- The Brain Image Library (BIL) is collecting huge amounts of brain data
- Scientists can easily access, visualize, and download the collected data
- BIL expects to acquire 10 petabytes of data within the next five years to further study the brain
The world is currently experiencing a data revolution. At no other point in history have humans been so inundated with information from every corner of our lives. In fact, the world’s population combined generates around 2.5 quintillion bytes of data every day.
While some of that data is social media posts and cat photos, a lot of it is invaluable to science. In fact, a big problem facing researchers today is not a lack of information, but rather a lack of easy access to it.
That’s where the Brain Image Library (BIL) comes in. Alex Ropelewski, BIL’s operations director and director of the Biomedical Applications Group at Pittsburgh Supercomputing Center, spoke with us about the initiative.
“People who are interested in getting very detailed, high-resolution images for particular cell types can come to our library, visualize the data, download it, and use that data,” says Ropelewski.
Through funding from the National Institutes of Health (NIH), this organization has created an impressive collection of brain data. Although the BIL doesn’t yet collect data from human brains, Ropelewski and his colleagues have high hopes for the future.
Follow the fluorescents
It’s important to note that the Brain Image Library isn’t working in isolation. Ropelewski states that they collaborate closely with the Brain Initiative Cell Census Network.
"The goal of that project is to acquire very detailed information about every cell type that’s in the brain," says Ropelewski. “They’re using lots of different types of animals, mostly mice. But there are plans to move on to marmosets, for example, and some folks are starting to look at how you can do human brains as well. This group is generating data and, as part of the overall Brain Initiative Cell Census Network, we are housing all their volumetric data for that project.”
Although there are many ways to study the brain, BIL houses data related to confocal fluorescencemicroscopy. This means using fluorescent dyes to map a brain’s pathways. For instance, a scientist could inject a virus along with a dye into the brain of a mouse. By following the fluorescent agent, the researchers can map where the virus goes.
But storing information on servers doesn’t lead to discovery. The Brain Image Library also works to bring computing resources to the scientists that need them most.
“The Brain Image Library is a national public resource that aims to both store confocal fluorescencemicroscopy data for the brain initiative, as well as to bring computing power to that data so that folks can analyze and mine that data,” says Ropelewski.
One way they provide these resources is via the Compute Through Bridges initiative. Bridges is a high-performance computing (HPC) system designed to support software for “both traditional and non-traditional HPC users.” This allows for users to attempt data analytics, simulation, and a host of other important research activities with information from the Brain Image Library.
Looking to the horizon
The Brain Image Library received a five-year grant from the NIH, and they have big plans for this time frame. For a starter, Ropelewski wants to provide visualization resources to researchers.
“When we initially put our proposal in, we intended to provide visualization services early,” says Ropelewski. “But that was before we had this great collaboration with the Brain Initiative Cell Census Network. So we had to readjust our time schedule. Over the next year we’re going to add visualization capabilities.”
That’s not all, either. Ropelewski states that the organization expects to have over 10 petabytes of high-resolution data available to users. What’s more, the Brain Image Library is enhancing its infrastructure to be able to securely handle restricted human data in the future.
Although this project is still growing and developing, it has great potential. Earlier this year, the BIL was named the ‘Best Use of HPC in Life Sciences’ by HPCwire. And with the next release at the end of April, the BIL will contain images of over 750 mouse brains.
“I’m interested in it because it’s a new area, and I also see the potential to link together the different data types,” says Ropelewski. “The work that the Brain Initiative Cell Census Network is doing, for example, on single cell sequencing, being able to apply that data to volumetric data, to learn new things that we haven’t learned thus far, really is exciting. That’s really what’s in it for me. There’s a lot of really interesting things that can be done once a scientist has this data in hand.”