It's a pitch that makes investors do a double-take. A three-person startup in San Francisco wants to ransack the genomes of ticks, intestinal worms, and other parasites—organisms that have spent millions of years perfecting the art of evading human immune defenses—and turn their molecular tricks into drugs for autoimmune disease.
Absurd? Maybe. But Ditto Biosciences isn't exactly working in isolation.
The autoimmune therapeutics market is hurtling toward what some analysts peg at $226 billion by 2035, even as Medicare's drug pricing program squeezes legacy biologics and regulators crack open pathways for experimental cell therapies. Behind the scenes, a cadre of biotech ventures is pursuing what you might call "tolerance engineering"—approaches that don't just suppress inflammation but aim to retrain the immune system itself. Some wrap antigens in nanoparticles. Others rewire T cells. Ditto's wager, laid out in a sparse company profile ahead of Y Combinator's Winter 2026 cohort, is different: mine parasite proteomes at scale, use AI to predict which molecules bind disease-relevant targets, then engineer those candidates into therapies that teach the body to tolerate itself again.
Whether that works—whether investors and eventual patients buy in—is a question mark the size of San Francisco Bay. But the science behind it is getting harder to dismiss.
---
The Numbers Behind the Noise
First, the backdrop. A nationwide study published in early 2025 in the Journal of Clinical Investigation—the first of its kind using electronic health records—estimated that roughly 15 million Americans, or about 4.6% of the population, have been diagnosed with at least one of 105 recognized autoimmune conditions. Women account for nearly two-thirds of cases. Globally, the autoimmune drug market clocked in at $72.1 billion last year and is expected to reach $92 billion by decade's end, according to Grand View Research.
But that growth isn't evenly distributed. AbbVie's twin powerhouses, Skyrizi and Rinvoq, are on track to surpass $31 billion in combined annual sales by 2027. In the final quarter of last year alone, Skyrizi brought in $3.78 billion; Rinvoq, $1.83 billion. Those figures underscore the continued dominance of IL-23 and JAK inhibitors in major autoimmune indications—psoriasis, Crohn's disease, rheumatoid arthritis.
Yet regulatory and pricing headwinds are mounting. Medicare's drug negotiation program set maximum fair prices for older biologics like Enbrel and Stelara starting this year, with the Centers for Medicare & Medicaid Services projecting $6 billion in net savings if those caps had applied to 2023 spending. The subtext for pharma: prove durable, differentiated value—or brace for pricing pressure.
That squeeze is opening cracks for new players. The FDA loosened some requirements for cell and gene therapies earlier this year, and biosimilar approvals continue to accelerate (76 and counting as of late last year). For venture investors and corporate development teams, the strategic question isn't whether the market will reward innovation. It's which innovations will deliver.
---
Evolution's Playbook
Enter the parasites—organisms that have been hacking human immune systems since long before Homo sapiens had a name for inflammation.
Ticks, for instance, produce a family of proteins called evasins: chemokine-binding molecules that neutralize multiple inflammatory signals at once, buying the tick time to feed. A 2022 study showed that structure-guided engineering could retarget evasin specificity to whichever chemokines a researcher—or drug developer—wanted to suppress. Tick salivary proteins called sialostatins, meanwhile, inhibit dendritic cell maturation and dampen T-helper responses; protective effects in models of multiple sclerosis and airway inflammation have been documented since 2009.
Parasitic worms offer a different toolkit. Last February, researchers detailed in Nature Communications a schistosome protein called TGM6, which mimics human TGF-β and antagonizes fibroblast signaling—potentially useful for fibrosis and immune modulation. ES-62, secreted by the nematode Acanthocheilonema viteae, protected mice from collagen-induced arthritis via a gut–bone marrow axis mechanism, according to a January 2024 paper in Frontiers in Tropical Diseases. Another schistosome peptide induced regulatory T cells in mouse models of colitis and psoriasis.
The academic literature stretches back years. What's missing is translation.
Earlier attempts to harness parasites for medicine stumbled badly. Trichuris suis ova programs—yes, actual pig whipworm eggs—were discontinued a decade ago. A randomized trial of experimental hookworm infection in celiac disease, published in late 2020, showed acceptable safety but mixed efficacy. Turns out dosing patients with live parasites is both logistically nightmarish and not especially effective. The field has since pivoted hard toward defined, purified molecules.
That pivot is where artificial intelligence comes in—and where Ditto claims an edge.
