Drug Repurposing
Old molecules.
New hope.
We use AI, advanced in vitro models, and Real World Data to find new treatments for rare diseases hidden in approved drugs around the world.
Get in touch
Drug Repurposing
Old molecules.
New hope.
We use AI, advanced in vitro models, and Real World Data to find new treatments for rare diseases hidden in approved drugs around the world.
Get in touch
Targeting
Heterogeneous patient populations lead to heterogeneous results.
We lead with the genetics, not the diagnosis.
A genetic mutation, not a symptom list, is our starting point. By targeting patients with a single, well-established monogenic cause, we avoid the trap that sinks most rare disease programs: treating biologically distinct conditions as if they were the same disease.
Screening
Most rare disease companies start with a disease.
We start with the entire rare disease landscape.
By evaluating millions of drug–disease pairs, we identify the programs with the strongest scientific rationale and development path to bring new treatments to patients faster.
Selection
Most drug development failures trace back to candidate selection.
We match the right drug, with the right target, with the right patients.
Over 300,000 distinct mutations are likely causal1 for the 6,000+ rare diseases cataloged.2 We screen approved and investigational medicines to find and formulate the right medicine for the causal mutation.
Approval
Only 12% of repurposed drugs in pediatric rare disease trials have reached approval3
We model which drug–disease pairs have a high probability of technical and regulatory success.
Clinical failure in rare disease follows identifiable patterns. By examining dozens of disease- and drug-related factors across more than 500 drugs and 7,000 rare disease clinical trials, we have characterized those patterns and use them to evaluate every candidate for its probability of technical and regulatory success.
90%
Over 90% of drugs that appear safe in animals fail to receive FDA approval4
Safety
Animal studies can't fully anticipate human responses.
We use New Approach Methodologies (NAMs): real tissue, grown from real patients, tested before anyone is exposed to risk.
Differences in target biology, pharmacokinetics, metabolism, and off-target effects between species mean many risks only surface in human trials. Using patient-derived organoids, we test candidate therapies directly in human tissue — before a compound ever reaches an animal model or a patient.
{{ safetyQuoteWords }}5
Trials
Randomized controlled trials are not always well suited to rare diseases.
We design small, focused trials around the patients most likely to respond.
In a conventional randomized controlled trial, responders are mixed with non-responders, so the treatment signal is diluted and many more patients are needed to detect it. We select patients by genetic subtype, so a larger share of each cohort is expected to respond. Natural history evidence serves as the external comparator, and patient-as-own-control designs compare each patient against their own pre-treatment course. Including patient-as-own-control designs and natural history comparators makes clinical trials smaller, faster, and more likely to succeed.6 Mathematically modeled designs reduce enrollment, cost, and timelines while maintaining scientific rigor.
Regulatory precedent
The FDA has approved 13 rare disease therapies on the basis of natural history or patient-as-own-control trial designs, with pivotal studies as small as 10 patients.6
Randomized controlled trial
A mixed population is split between treatment and control, and responders are a small share of each arm.
Advanced trial design
Genetically led and mathematically modeled trial designs improve trial speed and success rates.
Contact
Get in touch with the team.
Proxy works with government agencies, foundations, academic groups, and industry partners on rare disease programs. Reach out about a partnership, a collaboration, or a specific target you think belongs in our pipeline.
References
- 1 ClinVar. National Library of Medicine, National Center for Biotechnology Information. Accessed September 24, 2026. https://www.ncbi.nlm.nih.gov/clinvar/
- 2 Nguengang Wakap S, Lambert DM, Olry A, et al. Estimating cumulative point prevalence of rare diseases: analysis of the Orphanet database. Eur J Hum Genet. 2020;28(2):165-173. doi:10.1038/s41431-019-0508-0
- 3 Approval of repurposed drugs in industry-sponsored pediatric rare disease trials, 2001–2020: analysis of ClinicalTrials.gov, Orphanet, Open Targets, and openFDA data. Proxy Research. September 2026. Accessed September 25, 2026. Methods and data terms
- 4 FDA achieves year 1 goals in reducing animal testing in drug development. News release. US Food and Drug Administration. April 20, 2026. Accessed September 24, 2026. https://www.fda.gov/news-events/press-announcements/fda-achieves-year-1-goals-reducing-animal-testing-drug-development
- 5 FDA announces plan to phase out animal testing requirement for monoclonal antibodies and other drugs. News release. US Food and Drug Administration. April 10, 2025. Accessed September 25, 2026. https://www.fda.gov/news-events/press-announcements/fda-announces-plan-phase-out-animal-testing-requirement-monoclonal-antibodies-and-other-drugs
- 6 Summar ML, Woodcock J. Supply and demand in the mathematics of rare disease drug development: why choosing the right model is crucial. Clin Transl Sci. 2026;19(5):e70551. doi:10.1111/cts.70551