What do you think?
Rate this book
Virtual You: How Building Your Digital Twin Will Revolutionize Medicine and Change Your Life
The visionary science behind the digital human twins that will enhance our health and our future
Virtual You is a panoramic account of efforts by scientists around the world to build digital twins of human beings, from cells and tissues to organs and whole bodies. These virtual copies will usher in a new era of personalized medicine, one in which your digital twin can help predict your risk of disease, participate in virtual drug trials, shed light on the diet and lifestyle changes that are best for you, and help identify therapies to enhance your well-being and extend your lifespan―but thorny challenges remain.
In this deeply illuminating book, Peter Coveney and Roger Highfield reveal what it will take to build a virtual, functional copy of a person in five steps. Along the way, they take you on a fantastic voyage through the complexity of the human body, describing the latest scientific and technological advances―from multiscale modeling to extraordinary new forms of computing―that will make “virtual you” a reality, while also considering the ethical questions inherent to realizing truly predictive medicine.
With an incisive foreword by Nobel Prize–winning biologist Venki Ramakrishnan, Virtual You is science at its most astounding, showing how our virtual twins and even whole populations of virtual humans promise to transform our health and our lives in the coming decades.
Virtual You is a panoramic account of efforts by scientists around the world to build digital twins of human beings, from cells and tissues to organs and whole bodies. These virtual copies will usher in a new era of personalized medicine, one in which your digital twin can help predict your risk of disease, participate in virtual drug trials, shed light on the diet and lifestyle changes that are best for you, and help identify therapies to enhance your well-being and extend your lifespan―but thorny challenges remain.
In this deeply illuminating book, Peter Coveney and Roger Highfield reveal what it will take to build a virtual, functional copy of a person in five steps. Along the way, they take you on a fantastic voyage through the complexity of the human body, describing the latest scientific and technological advances―from multiscale modeling to extraordinary new forms of computing―that will make “virtual you” a reality, while also considering the ethical questions inherent to realizing truly predictive medicine.
With an incisive foreword by Nobel Prize–winning biologist Venki Ramakrishnan, Virtual You is science at its most astounding, showing how our virtual twins and even whole populations of virtual humans promise to transform our health and our lives in the coming decades.
336 pages, Hardcover
Published March 28, 2023
16 people are currently reading
269 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 8 of 8 reviews
October 20, 2025
Many experiments that were heretofore unachievable are now possible because of ever-increasing computer power and computing capabilities at bewildering speed. Supercomputers make it possible to model weather and climate with increasing accuracy. Using AI, we can research drug development in innovative ways not possible before. This book is an ambitious venture into modeling and simulation. Dr. Coveney presents ideas to create a digital twin of each one of us and experiment with it to improve and predict our future health. This book’s core concept is that science’s convergence is changing medicine. Patient data, theory, algorithms, AI, and powerful computers are leading us to a more quantitative and predictive approach. As Aristotle once remarked, knowing yourself propels all wisdom, and the digital twin aims to do just that, step by step.
The authors define the fundamental steps required to create a digital twin. They are: harvest diverse data about the body, craft a theory to make sense of all these data, and use mathematics to understand the fundamental limits of simulations. Next, harness computers to spark a mathematical understanding of the human body. This leads us to blend the insights of natural and artificial intelligence to interpret data and shape our understanding. This is the ‘Virtual You’, whose ultimate aim is to enable doctors to gaze into the future of a patient. Not a typical patient or an average patient but a specific patient, with their individual baggage of inheritance, upbringing and environment. Modeling aims to eliminate intrusive examinations. It will give us the freedom to explore “what if” scenarios. We can observe the effects of interfering with a protein using a drug molecule, or mutating or removing a gene from the models to understand their effects.
Large-scale models are easy to run on a computer, but are not accurate. Detailed microscale models are precise, but painfully time-consuming. With weather forecasting, for example, running a model that resolves down to a mesh comprising squares 10 km on each side will miss a lot of detailed weather. Other inaccuracies would exist, though it proceeds quicker than a 1 km resolution model, despite the unchanging mathematics underneath. While modeling the human body, we have to start our simulation somewhere. The cell is the basic building block of the body. It might be the right starting point to create the Virtual You. Beyond individual cells, we will simulate virtual tissues and organs. Let us look at how we can simulate some of the important body parts.
