Reach 2024: How do we create a google map of the human body?
By Jill Langlois
Illustration by Julia Schwarz
Researchers in the CIFAR MacMillan Multiscale Human program are on a mission to create a map of the human body, measuring across scales and time. How will they do it? CIFAR is convening experts across diverse fields of human health research to take on this unprecedented challenge.
Imagine having a roadmap of your body. Like the ones generated by Google, it would trace the best paths between one place and the next, taking into consideration the current circumstances. It would guide you around traffic accidents and construction sites and warn you of speed traps ahead.
But instead of linking geo-spatial landmarks, cities and countries with walking, driving and public transit routes, it would show you how each of the systems that make up your body are connected and what impacts they have on one another. Based on a universal encyclopedia of the human body, this roadmap would create the scaffolding for your digital twin, a computational predictive model of your body that would accompany you throughout your life and over time, become more attuned to the specifics of your genetics, your environment and your choices.
It would help health care professionals to better understand things like how your gut health is impacting your brain health, what the best treatment is for the cancer you’ve been diagnosed with and what your risk is for getting Alzheimer’s disease.
It might sound like science fiction, but this biological roadmap could soon exist. Much like the intricate interdependencies of the human body, research discoveries involve a complex web of investments and inputs — and thanks to the generous support of the MacMillan Family Foundation, we are closer to those breakthroughs becoming reality.
The program has so far brought together 16 researchers from around the world in an unprecedented effort to use their vast array of expertise to better understand the human body, all of the systems that make it work and how those systems, which exist on several different scales, are connected to bodily functions, such as digestion, respiration and cognition.
A revolutionary tool for medicine, this overarching map of the human body, which would then be tailored to each person’s specifications, could be the difference between just living and living well.
The human body operates at many different scales. Spatially, it functions at the nanometre with genes and molecules, at the micrometre with cells, and from millimetres to metres with tissues and organs. Temporally, the body can see changes as quick as an enzyme catalysis (which happens at a fraction of a second), and as slow as aging. And while researchers have studied each of these scales separately, there is still a significant gap in our understanding of how they are connected and what impacts they have on each other.
“In traditional research, we’re typically able to study one or two of those scales or layers at a time,” says Aviv Regev, a Fellow in the CIFAR MacMillan Multiscale Human program. “But in reality, they are intricately linked and dependent.
“For example, given a gene of interest, can we understand its impact on the cell, the tissue — the whole body — and how it might behave differently depending on your ancestry, environment or age?” asks Regev, a computational and systems biologist and the head and executive vice-president of genentech research and early development at Genentech.
“In traditional research, we’re typically able to study one or two of those scales or layers at a time. For example, given a gene of interest, can we understand its impact on the cell, the tissue — the whole body — and how it might behave differently depending on your ancestry, environment or age?”— Aviv Regev
Questions like the one posed by Regev — founding co-chair of a similar project to map all human cells called the Human Cell Atlas — are the kinds the program hopes to answer. Launched in April 2023 following CIFAR’s Global Call for Ideas, the program brings together researchers with varying areas of expertise, including genomics, cell and developmental biology, neuro-science, medicine, mathematical modelling, engineering, data science and AI.
Together, they plan to integrate data from various spatial and temporal scales across the global population and over time, mapping the human body’s systems and connections so that a 3D model can be made. It's a feat that has never been attempted on this scale before — other mapping projects, like the Human Cell Atlas or the Allen Brain Map, focus on just one level or organ — but one that the project’s team is sure they have the right tools for now.
Sarah Teichmann, who recently moved from the Wellcome Sanger Institute to the Cambridge University Clinical School, is one of the program’s Co-Directors and a co-founder and co-leader of the Human Cell Atlas.
She notes that one of the biggest technological hurdles she expects the team to face is getting the different types of data and the instruments used to measure them to work in sync.
“The challenge for us as a community is to actually achieve that linkage, which is not a small task,” she says. For Ed Lein, a Fellow of the CIFAR MacMillan Multiscale Human program, it’s about breaking up the work into achievable parts.
“It’s not something we can do all at once,” says Lein, a senior investigator at the Allen Institute for Brain Science, an affiliate professor in the departments of neurological surgery and laboratory medicine and pathology at the University of Washington.
“For example, we have groups that are represented in this effort that look at the brain from very different perspectives: from the imaging level, or the cellular level or connections across the brain. And these present opportunities to begin integrating datasets.”
In November 2023, members of the CIFAR MacMillan Multiscale Human program convened for the very first time. The program meeting took place at the Wellcome Sanger Institute in Cambridge, UK, where members shared their perspectives on the complex challenge of developing a multiscale map of the human body. Photo: Paul Fenn
Katy Börner, one of the program’s Co-Directors, the Victor H. Yngve distinguished professor of engineering and information science at Indiana University and the founding director of its Cyberinfrastructure for Network Science Center, likes to use the intestine to explain how connecting data across scales and layers will help build a multiscale 3D human reference atlas.
Like the brain, the intestine has already been studied at different levels, and several maps of it and its layers already exist, including a 1D rendering by the Helmsley Gut Cell Atlas Project and a 3D intestine that is part of the Human Reference Atlas. Computational methods can be used to map data across atlas systems, permitting tissue registered in the 3D reference (as well as associated cell types and biomarker expression values) to be viewed and understood in the 1D reference and vice versa, creating a more complete model of the intestine.
The same can be done for other genes, molecules, cells, tissues and organs in the body. Through the CIFAR MacMillan Multiscale Human program, researchers will build on the mapping they’ve already completed or are working on and collaborate with colleagues in other fields. Some are looking at a different layer of the same part of the body (one might look at the cells that make up the intestine, for example, while another looks at its tissues), while others are studying something completely different but that might have an important link that isn’t yet known (like the connection between gut health and brain health).
