Meet Harvey and Harvetta: The virtual humans shaking up personalization
04 Jun 2020 --- An international team of researchers has developed whole-body computational models to further propel research into personalized medicine, including the role of diet on the microbiome. Named Harvey and Harvetta, the virtual humans successfully predict known biomarkers of inherited metabolic diseases and enable the exploration of potential metabolic interactions between humans and their gut microbiomes at a personal level. They are touted as enabling the creation of personalized whole-body metabolic models using an individual’s physiological, genomic, biochemical and microbiome data.
“Harvey and Harvetta are available to the research community and can also be used by industry. Our interest is to enable others to apply these models broadly for a wide range of biomedical applications. Concerning industry, I think that the models will be of particular interest for the nutrition-oriented space,” Professor Ines Thiele, based at the National University of Ireland, Galway, and leader of the project, tells NutritionInsight. 
Her team of researchers developed the first whole-body, sex-specific, organ-resolved computational models of human metabolism, which mechanistically connect anatomy and physiology with molecular level metabolic processes.
The virtual male and female human metabolic models are built from literature and data on human metabolism, anatomy and physiology, as well as biochemical, metabolomic and proteomic data. They are anatomically interconnected as whole-body metabolic models, comprising more than 80,000 biochemical reactions distributed over 26 organs and six types of blood cells. Notably, these models can also be expanded to include gut microbial metabolism. The researchers provide a simulation toolbox that enables others to repeat this work and to use it as a starting point for their research questions.
The start of a personalized journey
Thiele defines personalized medicine as an emerging research area that focuses on improving disease prevention, diagnosis, treatment and care through a better understanding of genetics, biological mechanisms and interactions with the environment.
“Personalized medicine represents a paradigm shift, moving away from the one-drug-fits-all approach by accounting for inter-individual variations. Compared to the genetic mark-up, other factors influencing drug response, such as co-medication, diet and gut microbiota, can be modulated. Hence, the complexity of possible interactions requires sophisticated, innovative approaches, which include predictive computer models and ultimately ‘virtual humans,’” she says.
According to Prof. Thiele, Harvey and Harvetta provide a starting point for personalized medicine. Already, the models can be used to generate hypotheses on human metabolism and host-microbe co-metabolism. Importantly, biochemical and phsyological data of an individual can be used to personalize these models and to predict personalized metabolic features. These hypotheses and features need to be tested experimentally, which will further improve the predictability of the models.
“In the future, I envision that Harvey and Harvetta could be part of the aforementioned virtual human that would be used to predict the outcome of medical or nutritional treatment approaches in silico, for example. This will identify the best possible treatment option for an individual at a given time,” she adds.
Personalized medicine requires realistic, mechanistic computational models that capture the complexity of the human.A holistic approach
Personalized medicine requires realistic, mechanistic computational models that capture the complexity of the human. It must represent each individual’s physiology, dietary habits, metabolism and microbiomes. Prof. Thiele explains that while algorithms based on artificial intelligence and neural networks are gaining popularity and will undoubtedly have an impact on biomedicine, they do not provide actionable explanations on why correlations and patterns occur.
Projects within molecular biology have made useful discoveries into the “parts list” of human cells, but it remains challenging to integrate these aspects into a virtual whole human body. While the Virtual Human Physiome project had generated comprehensive computational models about the anatomy and physiology of human organs, it did not connect these with molecular-level processes and their underlying networks of genes, proteins and biochemical reactions.
“In contrast, integrating these data with multilevel, hierarchical, mechanistic computational models spanning from molecules to the entire human body allows novel mechanism-based hypotheses to be generated. Such computational models can predict the global, emergent behavior of the biological process under consideration,” Prof. Thiele explains.
“Through the use of computational modeling, systems biomedicine can contribute to the understanding and modulation of developmental and pathological processes in humans and will thus be instrumental in enabling personalized medicine and personalized nutrition,” she adds.
The researchers note that an interdisciplinary effort is required to generate a personalized whole-body metabolic model. Software for constraint-based modeling of high-dimensional biochemical networks, as well as novel algorithms, had to be developed.
“A whole-body model is generated by starting with a set of anatomically interconnected generic reconstructions of human metabolism. This draft model had in excess of 300 thousand dimensions, which was then pared down to approximately 80 thousand organ-specific reactions using efficient algorithms and high-performance computing facilities,” says Assistant Professor Ronan Fleming, a co-author of the study from the Leiden Academic Centre for Drug Research, Leiden University.
The creation of Harvey and Harvetta is just the latest step in the mainstreaming of tailored care. Last week, it was announced that precision nutrition is set to be the guiding force in a new strategic plan from The National Institutes of Health (NIH), part of the US Department of Health and Human Services. Meanwhile, the CEO of Persona Nutrition recently told NutritionInsight that COVID-19 is creating a “tremendous” growth opportunity in personalized nutrition.
By Katherine Durrell
