Back to the directory

Diagnostics & tools · Epigenetics

Epinoma

A four-person startup selling EpiAmp, a chemistry that copies DNA while keeping its methylation marks. Despite a tagline about engineering proteins that control biology, what it ships reads the epigenome rather than rewrites it.

StartedUC San Diego student project, 2018 · YC Summer 2021
BasedSan Carlos, California
ProductEpiAmp methyl-preserving amplification
Core licenseUniversity of Chicago, Chuan He laboratory
Reported fundingAbout $5 million seed, as of Sep 2025
CEOVarun Govil, co-founder

The copy that forgets

Run a polymerase chain reaction and you get millions of copies of a stretch of DNA. What you do not get is the methylation. A polymerase reads A, T, C and G and writes A, T, C and G. The methyl groups perched on cytosines, the marks that tell a cell which genes to keep quiet and that betray a tumor months before an image does, stay behind on the original strand.

Biology solved this long ago. Each time a human cell divides, an enzyme called DNMT1 finds the half-methylated DNA that replication leaves behind and restores the missing marks on the new strand. In a test tube, nobody could make it work at useful scale. “This has been sought after for decades,” Chris Mason, a sequencing specialist at Weill Cornell Medicine, told GenomeWeb in September 2025.

Epinoma, four people in San Carlos, California, says it now can. Its product is EpiAmp, and the pitch is narrow: amplify the sample first, marks and all, then run whatever methylation assay you planned to run.

Scarcity, not sensitivity

Methylation is a strong cancer signal. The obstacle is arithmetic. Ten nanograms of cell-free DNA from a blood draw holds roughly 3,000 haploid genome equivalents, and the tumor fraction inside it can be a few dozen parts per million. Bisulfite conversion, the long-running way to read methylation on a sequencer, then destroys a large share of that input. Enzymatic chemistries such as EM-seq and TAPS are gentler, but none create more molecules, and sequencing deeper only resamples the same pool.

The cost shows up at the early end. In an August 2026 analysis of Epinoma data, the genomics analyst Albert Vilella noted that colorectal adenomas have been measured at tumor fractions around 60 parts per million, and that blood tests struggle there: Guardant Health’s SHIELD detects 13 percent of advanced precancerous lesions and Freenome sits at 18 percent.

Two undergraduates and a competition

In 2018 an eleven-person team from UC San Diego entered a liquid biopsy concept named Epinoma in the International Genetically Engineered Machine competition and finished second out of 250 teams. The project lead was Varun Govil, then a third year bioengineering undergraduate. A teammate, chemistry and biology double major Zhijian Li, described the mechanism to the university news office: “We found that methyl-binding domain proteins (MBD) can specifically bind to DNA.”

In February 2020 the Regents of the University of California filed a patent application on that workflow, naming Govil, Li, Ruiyuan Zhang and Ishan Goyal. Govil then took the joint UC Berkeley and UCSF Master of Translational Medicine. “I’m currently the CEO of Epinoma, a company which started off as an undergrad thesis,” he told the program in November 2021. Epinoma joined Y Combinator’s Summer 2021 batch, and that October the Atlanta fund SeedtoB Capital announced an investment, naming the Harvard geneticist George Church among early backers.

The pivot that mattered

In 2022, per GenomeWeb, the team came across a 2020 paper in the Journal of the American Chemical Society from Chuan He’s laboratory at the University of Chicago. Chengqi Liu and colleagues reported 5mC-specific whole-genome amplification, retaining methylation from inputs as small as 10 picograms. The mechanism is the one cells use, slowed on purpose. “The whole idea is to provide a kinetic window so that DNMT1 has a chance to work on a newly synthesized strand,” He told GenomeWeb.

The university had listed the method for licensing. Epinoma took an exclusive license to research and diagnostic applications through the Polsky Center, then spent years turning a demonstration into a kit. “It is very gratifying to see this invention from our laboratory being commercialized,” He said in July 2025. Govil put the engineering problem plainly: “But it wasn’t plug-and-play for fragmented, real-world samples. That’s where we focused.”

What is proven, and what is still claimed

Operations track. Epinoma has no FDA clearance, no registered trial, no federal award and no SEC filing. The record is a university license, a paper it did not write, partner pilots, and data reviewed by outsiders.
EvidenceWhat the record showsSource type
Foundational paperLiu et al., DNA 5-Methylcytosine-Specific Amplification and Sequencing, Journal of the American Chemical Society, Mar 11, 2020, 142(10):4539 to 4543. A correction followed on Feb 24, 2021.Public record
The licenseExclusive license to research and diagnostic applications of the University of Chicago method, through the Polsky Center. GenomeWeb reports the team found the paper and licensed it in 2022.Independent
Epinoma publicationsNo peer reviewed paper under the company name in PubMed as of Sep 24, 2026. A whitepaper was promised in Sep 2025, a preprint for fall 2026.Not found
Analytical performancePreservation above 99 percent per cycle, de novo methylation under 0.05 percent per cycle, amplification from 1 nanogram. Limit of detection 0.1 percent at a single locus against 2 to 3 percent for current workflows.Company-stated
Outside review of dataMason reviewed a data package and found the amplification quality good. Vilella reviewed a dilution series showing 15 of 164 colorectal regions still detectable at 0.005 percent tumor fraction, where the unamplified control detected none.Independent
Partner pilotsPilots with several companies, only Vesica Health named. No pilot results published.Partner-stated
Clinical validationNone. Sensitivity figures come from spike-in mixtures and a simulation modeled on the TAPS literature, not patient samples.Not found
PatentsUS application 20200291483, filed Feb 18, 2020, published Sep 17, 2020. Applicant: the Regents of the University of California. Nothing found under Epinoma’s own name.Public record
Regulatory and federal moneyNo FDA 510(k) or PMA, no registered trial, no NIH or NSF award, no SEC filing of any kind.Not found
What it calls itselfThe Y Combinator listing reads “We engineer proteins that control the operating system of biology,” and job posts promised “a second generation of epigenetic medicines that are safer and more effective for patients.” Every shipped asset, license and pilot on the record is a diagnostic tool.Differs from the record