It’s like the “Blind Men and an Elephant” parable, says Börner, where each expert sees and understands different pieces of the multiscale human body, but nobody can yet imagine how all those parts deliver healthy human function together.
“We are really experts in our respective areas, but we’re also able to talk to each other and even collaborate with each other,” she says. “I think that’s where science shines. Interdisciplinary collaboration across scales and across institutional boundaries, and using some of the finest and highest quality data and efficient algorithms in existence today.” Lein says that collaboration can go beyond asking program colleagues to find the best way to map the human body — which makes CIFAR the perfect hub for this program.
“Have other fields had this same kind of problem of mapping complexity, like astronomy, perhaps?” he wonders. “Are there lessons from other fields that could accelerate this field? We don’t normally get to talk about these things. Usually, you’re in your own area and you get funding to do specialized projects, but some of these problems are much bigger than that and require a different way of thinking about them.”
“I think that’s where science shines. Interdisciplinary collaboration across scales and across institutional boundaries...” — Katy Börner
Artificial intelligence is another area Lein says could help make mapping the human body a reality, and one that Regev agrees could play an important role in the program.
“The advances of the past decade in machine learning, especially representation learning and generative AI, open an unprecedented opportunity for us,” she says. “Algorithms can now learn from separate experiments, operating at separate scales and with separate modalities and nominal variables how to map or translate one level to another, just like generating an image from a text description or translating between two languages.”
One example of how this works, she says, is when an algorithm is trained to take both an image of a tissue as well as cell profiles from the same specimen. It would then connect them so that it could tell us where the cells and their molecules came from in that tissue, allowing for innumerable medical applications, including better disease diagnosis, identifying and improving drug targets, and the development of new cancer and regenerative medicine therapies.
“There is much more like this that we need to do,” Regev says, “and not just for two levels but for multiple ones.”
Another aspect of the program that is important to its Fellows is ensuring equity. Having data that includes a wide variety of ages, sexes, genders, races and ethnicities is important to building a map that will be accurate for all people who use it, as is including researchers of all types in its construction.
“Much of the data we have in-hand right now doesn’t match the population of the planet,” Börner says, reiterating that the team is already working on including more Asian and Indigenous data. “And it’s not just the data used to construct the atlas. We also need to make sure atlas constructors represent major demographic groups, as they are making the decisions about what atlas we want and how it is built and used.”
“It’s like the eureka moment that every scientist lives for. And being in an interdisciplinary environment increases the chances of finding these solutions.”— Gary Bader
Just getting those map builders together was an obstacle the program had to overcome. Researchers from different disciplines tend to do their work alone and without consulting scientists from other areas, but with the help of CIFAR, that gap has been bridged. And now that they’re working together — they’ve already held two meetings to get the program started and more are already planned, each happening in a different city around the world — the Fellows expect to see many more advances in their individual work as well as the work they’re doing together.
“Interdisciplinary discussions and research are exciting,” says program Co-Director, Gary Bader, who also heads the University of Toronto’s Biological Systems Lab, a computational biology lab focusing on systems biology at the cell and tissue levels.
“The most exciting times in my career that I can think of are when I was just dumped into a room full of people to think about all sorts of different topics. It’s really energizing.”
One of the reasons, he says, is that discussions among colleagues from different disciplines can provide shortcuts to solve complex problems. A researcher in a different field might have already solved a problem that, while it seems different than yours, is essentially the same, just cast in a different way because of its discipline.
“And then you can take that solution from another field and make very rapid progress,” says Bader.
“It’s like the eureka moment that every scientist lives for. And being in an interdisciplinary environment increases the chances of finding these solutions.”
Teichmann remembers having that kind of a-ha moment when one of the program’s advisors — Hiroaki Kitano, CEO of Sony AI, executive vice-president and officer-in-charge for AI collaboration for Sony Corporation and CEO of Sony Computer Science Laboratories — presented one of the company’s artificial reality (AR) tools to help train surgeons at a program meeting.
“It was a completely unexpected potential application that had never even crossed my mind,” she says. “Having this community of Fellows and also these scientific advisory board members can lead to some really unexpected interactions and lightbulbs going on that you never thought would.”
While companies working to create these types of tools could benefit from this project, Teichmann notes that they could also benefit from those tools if they could use them to test their data to better frame them for success.
It’s that type of sharing across disciplines and institutions that Lein hopes will be a throughline for the CIFAR MacMillan Multiscale Human program.
“I think that’s really the value of this: bridging across fields to bring in new ideas of how this can be done,” he says. “I would like to see these integration efforts produce really transformative outcomes that the whole field can benefit from. To do that you have to engage different communities across the whole field at large. In CIFAR there are already people who are studying the same organs, but in very different ways, and they would never live in the same consortium. But now we can all interact and learn from one another and see each other’s perspectives and see how what we’re doing can be mutually informative if we put things together.” If that happens, the health applications could be boundless.
“By expanding our lens to understanding the links between biology at multiple levels,” says Regev, “we’ll gain powerful new approaches in our quest to turn biological clues, from any one level, into a fuller picture of how the human body works.”
Related Articles
-
Ross Mitchell, Amii: AI note-taking for doctors
November 16, 2024
-
Data Communities for Inclusion: Recentering technology in the knowledge and well-being of the collective
November 14, 2024
-
Former CIFAR Fellow Paul Hoffman awarded the 2024 Kyoto Prize for Groundbreaking Work in Earth Sciences
November 12, 2024
-
Government of Canada announces Canadian AI Safety Institute
November 12, 2024