According to the company's Y Combinator profile, Ditto has analyzed upward of one million parasite proteins using computational tools to predict binding to clinically validated immune targets. They say they've identified "thousands" of candidates, with early internal data showing binding affinities in the low-nanomolar range. Co-founder Dennis Sun posted in February that the team uses AI to predict targets of parasite effectors at scale, then engineers them into drug candidates.
The technical foundation is there, at least in theory. The developers of AlphaFold2 at DeepMind won the Nobel Prize in Chemistry two years ago for their work enabling protein structure predictions at unprecedented scale. A wave of peptide design frameworks—AfCycDesign, PepTune, CreoPep—emerged last year to help optimize stability and pharmacokinetics. Protein language models trained on metagenomic data are improving function prediction across millions of uncharacterized sequences, as a November 2024 arXiv preprint noted. Ditto estimates that more than 98% of parasite proteins remain uncharacterized, hinting at a vast design space.
Whether that space contains blockbuster drugs is another question entirely.
---
Who's Behind It—and Who's Alongside
Ditto Biosciences consists of three people, at least as of its Y Combinator application. The scientific firepower is real: Adair L. Borges holds a PhD from UCSF, did postdoctoral work as a Miller Fellow at UC Berkeley, and has more than 50 papers spanning parasitology, virology, and genomics. She's a co-inventor on an anti-CRISPR patent. Dennis Sun comes from Harvard and Berkeley with a background in evolutionary and developmental biology. Emily C.P. Weiss brings microbiology and genomics expertise, with prior stints at Illumina and DuPont.
The company operates out of MBC BioLabs in San Francisco—a shared wet lab space popular with early-stage biotech startups—and describes its mission as "evolutionary therapies for autoimmune disease." The stated plan: mine proteins from viruses, ticks, and worms to modulate human immunity. Beyond that? Details are scarce. No disclosed lead program, no IND timelines, no funding figures beyond whatever Y Combinator provides. The corporate website lists a sprawling indication map—rheumatoid arthritis, inflammatory bowel disease, psoriasis, lupus, multiple sclerosis, Type 1 diabetes—but no declared asset.
For venture scouts sizing up Ditto ahead of YC's Demo Day in late March, the core thesis is straightforward: parasites have already solved the hardest problem in immunology (suppressing host defenses without killing the host), and AI can now mine that solution space at industrial scale.
But Ditto is hardly alone in chasing immune tolerance. The field is crowded, well-capitalized, and moving fast.
Anokion reported positive symptom data from a Phase 2 trial of KAN-101, a liver-targeted gliadin antigen for celiac disease, early last year. (The trial was later terminated at sponsor discretion, though Anokion suggested the data was sufficient to advance the program.) COUR Pharmaceuticals uses tolerogenic nanoparticles; its lead asset CNP-104 for primary biliary cholangitis carries FDA Orphan and Fast Track designations, while CNP-103 for Type 1 diabetes got Fast Track status last March. A Genentech collaboration in 2024 brought big pharma validation. This past January, a Science Advances paper detailed COUR's mechanism: activation of the STING/Type I interferon pathway to drive tolerance. Parvus Therapeutics, which presents peptide-MHC complexes it calls "Navacims," struck a deal with AbbVie for inflammatory bowel disease; its lead program PVT401 hit a milestone last April and is moving toward IND-enabling studies.
The cell therapy side is equally active, if not more so. Sonoma Biotherapeutics released interim Phase 1 data last October for SBT-77-7101, a CAR-Treg targeting citrullinated proteins in refractory rheumatoid arthritis. Early results showed safety and improvement in joint counts. GentiBio plans to bring its Type 1 diabetes CAR-Treg candidate GNTI-122 into the clinic sometime this year. Quell Therapeutics presented preclinical data for QEL-005, a CAR-Treg for RA and systemic sclerosis, at last fall's ACR Convergence, with first-in-human trials expected in the first half of this year.
Then there's the "immune reset" narrative around CD19-targeted CAR-T. Cabaletta Bio's rese-cel showed what the company called "robust benefit" in lupus and myositis patients, with data releases spanning much of last year. Bristol Myers Squibb's NEX-T program BMS-986353, part of the Phase 1 Breakfree-1 trial, had treated 71 patients across systemic sclerosis, lupus, and inflammatory myopathies as of last October; 94% were off chronic immunosuppressants at time of analysis. The University of Chicago launched Phase 2 programs in January for lupus, myositis, and systemic sclerosis using CAR-T.