Read plainly: Epinoma is a diagnostics tooling company that reads methylation, not a therapeutics company that writes it. The distinction matters because the epigenetic writing field it gets grouped with plays a different game with different money. Tune Therapeutics closed more than $175 million in January 2025 to put an epigenetic silencer into hepatitis B trials, and nChroma Bio formed in December 2024 with $75 million. Epinoma has raised about $5 million and sells reagents. Its chemistry does use engineered proteins, a polymerase working alongside a methyltransferase, but they are pointed at a sequencer, not a patient.

What is real is the license, the underlying paper, and analytical data two outside specialists have looked at and not dismissed. What is missing is a publication of Epinoma’s own, a launched product with pricing, and a single clinical sample set.

What to watch

  • The preprint. A whitepaper was promised for September 2025 and a preprint for fall 2026. Neither has appeared as of September 24, 2026.
  • The early access program, said in September 2025 to be coming late that year or early in 2026. No launch announcement found.
  • A named partner moving from pilot to contract, starting with Vesica Health.
  • Clinical samples. Every sensitivity number on the record comes from spike-ins or simulation.
  • Headcount and a new round. Tracked headcount has sat at four since early 2024, and no Form D has ever been filed.
  • Whether 5-hydroxymethylcytosine gets added. The licensed method does not detect it, and rival chemistries increasingly do.

In their words

“This has been sought after for decades.”

Chris Mason, Weill Cornell Medicine, GenomeWeb, 2025 · Independent

“We know that DNA methyltransferases in the human body can faithfully transmit this epigenetic information, and do it routinely, but the in vitro solution had never really been solved.”

Chris Mason, Weill Cornell Medicine, GenomeWeb, 2025 · Independent

“The whole idea is to provide a kinetic window so that DNMT1 has a chance to work on a newly synthesized strand.”

Chuan He, University of Chicago, GenomeWeb, 2025 · Independent

“It is very gratifying to see this invention from our laboratory being commercialized.”

Chuan He, University of Chicago, Polsky Center, 2025 · University

“We can go down to 1 nanogram of starting material for certain applications.”

Varun Govil, CEO, GenomeWeb, 2025 · Company-stated

“The preliminary data from our pilot is very promising.”

Manuel Krispin, CSO, Vesica Health, GenomeWeb, 2025 · Partner-stated

“The dataset is limited because these are simulated spike-ins, not prospective clinical samples, so clinical sensitivity remains to be proven.”

Albert Vilella, Rhymes with Haystack, 2026 · Independent
Epinoma funding profile on HVCFHealthcare Venture Capital Fund

Related companies

Sources

  1. Public recordDNA 5-Methylcytosine-Specific Amplification and SequencingJ. Am. Chem. Soc. · Mar 11, 2020
  2. Public recordCorrection, 143(7):3015J. Am. Chem. Soc. · Feb 24, 2021
  3. Public recordUS patent application 20200291483USPTO · Sep 17, 2020
  4. Public recordEDGAR company and full text search, no filingsSEC · Sep 24, 2026
  5. Public recordRePORTER organization search, no awardsNIH · Sep 24, 2026
  6. IndependentEpinoma to Commercialize Methylation-Preserving Amplification ChemistryGenomeWeb, Andrew P. Han · Sep 19, 2025
  7. IndependentDNA Methylation Analysis Poised to Make StridesGenomeWeb · Dec 15, 2025
  8. UniversityThe Missing Piece: Epinoma Advances Novel Disease Detection TechPolsky Center, UChicago · Jul 29, 2025
  9. IndependentCompany profile Epinoma (EpiAmp)Rhymes with Haystack · Sep 2, 2025
  10. IndependentMethylation liquid biopsy reaches a new level of sensitivityRhymes with Haystack · Aug 17, 2026
  11. UniversityUndergraduate team wins 2nd place at global synthetic biology competitionUC San Diego · Dec 10, 2018
  12. InterviewAlumni Spotlight: Varun GovilBerkeley and UCSF MTM · Nov 22, 2021
  13. IndependentSeedtoB Capital invests in EpinomaVenture Capital Journal · Oct 22, 2021
  14. DataEpinoma headcount and investor profileCrustdata · Sep 2026
  15. IndependentEpigenetic biotech raises bumper round to the Tune of $175MFierce Biotech · Jan 13, 2025
  16. IndependentChroma Medicine, Nvelop Therapeutics Merge as nChroma BioGEN · Dec 12, 2024
  17. CompanySeedtoB Capital Invests in Innovative Epigenetics Startup EpinomaPR Newswire · Oct 21, 2021
  18. CompanyEpinoma home, technology, news and careers pagesEpinoma · Sep 24, 2026
  19. CompanyEpinoma company page and job listingY Combinator · Sep 24, 2026
  20. PartnerAssureMDx physician informationVesica Health · Sep 24, 2026

Profile researched and written by Healthcare Discovery. Last updated September 29, 2026.