNA and concluding it must apply to every human on Earth—was never subjected to large-scale blind validation before being deployed in courts worldwide. The stakes are not scientific; they are institutional, financial, and ideological.
Meanwhile, people sit in prison based on DNA evidence that fails when actually tested blind. Fathers are separated from children—or bound to children not their own—based on paternity tests with undisclosed error rates. Medical resources flow to genetic research that has failed to deliver on any of its promises while environmental causes of disease go unaddressed. Politicians avoid regulating harmful industries because genetic determinism lets them blame individual predisposition rather than corporate products.
The man in the Georgia case remains in prison. The forensic establishment continues operating exactly as before. The laboratories continue refusing blind validation. The rescue devices continue multiplying.
The science was never settled. The gold standard was always fool’s gold. The foundations were never checked—and those who try to check them are ignored, dismissed, or told they don’t understand.
The tests don’t work. The theory doesn’t predict. The observations were never made. And when you ask for blind validation, the laboratories refuse.
That tells you everything you need to know.
References
Peer-Reviewed Studies
Dror, I. E., & Hampikian, G. (2011). Subjectivity and bias in forensic DNA mixture interpretation. Science & Justice, 51(4), 204–208. https://doi.org/10.1016/j.scijus.2011.08.004
Krane, D. E., Ford, S., Gilder, J. R., Inman, K., Jamieson, A., Koppl, R., … Thompson, W. C. (2008). Sequential unmasking: A means of minimizing observer effects in forensic DNA interpretation. Journal of Forensic Sciences, 53(4), 1006–1007. https://doi.org/10.1111/j.1556-4029.2008.00787.x
Poetsch, M., Lüdcke, C., Repenning, A., Fischer, L., Mályusz, V., Simeoni, E., Lignitz, E., Oehmichen, M., & von Wurmb-Schwark, N. (2006). The problem of single parent/child paternity analysis—Practical results involving 336 children and 348 unrelated men. Forensic Science International, 159(2–3), 98–103. https://doi.org/10.1016/j.forsciint.2005.07.003
Ge, J., Eisenberg, A. J., Budowle, B., & Chakraborty, R. (2020). How many familial relationship testing results could be wrong? Forensic Science International: Genetics, 46, 102257. https://doi.org/10.1016/j.fsigen.2020.102257
Manolio, T. A., Collins, F. S., et al. (2009). Finding the missing heritability of complex diseases. Nature, 461(7265), 747–753. https://doi.org/10.1038/nature08494
Institutional Reports
Butler, J. M., & Coble, M. D. (2015). NIST Scientific Foundation Reviews: DNA Mixture Interpretation. National Institute of Standards and Technology. (ASCLD Lab presentation, January 2015)
President’s Council of Advisors on Science and Technology. (2016). Forensic Science in Criminal Courts: Ensuring Scientific Validity of Feature-Comparison Methods. Executive Office of the President.
National Research Council. (1996). The Evaluation of Forensic DNA Evidence. National Academies Press. https://www.ncbi.nlm.nih.gov/books/NBK232615/
Educational Materials
miniPCR bio. (2021). Electrophoresis Forensics Lab: Wrongfully Convicted? Instructor’s and Student’s Guide (Version 1.0). miniPCR bio.
Independent Research
Andrews, J. (2024). The DNA Hoax. The Virology Controls Studies Project. https://controlstudies.substack.com/p/the-dna-hoax
Andrews, J. (2024). The DNA Hoax 2. The Virology Controls Studies Project. https://controlstudies.substack.com/p/the-dna-hoax-4e4
Andrews, J. (2024). The DNA Hoax 3. The Virology Controls Studies Project. https://controlstudies.substack.com/p/the-dna-hoax-0a2
Andrews, J. (2024). Who’s The Daddy? The Virology Controls Studies Project. https://controlstudies.substack.com/p/whos-the-daddy
Andrews, J. (2024). Dissolving (DNA) Illusions. The Virology Controls Studies Project. https://controlstudies.substack.com/p/dissolving-dna-illusions
Cowan, T. (2024). A Look at the Human Genome Project (Livestream, October 9, 2024). https://drtomcowan.com
Tam. (2024). DNA discovery, extraction and structure: A critical review. Critical Check.