With a wealth of molecular and genetic data and increasingly detailed imaging of the entire lung, we can now create a Virtual Lung. We can simulate its lobes and geometry as a mesh. The virtual lungs can reveal the impact of asthma and the effects of smoking. It can show how they damage the airways and gas exchange tissue, making it harder to get enough oxygen. We can use the model to understand lung changes in patients with acute pulmonary embolism, a blockage of a pulmonary artery, often caused by a blood clot.
The pancreas produces juices that aid in digestion. Scientists have created artificial versions of the pancreas, offering hope to type 1 diabetes patients who struggle with insulin deficiency or resistance. The virtual pancreas helps patients regulate their blood glucose by delivering the right amount of insulin. It does this by using a mathematical model of human glucose metabolism and a closed-loop control algorithm. The algorithm models insulin delivery using data from a glucose sensor implanted in the patient. We can customize these into a patient-specific digital twin of the pancreas. It helps us to calculate continuously how much insulin is required and to control an implanted pump to maintain blood insulin concentrations.
The goal is to have mathematical models of a patient, which we will frequently update with information from the body to produce health forecasts. We can do it at different levels. There could be an average model that gives overall insights into lifestyle and symptoms. More sophisticated insights can come from Virtual Yous, primed with a person’s data, from DNA code to body scans, that are personalised. The digital twin would be a lifelong, personalized clone. It would age like us, receiving constant updates from measurements, scans, and medical exams. It would also incorporate behavioral, environmental, and genetic data. Maybe in the next twenty years, we can experiment with digital copies of ourselves. This would replace the use of trial data of people similar to us in predicting our fate. Going further, why stop at one virtual twin? Within a supercomputer, many virtual versions of us can co-exist and explore our many futures, depending on our diet, medication, lifestyle and environment. Our virtual future will arrive more quickly than actual reality.
Experimenting with our digital twins and trying various ‘what if’ scenarios sounds impressive and alluring, but the reality involves many bumps and blocks to negotiate. Let us examine certain constraints and difficulties.
The first is the accuracy of our digital twins. The usefulness of computer models hinges upon the dependability of data concerning ourselves. Our bodies are complex at the cellular level. The model could be incomplete, biased, or inconsistent because of its granularity. Hence, it can lead to unreliable predictions or diagnoses. Unreliable output can lead to loss of trust and faith in using the digital twin. Besides, the data is private, and hence becomes proprietary and limits access to it by researchers. This acts as an impediment to benefiting from the vast amount of data.
The human body is a marvel of intricate complexity. Even if we use powerful computers, the models would still be simplifications and approximations of intricate biological processes. When we use it to study the interactions between a drug and the body, predicting the side effects and long-term consequences with accuracy could be a significant hurdle. Our virtual body needs to ensure that the models reflect our actual physiology and represent the progress of our disease with accuracy. It is a critical and continuing challenge. The underperformance of a high-profile drug, created using virtual models, might undermine people’s faith in the digital twin.
The next thing to consider is the amount of computing power. Twenty years ago, when researchers invented cryptocurrencies, the amount of power used in mining them alarmed people. A decade later, AI used immense amounts of power to train its LLMs, and critics talked about the dangers of its climate impact. But the digital twin could dwarf even the AI’s greed for power. Running complex simulations of the human body requires immense computational resources. Today’s world has almost eight billion people. Imagine simulating eight billion digital twins and the interaction of each one with drugs to cure their ailments. Besides, it is a lifelong continuous process. These factors limit the pursuit of the idea.
Next, modifications to the regulatory framework must allow experimental drugs designed for each person, tested only on digital twins. How can we validate and seamlessly incorporate them into the current drug approval process?
