A biotechnology firm is creating comprehensive synthetic datasets to model the human body, aiming to address a critical lack of diverse and accessible medical data for research and development. The initiative by Mantis Biotech focuses on generating what are known as “digital twins,” virtual replicas that represent human anatomy, physiology, and behavior.
The company consolidates disparate and often siloed sources of health information to produce these synthetic datasets. This process is designed to create robust, privacy-compliant data pools that can be used by researchers and pharmaceutical developers without relying solely on real patient data, which can be scarce or difficult to access due to privacy regulations.
The Data Availability Challenge in Medicine
A significant bottleneck in medical innovation is the availability of high-quality, diverse clinical data. Drug development, treatment personalization, and medical device testing often require vast amounts of information that reflects a wide population. Real-world data collection is expensive, time-consuming, and fraught with privacy concerns under regulations like the General Data Protection Regulation (GDPR) in Europe and the Health Insurance Portability and Accountability Act (HIPAA) in the United States.
Mantis Biotech’s approach seeks to mitigate these issues. By building synthetic datasets, the company provides an alternative that mimics the statistical properties of real human data without being directly linked to identifiable individuals. This method is intended to accelerate research timelines and reduce dependency on limited clinical trial cohorts.
How digital twin Technology Works
The core product is a sophisticated digital model, or “twin,” of a human system. These are not simple avatars but complex computational models built from aggregated data on genetics, biomarkers, organ function, and even lifestyle factors. The models can be used to simulate disease progression, predict how a body might respond to a new drug, or test the safety of a medical device in a virtual environment.
The technology relies on advanced algorithms and artificial intelligence to integrate information from various sources. These sources may include anonymized electronic health records, genomic databases, data from wearable devices, and published clinical studies. The synthesis aims to create a holistic and dynamic representation that can be queried and tested computationally.
Potential Applications and Implications
The primary application seen for this technology is in the preclinical phases of pharmaceutical research. Companies could use digital patient cohorts to screen drug candidates more efficiently, potentially reducing the number of failed clinical trials. Other applications include personalized medicine, where a model could be tailored with an individual’s data to forecast treatment outcomes, and medical education, providing students with interactive models of disease.
However, the development of such technology also raises important questions. Experts note that the fidelity and predictive power of these digital twins depend entirely on the quality and breadth of the data used to train them. Biases in the original data sources could be perpetuated or amplified in the synthetic versions. Furthermore, the regulatory pathway for approving therapies developed or validated using synthetic data remains an area of active discussion among agencies like the U.S. Food and Drug Administration.
Looking Ahead
The field of synthetic data and digital twins is rapidly evolving. Mantis Biotech is part of a broader movement within biotech and healthcare information technology seeking computational solutions to physical world constraints. Industry observers anticipate increased collaboration between synthetic data firms and major pharmaceutical companies in the coming years. The next steps for technologies like this will involve rigorous validation against real-world clinical outcomes and the establishment of clear regulatory frameworks to govern their use in the drug approval process.
Source: GeekWire
