Digital Twins & In Silico Trials
Unlearn.AI
by Unlearn.AI, Inc.
AI-generated digital twins replacing placebo arms to accelerate clinical trials with fewer patients
Category
Digital Twins & In Silico Trials
Founded
2017
Headquarters
San Francisco, CA, USA
Overview
Unlearn.AI has developed TwinRCT, a platform that generates AI-powered digital twin controls to augment or replace placebo arms in randomized clinical trials. The system trains prognostic models on historical trial data for a specific disease, then uses these models to predict each patient's future disease trajectory if they had received placebo. These predicted counterfactuals — digital twins — are incorporated as synthetic control data alongside the actual control arm, increasing statistical power and enabling trials to be run with fewer randomized patients. FDA and EMA have engaged with Unlearn's methodology, and regulatory qualification submissions have been filed. Clinical development teams at pharmaceutical and biotech companies use Unlearn's platform to reduce the number of patients randomized to placebo — improving trial ethics, reducing costs, and shortening timelines. Applications span multiple sclerosis, Alzheimer's disease, ALS, and other CNS conditions where the company has trained validated prognostic models. In multiple sclerosis trials, Unlearn's TwinRCT has demonstrated the ability to reduce placebo arm size by 30-50% without compromising statistical validity. Unlearn's regulatory engagement is a core strategic asset: the company's statisticians and data scientists have worked directly with FDA and EMA to establish guidance on the use of digital twin controls under the 21st Century Cures Act framework for complex innovative trial designs. Unlike general Bayesian adaptive trial methods, Unlearn's approach generates patient-level synthetic data grounded in disease-specific mechanistic and statistical models, enabling transparent audit trails that satisfy regulatory requirements. The company has published peer-reviewed validations and has been adopted by leading academic medical centers and top-ten pharma companies.
Key Features
Multi-Scale Physiological Modeling
Connect molecular interactions to organ-level responses with multi-scale biological models.
In Silico Clinical Trials
Virtual clinical trials reduce time and cost of traditional Phase I-III studies by 30-50%.
Virtual Patient Modeling
Create digital patient models simulating drug responses across diverse population demographics.
Real-World Data Integration
Calibrate and validate models using real-world clinical data from healthcare systems.
Regulatory Evidence Generation
Generate computational evidence packages aligned with FDA guidance for regulatory submissions.
Pros & Cons
Pros
- +Integration with real-world clinical data improves model calibration and prediction accuracy
- +Predictive toxicology models identify safety liabilities before first-in-human dosing
- +Synthetic control arms reduce the need for placebo groups in rare disease clinical trials
- +Regulatory acceptance growing with FDA guidance on computational modeling for device and drug evaluation
- +Multi-scale modeling connects molecular interactions to organ-level physiological responses
- +In silico clinical trials reduce time and cost of traditional Phase I-III studies by 30-50%
Cons
- −Regulatory acceptance of in silico evidence varies across jurisdictions and therapeutic areas
- −Model validation against real clinical data is essential but time-consuming and expensive
- −Adoption requires significant cultural change in organizations accustomed to traditional trial designs
- −Requires extensive clinical data for initial model calibration and ongoing validation
- −Computational models cannot fully capture the complexity of human biological variability
Use Cases
Research Workflow Optimization
AI-powered optimization of research workflows to accelerate discovery timelines and improve reproducibility.
Data Analysis & Insights
Machine learning analysis of complex biological datasets to extract actionable insights and identify patterns.
Collaboration & Knowledge Management
Platform-enabled collaboration across distributed research teams with integrated data sharing and knowledge capture.