THE COVID-19 PANDEMIC BROUGHT A VAST MOBILIZATION of scientific ingenuity. At the same time, work on other medical threats continued to push forward. This week, Proto looks at the most transformational non-COVID-19 research of 2020.

Any medication is, in one sense, a molecular shape. The prescribing physician is hoping that it will lock on to complementary shapes in the body and create a positive change. As the body’s own molecular shapes have come into greater focus over the past century, it has enabled researchers to create even more precise medications for a wider range of conditions. Two major breakthroughs on that front in 2020—the first time “seeing” an atom with cryogenic electron microscopy, and a new way to predict how a protein folds—promises to dramatically help the drug development of tomorrow.

The task of delineating the body’s molecules began in the 1930s, in a field advanced early on by Dorothy Crowfoot Hodgkin, who mapped the structures of cholesterol, penicillin and, eventually, insulin. But she and most other scientists used a technique called X-ray crystallography, which requires a compound to be concentrated into a crystal, then bombarded with X-rays. Many molecules, including important proteins on the membranes of cells, important for drug delivery, resist being crystalized.

Cryogenic electron microscopy, or cryo-EM, is a more recent approach that flash freezes its target molecules. That greatly expands the kinds of molecules that can be mapped and gives more information about their possible ranges of motion. While the resolution of this technique lagged behind crystallography for decades, the past year saw two cryo-EM teams—one from the Max Planck Institute in Göttigen, Germany, and the second from the Medical Research Council of Molecular Biology in Cambridge, United Kingdom—achieve so-called atomic resolution, which means that researchers were able to observe proteins down to the atom.

In addition to this feat, the U.K. team achieved a historic resolution in seeing the GABAA receptor, “a major signaling molecule involved in practically all aspects of brain function,” says Radu Aricescu, a structural biologist at the MRC. A decline in GABAA receptor signaling can trigger insomnia, epilepsy and other neurological disorders, and the receptor is a “major drug target,” says Aricescu, whose team published its findings in Nature in November.

GABAA receptor signaling is also crucial in general anesthesia—and that connection has spurred a cross-continental collaboration with Massachusetts General Hospital. “This is the 175th anniversary of the first public demonstration of general anesthesia, which happened at MGH, and my team is now actively working with Keith Miller at the hospital, using the developments in our paper, to better understand how general anesthetics work and how could they be improved further,” Aricescu says.

The second major breakthrough happened when AlphaFold 2, a program developed by Google’s DeepMind, came close to predicting the shape of a protein based only on its sequence of amino acids—information that can often be found in the body from studying DNA. Christian Anfinsen, during his 1972 acceptance speech for the Nobel Prize in Chemistry, theorized that such a feat must be possible, though ever since then, teams have been unsuccessfully chasing the desperately complex mathematics behind protein folding.

An international competition—the biennial Critical Assessment of Protein Structure Prediction—has been the central measure for such efforts and is a forum for teams to try their protein folding models using a standard molecule, usually one whose shape has recently been established by more experimental methods. This year the AlphaFold 2 top model scored 92.5 out of 100, which is competitive with results obtained though cryo-EM and other traditional avenues.

Alone and in combination, these two technologies promise to become core tools in medical research, deciphering the shapes out of which all of us are made.