Latham, J. (2011). The Great DNA Data Deficit. Independent Science News. (Republished by Unbekoming, 2025)
Historical Sources
Watson, J. D., & Crick, F. H. C. (1953). Molecular structure of nucleic acids: A structure for deoxyribose nucleic acid. Nature, 171(4356), 737–738.
Franklin, R., & Gosling, R. G. (1953). Molecular configuration in sodium thymonucleate. Nature, 171(4356), 740–741.
Jeffreys, A. J., Wilson, V., & Thein, S. L. (1985). Hypervariable “minisatellite” regions in human DNA. Nature, 314, 67–73. https://doi.org/10.1038/314067a0
Support Independent Research
This work remains free because paid subscribers make it possible. If you find value here, consider joining them.
What paid subscribers get: Access to the Deep Dive Audio Library — 170+ in-depth discussions (30-50 min each) exploring the books behind these essays. New discussions added weekly. That’s 100+ hours of content for less than the price of a single audiobook.
[Upgrade to Paid – $5/month or $50/year]
Get in touch Essay ideas, stories, or expertise to share: unbekoming@outlook.com
Bitcoin: 3Q6BK8x8zjoPaXykQggzvoJxg5FiEbkb3U
Ethereum: 0x4CB0d39d8466a34609318FC1B003B745893788b3
New Biology Clinic
For those of you looking for practitioners who actually understand terrain medicine and the principles we explore here, I want to share something valuable. Dr. Tom Cowan—whose books and podcasts have shaped much of my own thinking about health—has created the New Biology Clinic, a virtual practice staffed by wellness specialists who operate from the same foundational understanding. This isn’t about symptom suppression or the conventional model. It’s about personalized guidance rooted in how living systems actually work. The clinic offers individual and family memberships that include not just private consults, but group sessions covering movement, nutrition, breathwork, biofield tuning, and more. Everything is virtual, making it accessible wherever you are. If you’ve been searching for practitioners who won’t look at you blankly when you mention structured water or the importance of the extracellular matrix, this is worth exploring. Use discount code “Unbekoming” to get $100 off the member activation fee. You can learn more and sign up at newbiologyclinic.com
Author's Note on Comments
Marc G. Wathelet, who identifies as a molecular biologist, raises points worth engaging directly.
On NMR visualization: NMR spectroscopy measures how nuclei respond to magnetic fields. Computational models interpret those signals into structural representations. The essay's point stands: every verification method detects electrical or magnetic properties and interprets them through frameworks that assume the structure being sought. NMR produces data interpreted as DNA structure—it doesn't image the molecule directly in living systems. The 2012 direct imaging attempts, after seven decades of certainty about DNA's structure, produced images requiring significant interpretation to reconcile with the canonical double helix.
On industry scale as validation: The biotech industry's size demonstrates commercial success, not theoretical accuracy. Empirical trial-and-error produces useful outputs regardless of whether underlying frameworks accurately describe biological mechanisms. Bloodletting persisted for millennia with institutional success and patient testimonials. The relevant question is whether claimed mechanisms survive blind validation. The Dror study says no. The NIST study says no. The German paternity research—95.8% of children matching with known non-fathers under standard protocols—says no.
On PCR specificity: Marc states it "never amplifies something that is not there." Early 2020 primer sets produced positives in negative controls with nuclease-free water. Jamie Andrews' independent testing found positive results from ionic household materials. These results need accounting for.
On ancestry matches: Te Reagan found a half-sister through Ancestry. Pattern recognition can produce correlations that align with known family structures without validating the theoretical mechanism claimed. Ancestry companies have returned specific human ethnic percentages from dog DNA. Pet DNA companies have returned breed percentages from human samples. The tests find patterns. What those patterns measure is the question.
There was a link to this site in Jamie Andrews comments, which covers much of the same science, or lack thereof. It's very well written and referenced (personally I found it easier to follow that J.A.'s articles). There are only a handful of articles on the site, I would suggest checking them all.
https://criticalcheck.wordpress.com/2021/12/15/dna-discovery-extraction-and-structure-a-critical-review/