Recently, digital twins (AI‑powered, real‑time virtual replicas of individual patients) have emerged as one of the most transformative innovations in personalized medicine. Researchers at the Weizmann Institute’s Human Phenotype Project have unlocked the potential to predict disease risks and tailor preventive treatments by creating digital twins rooted in deep clinical and multi-omic datasets. In India, surgeons are increasingly leveraging digital twins based on patients’ live hemodynamic data to simulate different surgical scenarios, helping pinpoint the safest and most effective interventions, especially in complex cardiovascular cases.
These advances underscore a growing trend: digital twins are moving beyond concept into practical applications, offering an unprecedented opportunity to shift from reactive care to simulations-driven, precision-guided treatment strategies that could substantially enhance outcomes and efficiency in clinical decision-making.
We have covered this trend before.
The offer to create a biological twin
That’s how I have come across a company called BioTwin. They offered me to create my biological twin and I accepted the offer, as always, without any payment of sponsorship.
To achieve that, I have to provide 10 dried blood spot (DBS) samples. One sample a day for a week, and maybe a few more before and after exercise.
They don’t analyze DNA, but create untargeted metabolomic signatures, generating an advanced biological profile for me.
What do they do?
BioTwin is developing a digital twin platform as a clinical decision support tool, with a particular focus on early disease detection and personalized treatment planning. Their approach relies on untargeted metabolomic profiling, which captures metabolic signals from blood samples rather than focusing on a few predefined biomarkers.
The idea is that subtle metabolic disturbances could indicate health changes before they are visible through imaging or symptomatic to the patient. This concept is being evaluated in a validation study with Cleveland Clinic Abu Dhabi, particularly in oncology applications like breast and colon cancer.
Beyond detection, BioTwin envisions the use of “virtual twin” simulations: clinicians could adjust a patient’s digital twin with hypothetical interventions, such as adding a medication, and observe predicted metabolic responses. This might make it possible to compare how different treatment strategies could play out for an individual, or benchmark a patient’s twin against profiles of others who responded well or poorly to a given therapy. Current pilot developments include oncology treatment planning and support for Parkinson’s medication management.
In parallel, BioTwin is exploring how metabolomic data can be used to generate lifestyle and physiological health scores. These are derived entirely from blood-based metabolomic fingerprints, without direct reliance on wearable or survey data, in order to reduce external bias. The predictive models were trained on thousands of metabolomic profiles paired with real-world physiological measures (such as weight, body composition, heart rate, HRV, and sleep patterns), allowing the algorithms to learn metabolic signatures that correspond to these traits. When only the blood sample is available, the system can still output estimates of metrics such as body fat, muscle and bone mass, heart rate, and stress indices.
The company also acknowledges the current limitations of this work. Some models are trained on relatively small datasets (for example, a few hundred ground-truth measurements for heart rate variability or body composition) which increases uncertainty in certain predictions. Annexes in their pilot reports flag these margins of error. Furthermore, while the models can currently provide lifestyle-related metrics, they do not yet generate diagnostic outputs for diseases such as diabetes or cancer. BioTwin emphasizes that clinical-grade disease predictions are still under development and require ongoing validation within research collaborations.
My biological twin
Here are some preliminary results:
- I cannot believe that my body fat is 22%. It must be lower than that.
- But my muscle mass is bigger than average. Hooray! All those exercises are worth it.
- I’m well hydrated.10 glasses of water a day.
- My resting heart rate, in fact, is around 57, not 68.
- My VOMax is 37 as I got that measured.
What’s next?
They told me that they treat these results as a “thank you” to research participants (e.g., early-oncology detection). And that it might have two core benefits. For clinicians, the body-composition insights are handy when they don’t see the patient in person (BMI alone is a poor proxy). For other users, everyone’s metabolism is different; BioTwin personalizes scores to their metabolomic profile. Users are more likely to follow advice that comes from their own virtual twin than from population averages.
My verdict
I have mixed results to say the least. My biological age, which the company generated, changed twice during the process. The initial results I shared above also changed. Even though I filled in many questionnaires about my health and lifestyle, it seemed some results must be off because I measure my health with so many other methods too.
But the most important part is that I still have no idea what I could use the results for.
I don’t feel like I have a biological twin, which helps me make better decisions about my health or disease management. The service sounds like a metabolomic profile (with many open questions about accuracy) but this experience is far from what real biological twins will able to deliver such as being able to test drugs or treatments on them before giving them to patients.
What I have learn along the way is that we might be in the phase where many companies will call their products digital twins, just like how many companies announced that they would be using machine learning, while in fact, it might have only been an Excel spreadsheet and a macro.
In the meantime, I’ll keep on trying other biological age tests too.
The post What Can You Learn From Your Digital Twin: BioTwin Review appeared first on The Medical Futurist.