There are justifiable elements of excitement and high expectation surrounding computer models and the use of AI in drug discovery. Tangible successes and growing adoption show it is a powerful and transformative tool. As our understanding of biology continues to advance, these models will get better in their accuracy of representing the human body. They will then play a more crucial role in the future of medicine, making drug development faster, cheaper, and more effective for everyone. But we remember that a lot of predictions based on computer models have failed in history, especially when the models depended on human societies and their behavior. Examples include computer models from the 1960s predicting population explosions and famines, and models from the 1970s predicting resource depletion by the 1990s. Another example is climate models predicting the Arctic Ocean becoming iceless in summer by 2015. Hence, it will be prudent for the average citizen to tread with caution and treat the virtual representation as an approximation instead of a carbon copy.
The book offers a broad perspective on scientific efforts to create digital representations of ourselves. It has an analytical foreword by the Nobel-prize-winning biologist, Venki Ramakrishnan. It is difficult reading but worth the time to learn the marvels of where science is heading.
The authors define the fundamental steps required to create a digital twin. They are: harvest diverse data about the body, craft a theory to make sense of all these data, and use mathematics to understand the fundamental limits of simulations. Next, harness computers to spark a mathematical understanding of the human body. This leads us to blend the insights of natural and artificial intelligence to interpret data and shape our understanding. This is the ‘Virtual You’, whose ultimate aim is to enable doctors to gaze into the future of a patient. Not a typical patient or an average patient but a specific patient, with their individual baggage of inheritance, upbringing and environment. Modeling aims to eliminate intrusive examinations. It will give us the freedom to explore “what if” scenarios. We can observe the effects of interfering with a protein using a drug molecule, or mutating or removing a gene from the models to understand their effects.
Large-scale models are easy to run on a computer, but are not accurate. Detailed microscale models are precise, but painfully time-consuming. With weather forecasting, for example, running a model that resolves down to a mesh comprising squares 10 km on each side will miss a lot of detailed weather. Other inaccuracies would exist, though it proceeds quicker than a 1 km resolution model, despite the unchanging mathematics underneath. While modeling the human body, we have to start our simulation somewhere. The cell is the basic building block of the body. It might be the right starting point to create the Virtual You. Beyond individual cells, we will simulate virtual tissues and organs. Let us look at how we can simulate some of the important body parts.
With a wealth of molecular and genetic data and increasingly detailed imaging of the entire lung, we can now create a Virtual Lung. We can simulate its lobes and geometry as a mesh. The virtual lungs can reveal the impact of asthma and the effects of smoking. It can show how they damage the airways and gas exchange tissue, making it harder to get enough oxygen. We can use the model to understand lung changes in patients with acute pulmonary embolism, a blockage of a pulmonary artery, often caused by a blood clot.
The pancreas produces juices that aid in digestion. Scientists have created artificial versions of the pancreas, offering hope to type 1 diabetes patients who struggle with insulin deficiency or resistance. The virtual pancreas helps patients regulate their blood glucose by delivering the right amount of insulin. It does this by using a mathematical model of human glucose metabolism and a closed-loop control algorithm. The algorithm models insulin delivery using data from a glucose sensor implanted in the patient. We can customize these into a patient-specific digital twin of the pancreas. It helps us to calculate continuously how much insulin is required and to control an implanted pump to maintain blood insulin concentrations.
The goal is to have mathematical models of a patient, which we will frequently update with information from the body to produce health forecasts. We can do it at different levels. There could be an average model that gives overall insights into lifestyle and symptoms. More sophisticated insights can come from Virtual Yous, primed with a person’s data, from DNA code to body scans, that are personalised. The digital twin would be a lifelong, personalized clone. It would age like us, receiving constant updates from measurements, scans, and medical exams. It would also incorporate behavioral, environmental, and genetic data. Maybe in the next twenty years, we can experiment with digital copies of ourselves. This would replace the use of trial data of people similar to us in predicting our fate. Going further, why stop at one virtual twin? Within a supercomputer, many virtual versions of us can co-exist and explore our many futures, depending on our diet, medication, lifestyle and environment. Our virtual future will arrive more quickly than actual reality.
