The third time’s the charm for the Allen Institute for Brain Science’s 3-D atlas of the mouse brain.
Version 3 of the atlas, known as the Allen Mouse Brain Common Coordinate Framework or CCFv3, is the subject of a research paper published today in the journal Cell. It builds on a partial brain map that focused on the mouse cortex and was released in 2016.
Previous versions of the atlas were rendered with lower-resolution 3-D maps. The latest high-resolution maps are fine enough to pinpoint the locations of individual brain cells — which is crucial for interpreting datasets that contain thousands or millions of pieces of information.
“In the old days, people would define different regions of the brain by eye. As we get more and more data, that manual curation doesn’t scale anymore,” Lydia Ng, senior director of technology at the Seattle-based Allen Institute for Brain Science, explained in a news release. “Just as we have a reference genome sequence, you need a reference anatomy.”
To produce the high-resolution atlas, the Allen Institute team created an “average” template from scans of 1,675 mouse brains, and segmented that template into tiny virtual 3-D blocks known as voxels.
The team then assigned the voxels to hundreds of different known regions of the mouse brain, drawing careful borders between distinct areas.
This video moves through the average mouse brain, from front to back. To map the true 3-D geometry of the cortex, a mathematical equation was used to find the shortest curved paths to the brain surface. Some of these paths are shown in the model as multicolored “streamlines.”
The full-brain atlas has been openly available to neuroscientists since late 2017, and it’s already pointed some researchers to new insights.
For example, one recent study observed brain-cell activity as mice chose among the images they saw during a laboratory test. The researchers conducting the study used a system of electrical probes known as Neuropixels to document the activity of hundreds of neurons simultaneously, across several different regions of the brain.
As they analyzed the data they collected, the researchers realized that the activity extended to more brain regions than they expected. CCFv3 helped them make sense of the big picture.
“The atlas was a really necessary resource that enabled the very idea of doing studies at the brain-wide level,” said Nick Steinmetz, a University of Washington assistant professor who’s also affiliated with the Allen Institute. “When you’re recording from hundreds of sites across the brain, that introduces a new scale of investigation. You have to have a bigger view of where all the recording sites are, and the CCF is what made that possible.”
Future versions of the atlas will almost certainly rely on artificial intelligence tools and automation rather than the laborious process that produced CCFv3.
“As we know now, atlases should be evolving and living resources, because as we learn more about how the brain is organized, we will need to make updates,” said Julie Harris, the institute’s associate director of neuroanatomy. “Building atlases in an automatic, unbiased way is where the field is likely moving.”
Ng and Harris are senior authors of the report published by Cell, “The Allen Mouse Brain Common Coordinate Framework: A 3D Reference Atlas.” Other authors include Quanxin Wang, Song-Lin Ding, Yang Li, Josh Royall, David Feng, Phil Lesnar, Nile Graddis, Maitham Naeemi, Benjamin Facer, Anh Ho, Tim Dolbeare, Brandon Blanchard, Nick Dee, Wayne Wakeman, Karla Hirokawa, Aaron Szafer, Susan Sunkin, Seung Wook Oh, Amy Bernard, John W. Phillips, Michael Hawrylycz, Christof Koch and Hongkui Zeng. The Allen Institute for Brain Science is a division of the Allen Institute, founded by the late Microsoft co-founder Paul Allen.