Before an important meeting or when a big decision needs to be made, we often mentally run through various scenarios before settling on the best course of action. But when it comes to our health – be it choosing a treatment for an ailment or even selecting a dietary regimen – it is a lot harder to predict how each choice will affect our bodies and whether it will suit us personally. Recently, researchers from Prof. Eran Segal’s laboratory at the Weizmann Institute of Science have harnessed artificial intelligence to create a personalized “digital twin” that allows them to detect a risk of developing diseases, initiate preventive treatment and even run simulations to predict which treatment will be most effective. This new development, detailed in Nature Medicine, was made possible by the Human Phenotype Project, in which scientists involved in the initiative, along with colleagues worldwide, have collected extensive, in-depth medical information from over 13,000 people.
Before the Human Genome Project was launched in 1990 to explore the fundamental question of what makes each of us who we are, only a fraction of human genes were known to science. The project led to the identification of tens of thousands of genes that shape our traits, and it revealed the genetic basis of numerous diseases. Today, however, it is clear that genes alone provide only a partial picture. Many of the characteristics that define us and the diseases that threaten us are linked to environmental factors, the community of microorganisms residing in our bodies (our microbiome), the aging process and other factors. In an effort to attain a broader picture, Segal, from Weizmann’s Computer Science and Applied Mathematics Department, launched the Human Phenotype Project in 2018. This project tracks thousands of participants who undergo extensive medical assessments and testing every two years over a 25-year period. These evaluations cover 17 different body systems and include a wide array of tests, such as body measurements, nutritional logs, ultrasounds, bone mineral density tests, voice recordings, home sleep tests, continuous glucose monitoring over two-week periods, gene sequencing, cellular protein analysis and microbiome analysis of samples from the gut, vagina and oral cavity.
“”While men’s biological age generally increases relatively linearly, we observe an acceleration in women’s biological aging during their fifth decade of life”
“When we launched the project in Israel in 2018, our initial goal was 10,000 participants,” Segal says. “Since then, more than 30,000 people have signed up, and we hope to reach 100,000 in the future. To deepen our understanding of ethnic, environmental and cultural variations, we set up a branch in Japan and are currently finalizing the establishment of another in the United Arab Emirates, in collaboration with Professor Eric Xing from the Mohamed bin Zayed University of Artificial Intelligence. We are also broadening the age range of our participants; initially, we recruited people between 40 and 70 years of age, but now younger and older people are also joining the study. This research has led to the creation of an advanced database that is not only extensive but also represents the most in-depth collection of human data currently in existence. We recognized the importance of sharing this resource with the scientific community and have now made it accessible digitally to research groups worldwide, while maintaining the privacy of the participants. We believe that the data we have compiled will profoundly affect the field of medicine.”
What’s your biological age?
Modern medicine largely relies on conducting tests and comparing the results to the average ranges for a person’s age and sex. However, the underlying health status and the aging process vary considerably among individuals. A research team led by Drs. Lee Reicher and Smadar Shilo from Segal’s lab has developed an AI model that studies typical physiological changes, which occur throughout a person’s lifespan, in 17 human body systems and learns to identify deviations from expected patterns. The model is built on a platform developed by Pheno.AI, a company specializing in AI research for healthcare. “The model assigns scores to each body system and compares these values to the expected values for the participant’s chronological age, sex and body mass index,” explains Segal. “Based on the deviation from these predicted values, the model determines the participant’s biological age. The older the apparent age of a body system, the greater the risk of associated diseases. For instance, by tracking participants’ glucose levels, we determined the normative rate of increase in blood sugar for men and women over the years. Our model detects any deviation from this pattern and thus successfully identifies pre-diabetes in 40 percent of people who were classified as healthy by conventional testing methods.”
The study of biological age has revealed significant differences between the sexes. “While men’s biological age generally increases relatively linearly, we observe an acceleration in women’s biological aging during their fifth decade of life,” Segal notes. “Menopause is a pivotal event in many medical respects, and it appears to reset the biological age clock. For example, we found that a decrease in bone density is more strongly correlated with the time that passed since the onset of menopause than with chronological age. Furthermore, our measurements make it possible to detect the start of menopause early, so that hormonal treatment can be planned accordingly.”
The Human Phenotype Project has also uncovered new avenues for the early diagnosis of a multitude of medical conditions, including breast cancer, inflammatory bowel disease and endometriosis. That’s because these conditions are characterized by a change in the composition of the patient’s microbiome, and this change acts as a unique and identifiable “signature.”
Still, the most significant promise of the Human Phenotype Project lies in its potential to advance personalized or precision medicine. Researchers aim to achieve this through a unified computer model that will integrate all the information collected from each participant in the project, creating a digital twin of that person. This model – currently under development, in a project led by doctoral student Guy Lutsker – will predict what medical events the participant is likely to experience in the future and how best to prevent them. To train the model, the scientists let it study the medical records of each participant and then ask it to make minor predictions. A specific piece of information is withheld each time, and the model is tasked with predicting it based on the existing data. This training approach helps create a generative AI model that can predict medical events and in the future is expected to construct an entire personalized “health trajectory” outlining a person’s future health status years in advance.
The research team has already developed a model that, by analyzing participants’ glucose levels, has successfully predicted not only their future glucose levels but also which pre-diabetic individuals are at the highest risk of developing diabetes within the next two years. Such predictions help prevent the disease, or delay it at an early stage. Moreover, the researchers are already using the digital twin to check which dietary changes or drugs would be most beneficial for each participant. In the future, the model is expected to encompass all the information within the database, enabling it to predict a wide range of medical events and spare patients the often lengthy trial-and-error process of finding the most effective treatment.
“This achievement is primarily made possible by the community of participants in the Human Phenotype Project. It is a dedicated group of individuals committed to advancing medicine and to the continuous monitoring of their health. We are developing an application that will bring all the collected information to the participants’ fingertips and in the future will provide them with a personal ‘health trajectory,’” Segal adds. “We are living in an era of incredibly rapid change. The realms of health and medicine will undergo dramatic transformations in the coming years, becoming increasingly AI-driven. Our project is poised to be a leading global source of information and innovation, and this is all thanks to our participants. I want to take this opportunity to express my sincere gratitude to each and every one of you – your exceptional collaboration is the true driving force behind this revolution in medicine.”