Experimenting with our digital twins and trying various ‘what if’ scenarios sounds impressive and alluring, but the reality involves many bumps and blocks to negotiate. Let us examine certain constraints and difficulties.
The first is the accuracy of our digital twins. The usefulness of computer models hinges upon the dependability of data concerning ourselves. Our bodies are complex at the cellular level. The model could be incomplete, biased, or inconsistent because of its granularity. Hence, it can lead to unreliable predictions or diagnoses. Unreliable output can lead to loss of trust and faith in using the digital twin. Besides, the data is private, and hence becomes proprietary and limits access to it by researchers. This acts as an impediment to benefiting from the vast amount of data.
The human body is a marvel of intricate complexity. Even if we use powerful computers, the models would still be simplifications and approximations of intricate biological processes. When we use it to study the interactions between a drug and the body, predicting the side effects and long-term consequences with accuracy could be a significant hurdle. Our virtual body needs to ensure that the models reflect our actual physiology and represent the progress of our disease with accuracy. It is a critical and continuing challenge. The underperformance of a high-profile drug, created using virtual models, might undermine people’s faith in the digital twin.
The next thing to consider is the amount of computing power. Twenty years ago, when researchers invented cryptocurrencies, the amount of power used in mining them alarmed people. A decade later, AI used immense amounts of power to train its LLMs, and critics talked about the dangers of its climate impact. But the digital twin could dwarf even the AI’s greed for power. Running complex simulations of the human body requires immense computational resources. Today’s world has almost eight billion people. Imagine simulating eight billion digital twins and the interaction of each one with drugs to cure their ailments. Besides, it is a lifelong continuous process. These factors limit the pursuit of the idea.
Next, modifications to the regulatory framework must allow experimental drugs designed for each person, tested only on digital twins. How can we validate and seamlessly incorporate them into the current drug approval process?
There are justifiable elements of excitement and high expectation surrounding computer models and the use of AI in drug discovery. Tangible successes and growing adoption show it is a powerful and transformative tool. As our understanding of biology continues to advance, these models will get better in their accuracy of representing the human body. They will then play a more crucial role in the future of medicine, making drug development faster, cheaper, and more effective for everyone. But we remember that a lot of predictions based on computer models have failed in history, especially when the models depended on human societies and their behavior. Examples include computer models from the 1960s predicting population explosions and famines, and models from the 1970s predicting resource depletion by the 1990s. Another example is climate models predicting the Arctic Ocean becoming iceless in summer by 2015. Hence, it will be prudent for the average citizen to tread with caution and treat the virtual representation as an approximation instead of a carbon copy.
The book offers a broad perspective on scientific efforts to create digital representations of ourselves. It has an analytical foreword by the Nobel-prize-winning biologist, Venki Ramakrishnan. It is difficult reading but worth the time to learn the marvels of where science is heading.
Read
October 13, 2025'Those who have handled sciences have been either men of experiment or men of dogmas. The men of experiment are like the ant, they only collect and use; the reasoners resemble spiders, who make cobwebs out of their own substance. But the be takes a middle course: it gathers its material from the flower of the garden and of the field, but transforms and digests it by a power of its own.'
- Francis Bacon, Novum Organum, sive Indicia Vera de Interpretatione Naturae (1620)
- Francis Bacon, Novum Organum, sive Indicia Vera de Interpretatione Naturae (1620)
September 3, 2023
Virtual You is not a light reading but excellent food for thought.
Highfield and Coveney’s dream of the Virtual You goes one step beyond the precision medicine, which creates your predictions based on your doppelgängers, or others “like you”. They envision a healthcast, just like the earth model, which leads to increasingly accurate Nowcasts in your local area.
Wouldn’t it be clever to fix all your future health mistakes without actually making them? “Don’t eat this stuff - it’ll make you sick! Don’t worry about triathlons, just do yoga and you’ll be fine. And oh, do your bone scans every year instead of never, will you?” That kind of stuff.
