Squint your eyes so tight that they’re almost shut. Now keep them that way as you look at the world. You can’t see its beauty in detail. Leaves on the trees, constellations in the night sky, the words on this screen — they’re all a blur.
In a sense, biologists have spent the last century “squinting” at human cells — observing their size, shape, structure, and basic functions but missing vital details about their molecular identities. The time has come to fully open our eyes. An international project called the Human Cell Atlas, which we co-chair, aims to help do just that. A comprehensive atlas of all human cell types would define all the cells in the human body based on the molecules they produce. On Thursday in Stockholm, nearly 200 scientists and collaborators from around the world are gathering for the third meeting to coordinate this global effort.
DNA sequencing has shown us how genes vary from person to person, and how these variants track with diseases. RNA sequencing has revealed where in the body different genes are most active. But we’ve largely applied these two techniques to tissues we mash up before sequencing, losing many of the defining characteristics of individual cells along the way.
New techniques for profiling individual cells have emerged in the last few years, but they haven’t yet been applied in a comprehensive and high-throughput fashion. As a result, we have barely begun to identify all the different kinds of cells in the human body — and there may be hundreds we have yet to describe.
As brain maps proliferate, scientists aim to sync them up
This isn’t just a problem that matters to biologists. It should matter to everyone. Without a deep understanding of what our cells are and how they organize themselves to keep us healthy, we can’t understand how they go awry in disease. No wonder two out of three drugs fail when tested in humans — we don’t understand the cells the drugs are acting on.
Last fall, we and a group of colleagues launched the international Human Cell Atlas consortium. It will build a sort of “Google Maps” of the human body: Instead of geographical features such as continents, countries, cities, streets, and houses, this map of the human body will zoom in on molecular and organizational features of organs, tissues, and cells.
Three breakthroughs have made it possible to realize such an ambitious goal: the invention of technologies to create molecular profiles of individual cells; the development of machine learning and related techniques for parsing huge datasets; and the rise of collaborative science across nations and disciplines.
Single-cell genomic measurement techniques now let us look inside individual cells and gather information on what their DNA says, how it is packaged, what it produces, and how these shift as cells respond to changes in their environments. These measurements can now be quickly made in thousands of cells at a time, for a surprisingly low cost per cell. Emerging techniques will even let us measure cells when they are embedded in their native tissues.
The amount of information produced by these techniques is enormous; making sense of it is an equally enormous challenge. Luckily, we live in the age of big data, and machine learning techniques can now be brought to bear on biological information. To use them, the Human Cell Atlas consortium is planning to build a cloud-based, open-source platform that will make it possible for scientists around the world to upload information, perform joint analyses, and compare cells of many types to gain insights into health and disease. At the meeting in Stockholm, the Chan Zuckerberg Initiative announced it will fund the development of this data coordination platform and collaborate with three of the Human Cell Atlas partner institutions to make it a reality.
A project of this scope requires a global effort. Thanks to the Human Genome Project, which kicked off an era of deep collaboration in biology, most scientists no longer want to work alone — nor can they if they want to accomplish their goals! About 500 scientists from about 20 countries are already on board with the Human Cell Atlas project, participating in meetings, online discussions, and collaborative project planning. We expect (and encourage) others to join in.
Sign up for our First Opinion newsletter
The consortium has launched pilot projects to build exhaustive cellular maps of cells from the immune system, brain and nervous system, skin, and various types of tumors. Early work is already yielding important findings. For example, a recent single-cell genomics study of the immune system led to a surprising reclassification of known immune cells and the discovery of new ones, with implications for the way that immunotherapies for cancer should be designed.
Beginning in ancient Greece, the early anatomists, often working in secret, created maps of the human body that showed the physical — and sometimes physiological — relationships between organs and tissues. Working completely in the public eye, the Human Cell Atlas project aims to usher in a new era of high-resolution human anatomy that will be able to interpret our growing wealth of genomic data, view human biology directly (instead of through the lens of cell culture), and decipher the regulatory codes that make one cell type different from another, let them interact with each other, and maintain their identities. The Human Cell Atlas will also inspire the development of new drugs, let us be smarter in our use of existing drugs, and improve the diagnosis of disease.
Working together towards this ambitious goal will be an exciting journey for the international scientific community over the next decade.
We don’t expect the Human Cell Atlas to “solve” all of biology — in science, every conclusion inspires more questions. We do, however, expect it to transform fields and benefit scientists and medical doctors around the world.