Living Molecular Twins™ to Accelerate Drug Development and Clinical Impact
Living Molecular Twins™ (LMTs™) are digital representations of real patients used to reduce the time and cost of drug development by simulating the clinical outcomes of each patient. Each LMT™ is the answer to the pitfalls of preclinical and clinical development, characterized by expensive and lengthy activities disconnected from the actual patients. Drug candidates and their mechanisms are not tested in a patient until phase 2 trials, while distribution of responders and non-responders is rarely, if ever, factored into trial design. These deficits lead to a 3-5% probability of success from initial clinical trials to approval. Unravel Biosciences uses its patient-focused Predictable Medicine™ innovation stack to generate populations of LMTs™, align R&D activities with real patient biology, and maximize the probability of success in the clinic.
Unravel uses BioNAV™, our computational AI platform, to generate LMTs™. We collect RNA directly from living patients and healthy relatives and use the data as an input to BioNAV™. We assess the transcriptome using a large causal inference network programmed with 121 million drug-gene interactions, 364 million unique network patterns, and over 40,000 molecules. This gives it the ability to assess 40,000 molecules’ interactions with the entire patient biological state and all downstream effects associated with each molecule, essentially conducting an in silico personalized clinical trial at the molecular level. Beyond its ability to generate LMTs™, BioNAV™’s extensive analysis provides deep insights into the mechanisms of disease and health, the ability to predict treatment responders and non-responders, identify biomarkers of treatment progression, as well as mechanisms of action that can be missing from the literature and therefore invisible to today’s AI LLMs.
LMTs™ can be generated for as many patients as necessary, creating a population of patients that truly represent its heterogeneity, outperforming preclinical models that, at best, recapitulate only one subgroup of a disorder. This requires patient transcriptomic data, which can be gathered rapidly and non-invasively with at-home collection from nasal swabs, along with other tissue collection methods. Our ever-growing datamine contains transcriptomes of approximately 400 patient & healthy control pairs across over 100 disorders, primarily from patients with rare neurodevelopmental disorders. The population is global, ranges from 0-70 years old, and includes longitudinal sampling of over 100 patients. LMTs™ are true digital twin models, acting as a bridge between the living patients and computational simulations, translating data into actionable insights to improve the lives of the patients.
Living Molecular Twins™ enable rapid clinical translation and a clear path to the clinic.
Benefit
Target-agnostic identification of novel drug targets and disease mechanisms
Value
By discovering novel drug targets and molecular pathways, BioNAV™ enables drug discovery in mere hours; side-stepping the current pace of research and discovery, which takes years and often decades, and extensive funding.
Preclinical model alignment with molecular assets
95% of molecules for new targets fail in clinical trials after successful preclinical studies. BioNAV™ identifies preclinical models that accurately represent the patient population’s response of each drug candidate to minimize clinical risk.
The biggest question in drug development is, will it work? BioNAV™ ensures the success of investigational molecules by providing a preview of clinical outcomes for any drug candidate, without putting a patient at risk.
Predict therapeutic efficacy in silico for any patient
No molecule is effective in all patients with the same indication. Prevent failure in clinical trials by stratifying patients into responders and non-responders prior to investing millions. Enables indication expansion into new patient populations. Develop companion diagnostics and biomarkers to improve clinical trial outcome measures and protect assets with additional IP.
Patient stratification by therapeutic response and identification of biomarkers for clinical trial I/E criteria
Model disease progression and treatment response with longitudinal sampling
Repeated BioNAV™ analyses over time increases understanding of the molecular mechanisms behind disease progression to identify drug candidates that will benefit patients in the long term. It also monitors molecular changes upon dosing with a drug candidate. Improve patient outcomes from gene therapies.
Case Studies
Stratification of COVID-19 Patient Population based on Predicted Therapeutic Response:
In one peer-reviewed study of 4,000 patients, BioNAV™ was used to predict the efficacy of statins in COVID-19 patients. During the pandemic, the drug class of statins were viewed interchangeably by clinicians. However, Unravel’s platform has demonstrated that not all molecules of the same class are interchangeable, and patients respond to different molecules differently. Using electronic health records of 4,000 patients, BioNAV™ predicted the efficacy of each statin on COVID-19 patient mortality with 80% clinical outcome accuracy. https://doi.org/10.1371/journal.pcbi.1011050
N-of-1 clinical success in an IRF2BPL-related disorder patient:
A Living Molecular Twin™ was generated for a 9-month-old with a life-altering IRF2BPL-related genetic disorder who was experiencing seizures. Though her clinicians had prescribed multiple steroids for her, none were working. After generating an LMT™ of the patient and screening 40,000+ molecules, a specific steroid was predicted to have high efficacy at treating her disease. The patient was prescribed the highly predicted steroid by her clinician and has since begun vocalizing, seen abolished seizures, has a cleaner EEG, and improved quality of life. The predictions were also used to support the choice of immunosuppressants for her gene therapy.
N-of-1 clinical success to phase I clinical trial in Rett syndrome:
A Living Molecular Twin™ was generated for a young boy with severe Rett syndrome in palliative care. His clinician prescribed vorinostat based on prior work published by Unravel’s founding members, which was corroborated specifically for this patient from his LMT™. Nearly two years later, he has seen improvement of multiple symptoms, including improved cognitive and motor function, sleep, and blood markers. Unravel launched a phase I clinical trial to evaluate vorinostat in 15 Rett syndrome female patients. https://clinicaltrials.gov/study/NCT07150013, https://doi.org/10.1038/s43856-025-00975-8
Authorship
This white paper was authored by Mikayla Reitsma, Frederic Vigneault, Richard Novak, and Eleni Pitsiniaga.
Last edited March 27, 2026.
©Unravel Biosciences 2026 | Contact@unravel.bio
| BioNAV™, Living Molecular Twin™, Predictable Medicine™ are exclusive trademarks of Unravel Biosciences |