Ethan Perlstein and the other n-of-1

The n-of-1 conversations in rare disease are dominated by the antisense oligonucleotide and adeno-associated virus programs that produce custom-built molecules for one identified mutation. Ethan Perlstein has spent the last decade arguing that there is a different, faster, and cheaper way to help many ultra-rare families: screen the thousands of already-approved small-molecule drugs against models of the affected person's specific genetic disease, and find the ones that already happen to work.

Perlstein founded Perlara in 2014 as the first biotech public benefit corporation, structured to take rare-disease cases families brought to him and run drug-repurposing screens in yeast, nematodes, fruit flies, and zebrafish carrying the patient-specific mutation. The premise was that an approved drug, already tested in millions of people for some other indication, that turns out to rescue a yeast model of a child's particular metabolic defect can move into compassionate use far faster than any custom drug that has to be designed, manufactured, and toxicity-tested from scratch. The Perlara model never produced a milasen-scale outcome. It produced something more diffuse and arguably more scalable.

Training

Perlstein earned his Ph.D. in molecular and cell biology at Harvard University, working with Stuart Schreiber. The Schreiber lab is one of the world's foundational chemical-biology programs, and Perlstein's graduate work used yeast as a screening platform for understanding small-molecule mechanism of action, with particular emphasis on the relationship between drug structure and cellular response across model organisms.

His postdoctoral work was at the Lewis-Sigler Institute for Integrative Genomics at Princeton University, as an independent postdoctoral fellow. The Lewis-Sigler model gives selected fellows their own laboratories at the postdoctoral stage, allowing them to develop independent research programs without immediately entering a faculty position. Perlstein used the time to develop what he later called evolutionary pharmacology, the framework for screening drugs against engineered models of human disease across the tree of life.

He left Princeton without taking a tenure-track faculty position, an unusual choice for someone with the academic profile he had built. The Perlara company was the alternative he chose.

Perlara

Perlstein incorporated Perlara, a public benefit corporation, in 2014 in the San Francisco Bay Area. The company was the first biotech to use the public benefit corporation legal structure, which requires the corporation to balance its for-profit mission with a stated social-benefit purpose and to report on both. Perlara's stated benefit purpose was rare-disease drug discovery for conditions too rare to support standard pharmaceutical development.

The company's operating model paired with affected families directly. A family with a child diagnosed with a rare genetic disease would contract with Perlara to develop a model organism carrying the child's specific mutation, screen approved drugs and other small-molecule libraries against that model, and identify candidate compounds that rescued the model's disease-related phenotype. The candidates were then made available to the family and their treating physician for off-label or compassionate-use evaluation under standard medical practice.

Perlara was an early Y Combinator portfolio company, the first biotech in the Y Combinator program. It raised approximately ten million dollars across multiple rounds before 2019, partnered with Novartis on a discovery collaboration in 2016, and partnered with Mission: Cure and other condition-specific foundations on individual rare-disease programs. The company worked with more than thirty client families over its first iteration. One of the spinouts, Maggie's Pearl, focused on a drug already approved in Japan for congenital disorder of glycosylation type 1a (PMM2-CDG), based on Perlara's screening.

Perlara wound down the original company structure in 2019. The pressures were the standard ones for a young biotech operating outside the conventional venture-capital model: the investment thesis was difficult to scale, the per-program revenue did not match the per-program cost, and the public benefit corporation form constrained the kinds of exits the company could pursue. Perlstein relaunched Perlara in approximately 2021 with a different operating model, focused on partnering with patient organizations and academic groups rather than serving as a contract research organization.

What evolutionary pharmacology actually does

Perlstein's technical contribution is a screening strategy that deserves to be understood on its own terms, separately from the company history.

The core observation is that many ultra-rare genetic diseases are caused by mutations in genes that are conserved across yeast, worms, flies, and humans. A defect in a metabolic pathway in a human child often produces a measurable defect in the same pathway in a yeast strain carrying the analogous mutation, and the yeast strain can be screened against thousands of approved drugs in days, where a mouse model would take months and cost ten times as much. The yeast screen identifies candidate compounds. Cross-species validation in worms and flies, which capture more of the cellular complexity humans have but yeast lack, refines the candidate list. The final candidates can be brought to mammalian validation only when there are a handful, not thousands.

The argument for the approach is speed. A child with a fatal metabolic disease cannot wait the ten years a standard drug-discovery pipeline takes. A yeast screen can produce candidate compounds in weeks. If the compounds are already approved drugs, the regulatory path to use them in the child is short.

The argument against the approach is hit rate. Many rare diseases do not have a tractable yeast or invertebrate model. Some mutations affect genes that are not conserved or affect human-specific cellular machinery (the central nervous system in particular has many human-specific features). When the model organism does not capture the human disease, the screen produces nothing useful.

In practice, Perlara's published programs and its public client outcomes split between conditions where the approach worked (PMM2-CDG, several rare metabolic disorders) and conditions where it did not. The successes were substantial enough to justify the model and limited enough to constrain the business case. Both outcomes are consistent with the underlying biology.

Curetopia

In late 2024, Perlstein launched Curetopia, his first decentralized-science (DeSci) project, structured as a BioDAO (a biotechnology decentralized autonomous organization). The Curetopia thesis is that the funding mechanism for ultra-rare disease drug discovery is the bottleneck more than the science. Conventional venture capital and standard philanthropy do not address the n-of-few cases efficiently. A DeSci structure, in which token-holding community members vote on which programs to fund and on what terms, could in principle aggregate small contributions from many supporters and direct them at specific n-of-1 or n-of-few programs without the overhead of traditional foundation infrastructure.

Curetopia raised approximately $1.77 million in its initial funding round in late 2024 and early 2025. The organization's stated goal is to fund rare-disease cures using the BioDAO model, with the affected community participating in governance. Perlstein serves as director of the Curetopia foundation.

He is also a scientific advisor to Cureledger, the parent platform that Zebra Reports operates within. Cureledger's data-trust architecture for rare-disease evidence and Curetopia's funding model address adjacent pieces of the same infrastructure problem: how to capture the evidence of what works for one ultra-rare patient, and how to fund the work that produces the evidence in the first place.

What the work has demonstrated

Perlstein's career arc is more disjointed than the careers of Yu, Crooke, Krainer, Chung, or Gray. The Perlara company has been wound down once and relaunched, the Curetopia BioDAO is at an early stage, and the underlying technical approach (model-organism drug repurposing) has produced results that are scientifically valid but commercially difficult.

The thesis that has held across all of Perlstein's work is that the ultra-rare disease problem is not primarily a chemistry problem. The chemistry of small-molecule drug repurposing, like the chemistry of antisense oligonucleotides and adeno-associated viruses, has been mature enough for years. The hard problems are funding mechanisms, family-and-academic partnerships, and the data infrastructure that connects what is tried for one patient to what gets tried for the next. Perlara was an attempt to solve the partnership problem at the company level. Curetopia is an attempt to solve the funding problem at the community level. Cureledger, where Perlstein advises, is an attempt to solve the data-infrastructure problem at the platform level.

The three together describe a complete model of what an ultra-rare drug-development infrastructure could look like outside the milasen-class ASO and the Pirovolakis-class AAV programs. Whether the model holds at scale is the question of the next decade. Perlstein has been asking it for one.