Niven R. Narain, Ph.D., President & CEO, BPGbio.
In Gattaca, the 1997 sci-fi classic, a pair of siblings—one genetically engineered for perfection, the other naturally born—illustrate a dystopian vision where DNA determines destiny. The “in-valid” brother is denied opportunity not because of who he is, but because of what his genome might predict. At its core, the film is a cautionary tale of biological determinism—how genetic data, when misused, can restrict rather than empower.
Fortunately, real-world biology has followed a more promising trajectory. Since the Human Genome Project, the life sciences ecosystem—spanning government, academia and industry—has embraced biologic data to advance inclusion and precision. We’ve developed gene-editing technologies, created gene therapies and unlocked precision medicine, all fueled by an explosion of biologic insight.
But the future with biologic data doesn’t end here.
While much of the spotlight in AI over the past few years has focused on generative models and chatbot-like tools, the future lies in something far more biologically grounded: a dynamic, data-driven model of an individual’s biology, built using real patient samples, clinical records and multiomics data, that can simulate how real people respond to disease and treatment. In short, biological digital twins.
Why Biological Digital Twins Matter
Drug development today is slow, expensive and inefficient. It often takes over 10 years and $2 billion to bring a single drug to market—and more than 90% fail. Much of this stems from outdated approaches: reliance on population averages, poor preclinical models and siloed workflows across discovery, clinical and regulatory phases.
Biological digital twins flip that model. By creating a computational replica of an individual’s biology, informed by real patient data, researchers can simulate how diseases progress and how different interventions—drugs, devices or lifestyle changes—might alter that trajectory. Importantly, all teams from early discovery to commercial not only have access to the model but also derive actionable insights from computational analyses that may be used in target discovery, clinical trial design, health economics and reimbursement, among others.
This digital twin approach makes it possible to:
• Create a “biological passport” (a living in silico model of an individual’s multiomics signature over various times).
• Personalize drug response prediction before a trial ever begins or a patient is provided a therapy.
• Run in silico trials to prioritize compounds with the highest likelihood of success or matches to correcting a disease phenotype.
• Optimize clinical trial design with smarter patient selection and mechanistic endpoints.
• Shorten timelines by eliminating dead ends early.
In other words, digital twins can make drug development faster, more efficient and impactful to patient outcomes.
Why Digitizing Biology Begins With The Right Biobank
To build a digital twin, the raw materials matter. That starts with biobanking—the systematic, validated and coordinated collection, processing and storage of biological samples like organs, blood, urine, tumor tissue and cerebrospinal fluid—paired with deep and accurate clinical annotation. These longitudinal samples, collected from real patients over time, are the foundation for training AI models that reflect the true complexity of human disease.
At BPGbio, we’ve spent more than a decade building one of the most comprehensive clinically annotated biobanks in the U.S., covering oncology, neurology, rare diseases, inflammation and more. These samples are not just anonymous datapoints; they are living snapshots of patient biology, collected before, during and after diseases and treatments.
When paired with de-identified medical records, social and lifestyle information, imaging and outcomes data, they provide a uniquely rich substrate for building causal, mechanistic models that replicate biology—not just data artifacts.
From Biobank To Biological Digital Twin
The next step is digitalization—transforming those biospecimens into quantifiable data across multiple biological layers: genomics, transcriptomics, proteomics, lipidomics, metabolomics. This multiscale, multiomics approach reveals not only what’s happening at the genetic level, but how that cascades into cellular behavior, tissue dysfunction and whole-body disease.
But the human body is not a simple machine; it is an engineering marvel. That’s where Bayesian causal AI and supercomputing come in. Unlike black-box algorithms that can only detect statistical correlations, biology-first AI platforms can infer cause-and-effect relationships within complex biological systems. They help us ask not just what is happening in a disease, but why, where and what we can do about it.
By integrating real-time clinical data with longitudinal biology, digital twins allow researchers and clinicians to test hypotheses in silico before committing to costly and risky clinical trials. They help us design smarter trials with adaptive protocols, select the patients most likely to respond and de-risk drug development from preclinical through Phase III. In a world where fewer than 10% of drugs that enter clinical trials ultimately succeed, this kind of simulation power isn’t just nice to have—it’s essential.
A New Path Forward
The U.S. FDA’s recent steps, such as its initiative to phase out animal testing for biologics and its support for Bayesian methods and AI-assisted review, reflect a growing regulatory alignment to this approach and a utility in a broader context.
Digital twins represent more than a technological evolution. It’s also one that is philosophical and ethical. Instead of treating patients as averages or relying on outdated models, we now have the tools to understand biology at an individual level and give the right patients a higher chance of responding to a therapy.
Unlike the genetically “perfect” brother in Gattaca, a biological digital twin isn’t fixed at birth. It evolves with data; it learns. And, most importantly, it empowers patients by helping science work with their biology, not against it.
Closing Thoughts
The era of digital twins offers a future where deep biology, AI and real-world data combine to make medicine more personal, more precise and more human. The challenge now is not whether we can build digital twins. The industry already has, with projects like the artificial pancreas and the “Living Heart Project.” The challenge is how quickly we can scale this approach across disease areas, integrate it into regulatory pathways and ultimately deliver on its promise for patients around the world.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?