So the Virtual You presents a perfect recipe:
Create a 3D model of a cell/tissue/ organ/entire body, work out the underlying theory and a mathematical model that uses a variety of physics and biological processes, add a judicious sprinkling of AI to stir it all up and make sense of data in the multi-scale cauldron. Then display the outcome in an avatar that is your replica, a virtual You that can predict how to make your biological You live and feel healthier and better.
The recipe is clear but the cooking is hard because it covers everything from genes and physics models of cells, tissues and organs, to quantum and Exascale computers that are needed to handle the universe of data. Coveney and Highfield are very careful about the use of AI which they expect to help making sense of all the data that we collect about ourselves, but only under supervision from humans.
Such a compelling dream.
For years now, chief digital officers and CIOs have been dreaming of digital twins for their smart cities, factories, goods. Digitisation, monitoring and predictive analytics are massive building blocks needed for this.
Us, humans, we also need data, wearable and AI before we can have our Virtual Us. We generate a huge digital trail but actually know so little about our biological digital twins, Virtual Us - how we tick from the inside: cells, organs, hormones, brain.
Turns out a handful of Bioengineering Labs have been quietly and painstakingly recreating the biological and cellular processes that make our human bodies in digital format. But so so much more needs to be done.
Fascinating read. Worth the effort.
Highfield and Coveney’s dream of the Virtual You goes one step beyond the precision medicine, which creates your predictions based on your doppelgängers, or others “like you”. They envision a healthcast, just like the earth model, which leads to increasingly accurate Nowcasts in your local area.
Wouldn’t it be clever to fix all your future health mistakes without actually making them? “Don’t eat this stuff - it’ll make you sick! Don’t worry about triathlons, just do yoga and you’ll be fine. And oh, do your bone scans every year instead of never, will you?” That kind of stuff.
So the Virtual You presents a perfect recipe:
Create a 3D model of a cell/tissue/ organ/entire body, work out the underlying theory and a mathematical model that uses a variety of physics and biological processes, add a judicious sprinkling of AI to stir it all up and make sense of data in the multi-scale cauldron. Then display the outcome in an avatar that is your replica, a virtual You that can predict how to make your biological You live and feel healthier and better.
The recipe is clear but the cooking is hard because it covers everything from genes and physics models of cells, tissues and organs, to quantum and Exascale computers that are needed to handle the universe of data. Coveney and Highfield are very careful about the use of AI which they expect to help making sense of all the data that we collect about ourselves, but only under supervision from humans.
Such a compelling dream.
For years now, chief digital officers and CIOs have been dreaming of digital twins for their smart cities, factories, goods. Digitisation, monitoring and predictive analytics are massive building blocks needed for this.
Us, humans, we also need data, wearable and AI before we can have our Virtual Us. We generate a huge digital trail but actually know so little about our biological digital twins, Virtual Us - how we tick from the inside: cells, organs, hormones, brain.
Turns out a handful of Bioengineering Labs have been quietly and painstakingly recreating the biological and cellular processes that make our human bodies in digital format. But so so much more needs to be done.
Fascinating read. Worth the effort.
August 19, 2023
I saw this on my educator list through libro.fm and snagged it, because I'm intrigued by the topic. I work with a lot of scientists who talk about "digital twins", and I find the concept super fascinating but don't totally understand how it work (big picture, yes, but like, how does it actually work?).
This book is not for the feint of heart, as it is basically a 400-level (or higher) university course. So much math. So much computer science. Quite a bit of medical science. I understood all the words in terms of I knew that they were words and loosely what they meant, but I did not understand a lot of it in terms of the context in which they were used.
It is definitely not a book for the layperson -- it seems more like a book written for industry peers. I took differential and integral calc in college and PhD level statistics, and most of the math (and associated computer science) in the book was way above my head. But, super interesting nonetheless, and I enjoyed learning about it. I'm not sure how much I retained or how useful it will be, but I do feel like I'll be better prepared to understand the AI engineers and developers at work when they talk about digital twins.
Thanks libro.fm and the publisher for the complimentary audiobook.
This book is not for the feint of heart, as it is basically a 400-level (or higher) university course. So much math. So much computer science. Quite a bit of medical science. I understood all the words in terms of I knew that they were words and loosely what they meant, but I did not understand a lot of it in terms of the context in which they were used.