Meanwhile, the incumbent biologics keep expanding. UCB's bimekizumab—first dual IL-17A/F inhibitor—picked up U.S. approvals for psoriatic arthritis, non-radiographic axial spondyloarthritis, and ankylosing spondylitis in late 2024. Johnson & Johnson's guselkumab won ulcerative colitis approval last September and Crohn's disease approval this past March, becoming the first IL-23 inhibitor with both subcutaneous and intravenous induction options. It also showed superiority over Stelara on pooled endoscopic endpoints. AbbVie's risankizumab added ulcerative colitis mid-last year. Sanofi's teplizumab, approved in 2022 for Stage 2 Type 1 diabetes, filed a supplemental biologics license application last October to expand into newly diagnosed Stage 3 disease.
The competitive landscape, in other words, is sophisticated and unforgiving. Any new mechanism—parasite-derived or otherwise—will be benchmarked against high remission rates from IL-23 and IL-17 inhibitors, the durability signals from CAR-T, and the antigen-specific precision of nanoparticle platforms.
---
The Hard Questions
Which brings us to the make-or-break issues for Ditto and anyone else pursuing parasite-inspired therapies.
First: Can engineered parasite proteins match or beat current standards while offering differentiated durability or safety? The academic literature is rich with preclinical signals—evasins neutralizing chemokines, sialostatins reprogramming dendritic cells, helminth-derived mimics modulating TGF-β pathways. But preclinical models are not patients. The graveyard of autoimmune drug development is littered with candidates that looked great in mice.
Second: Can the practical challenges be solved? Many parasite proteins are cysteine-rich, requiring specialized expression systems. Extending half-life will almost certainly demand Fc fusions, PEGylation, or macrocyclization—none of which are trivial. Immunogenicity is a wildcard; Ditto claims to be building a tissue biobank to map immune memory against parasite proteins, a clever risk-mitigation move if it pans out. But that needs human validation, and soon.
Third—perhaps most critically for a three-person startup—what's the regulatory and commercial path? Purified parasite-derived proteins will likely be regulated as biologics, aligning with established FDA frameworks. That's far more tractable than live-organism approaches. The FDA's recent guidance on cell and gene therapy CMC flexibility, plus ongoing acceleration of biosimilar pathways, signals a regulator willing to meet innovation halfway.
But pricing dynamics are shifting underfoot. The 2026 Medicare maximum fair prices on Enbrel and Stelara, combined with expanding biosimilar competition, mean payers will demand clear value propositions. Durable remission, drug-free intervals, or true disease modification might command premium economics—but the evidentiary bar is high, and tolerance-based therapies inherently take longer to show durable benefit than drugs that simply block inflammation downstream.
Ditto hasn't disclosed a lead program, let alone IND timelines or funding beyond Y Combinator. The corporate site lists a sprawling indication map but no declared asset. For VCs and potential pharma partners sizing up the company ahead of Demo Day next month, the pitch is essentially this: parasites have already solved the hardest problem in immunology, and AI can now mine that solution space at scale. Believe it or not.
---
The Broader Bet
Step back, though, and a pattern emerges. Tolerance-based and immune-reset modalities—nanoparticles, CAR-Tregs, CAR-T therapies, and now parasite-derived proteins—are converging on the same strategic bet: that the next generation of autoimmune therapies won't just block inflammatory signals downstream but will reprogram the immune system upstream to restore self-tolerance.
It's an elegant idea. Also a hard one to execute. The human immune system evolved to be aggressive, promiscuous, and unforgiving—traits that kept our ancestors alive long enough to reproduce but now manifest as a rising tide of autoimmune disease in modern, sanitized environments. Teaching that system to stand down, selectively and durably, without opening the door to infection or cancer, is a high-wire act.
Whether Ditto's approach proves out is anyone's guess at this stage. The company is pre-clinical, pre-funding (beyond YC), and pre-almost everything that would let an outsider assess the technology rigorously. But the underlying science is real. The market pull is enormous. And the regulatory and reimbursement environment, for all its complexity, is evolving—perhaps faster than many expected—to reward durable innovation.
The question is no longer whether evolution's immune evasion tricks are worth studying. Academics have been doing that for decades. The question is whether a three-person biotech in a San Francisco wet lab can engineer those tricks into medicines fast enough to matter—and whether the tolerance-engineering thesis, writ large, can deliver on its considerable promise.
The answer will shape not just Ditto's fate but the trajectory of autoimmune drug development for years to come.