It is definitely not a book for the layperson -- it seems more like a book written for industry peers. I took differential and integral calc in college and PhD level statistics, and most of the math (and associated computer science) in the book was way above my head. But, super interesting nonetheless, and I enjoyed learning about it. I'm not sure how much I retained or how useful it will be, but I do feel like I'll be better prepared to understand the AI engineers and developers at work when they talk about digital twins.
Thanks libro.fm and the publisher for the complimentary audiobook.
May 3, 2023
3.5 stars...
A REALLY interesting premise: the use of a virtual/digital "twin" that is customized to mirror an individual in order to diagnose and treat medical conditions. This book considers how far/close this might be and what would be necessary to get there. The underlying premise is that maybe biology has done less to evolve theoretically than other sciences and that this will be necessary for the next step in biomedicine.
Wow. They did NOT "dumb it down" for a layperson. Or maybe they did and it's still over my head. I feel like I have a pretty good foundation in understanding science (and general, non-theoretical math), and most of this felt overwhelming. I thought it would be mostly a consideration of biology, where I have more comfort than some others, and it perhaps showed the necessity of having far more depth in all of the areas in order for things to work because ALL of the other things were absolutely foundational to building the case for how this could work (computer science, complex math, physics, chemistry... and more, I'm sure, that I'm forgetting).
I suppose I got the gist of most of what was there, but I certainly couldn't think through the details of how and why because the language of differential (and partial differential) equations and other things are no part of my natural language. The author works to provide some basic groundwork for these ideas, but the concepts are so foundational that I could only get a general sense rather than a good understanding.
But I was definitely interested in the possibilities. I hope people who have more facility in all of the necessary areas will read this and get more out of it than I did. And I hope that young people going into any/all of these fields will read this as a foundation for where the future might be going (and the skills/understandings/interrelationships that might be necessary in the future of these fields). Even if it is beyond total comprehension, it might be worthwhile in sparking desire to build these skills and understandings.
A REALLY interesting premise: the use of a virtual/digital "twin" that is customized to mirror an individual in order to diagnose and treat medical conditions. This book considers how far/close this might be and what would be necessary to get there. The underlying premise is that maybe biology has done less to evolve theoretically than other sciences and that this will be necessary for the next step in biomedicine.
Wow. They did NOT "dumb it down" for a layperson. Or maybe they did and it's still over my head. I feel like I have a pretty good foundation in understanding science (and general, non-theoretical math), and most of this felt overwhelming. I thought it would be mostly a consideration of biology, where I have more comfort than some others, and it perhaps showed the necessity of having far more depth in all of the areas in order for things to work because ALL of the other things were absolutely foundational to building the case for how this could work (computer science, complex math, physics, chemistry... and more, I'm sure, that I'm forgetting).
I suppose I got the gist of most of what was there, but I certainly couldn't think through the details of how and why because the language of differential (and partial differential) equations and other things are no part of my natural language. The author works to provide some basic groundwork for these ideas, but the concepts are so foundational that I could only get a general sense rather than a good understanding.
But I was definitely interested in the possibilities. I hope people who have more facility in all of the necessary areas will read this and get more out of it than I did. And I hope that young people going into any/all of these fields will read this as a foundation for where the future might be going (and the skills/understandings/interrelationships that might be necessary in the future of these fields). Even if it is beyond total comprehension, it might be worthwhile in sparking desire to build these skills and understandings.
May 28, 2023
While it was interesting and informative, I felt that this book did not really sound like the summary. If you don’t have a strong science, background or interest in listening to science, as if it was a college course, I would skip this book.
February 18, 2024
Interesting concept, no overarching story
Very detailed science that will quickly become an outdated read. Poor narrative and a waste of an opportunity to outline the potential and benefits rather and inspire readers.
Very detailed science that will quickly become an outdated read. Poor narrative and a waste of an opportunity to outline the potential and benefits rather and inspire readers.
April 14, 2023
Clear, detailed, visionary
Displaying 1 - 8 of 8 reviews