A 3 Billion Dollar Mistake: Why the American government should think twice about a Brain Activity Map (BAM)
Update: The United States government has released more information about the specifics of their brain mapping project, now called the BRAIN Initiative. I break down the details and discuss the pros and cons here.
By Gabrielle Rabinowitz, @GabrielleRab
In the 2013 State of the Union address, President Obama praised American scientists for developing drugs, engineering new materials, and “mapping the human brain.” This scientific shout out was not just a pat on the back for American researchers. Rather, it was a veiled reference to a new multi-billion dollar research initiative planned by the Obama administration and the National Institutes of Health (NIH). So what is this plan? As the director of the NIH, Francis S. Collins tweeted:
— Francis S. Collins (@NIHDirector) February 13, 2013
The Brain Activity Map (BAM) is a project that will bring together federal agencies, neuroscientists, and private research foundations to create a functional map of every connection in the human brain – numbering in hundreds of trillions. Obama plans to spend $300 million each year on this project over the next 10 years. The researchers and investors involved hope that BAM will drive advances in artificial intelligence, help us understand diseases like Alzheimer’s and schizophrenia, and stimulate the American economy.
Sounds too good to be true, right?
Redundancy and confusion
When I began researching this article, I quickly found bold announcements of an exciting multi-million dollar brain-mapping project, as well as skeptical critiques of the expensive proposition… but they were all from 2011. It became clear that these articles did not refer to the new “Brain Activity Map” proposal, but to an existing multi-million dollar NIH-funded Human Connectome Project launched in 2010.
So what gives?
While Story C. Landis, the director of the NIH’s Institute of Neurological Disorders and Stroke (NINDS), initially shared my confusion over the announcement, she now acknowledges that the BAM project will piggyback on the existing Human Connectome Project. But what is the human connectome, why is it useful, and what makes the BAM proposal different?
Mapping the brain on many scales, in many animals
The “connectome” is the complete map of connectivity in a brain- but what does that really mean? There are many types of connections and scientists can look at the connectome on three different levels:
The Microscale: Connections between individual nerve cells (neurons).
The Mesoscale: General area of connection between groups of similar neurons (individual neurons cannot be distinguished).
The Macroscale: Overall distribution of nerve bundles made up of many different kinds of neurons (actual connections are hard to identify).
Scientists are making good progress on the microscale connectome of the nearly 100,000 neurons in the fly brain and others have begun mapping the mouse connectome on both the mesoscale and the microscale.
The only complete microscale connectome was mapped back in 1986: the nervous system of a nematode worm called C. elegans. This worm has a total of 302 neurons, which make only 7,000 connections – far fewer than the trillions in our own brain. So can we predict or simulate a worm’s actions as the BAM proposal suggests we could?
In a word… No.
In spite of the beautiful wiring map, much of the function of the C. elegans nervous system remains a mystery. As Dr. Cori Bargmann, an award winning researcher of these worms and their neurons, tweeted:
@mbeisen Nothing wrong with starting with a connectome — that’s why I went to worms. It poses questions, but doesn’t answer (many of) them.
— Cori Bargmann (@betenoire1) February 7, 2013
One of the biggest problems with “structural connectomes” like the map of the C. elegans nervous system is that they do not account for changes in connectivity. Because neurons are always reacting and adapting, the connectome is not a definitive representation of what is actually happening. It is more of a collection of all possible firing pathways in a brain.
Keeping in mind this caveat, the scientists behind the BAM proposal are hoping to go beyond predicting all possible outcomes. They are working toward showing the actual outcome of neuronal connectivity. But is this really possible?
The Human Brain Activity Map: “Functional Connectome” or Foolish Claim?
The New York Times describes the new BAM project as “a decade-long scientific effort to examine the workings of the human brain and build a comprehensive map of its activity.” The key word here is “activity.” According to the proposed plan, this new brain map will be more than just a connectome. It’s going to be Connectome 2.0: a “dynamical mapping of the ‘functional connectome,’ the patterns and sequences of neuronal firing by all neurons.”
This sounds great. Mapping both the physical connections of neurons and their dynamic activity could drastically improve our understanding of the brain. Unfortunately, there are at least 3 big problems associated with this plan.
Problem 1: Choosing a brain
The authors of the BAM paper write, “For midterm goals (10 years), one could image the entire Drosophila brain (135,000 neurons), the Central Nervous System of the zebrafish (∼1 million neurons), or an entire mouse retina or hippocampus, all under a million neurons.” But Obama and the New York Times have not been heralding the complete mapping of the zebrafish central nervous system.* They’re promising us a human brain activity map. In ten years. Even the extremely ambitious authors of the BAM paper won’t promise that. They say that they “do not exclude the extension of the BAM Project to humans” and will not even try to give a timeframe for such an extension. People expecting a human brain activity map in 10 years are going to be sorely disappointed.
Problem 2: Scale
Remember the Human Connectome Project- the structural connectome on which BAM is supposed to piggyback? That’s a macroscale project. It looks indirectly at mixed bundles of neurons in living human brains. And BAM? Microscale. How will the researchers know from which neurons they’re recording? Will they build up their own microscale connectome from scratch? This is a daunting challenge. Even if they restrict themselves to animal models, the fly microconnectome is still a work in progress.
Problem 3: Technology
Another huge problem for BAM is that the technology to measure every firing of every neurons in the brain doesn’t actually exist. The Neuron paper that inspired the BAM proposal lists a host of “novel methods” (translation- no one’s tried them yet), but does not include any examples of those techniques in practice. Story Landis has said that one of the goals of BAM is to “develop the tools.” Ten years is a very short time to both develop new technologies and apply them.
Flaws in Funding
Back in April, Dr. Cori Bargmann tweeted:
#brainbrawl Ultimate connectome Q is about resources. Fund Seung, Denk, Lichtman? Definite yes – smart, creative. $1B scaleup? Um, vote no.
— Cori Bargmann (@betenoire1) April 3, 2012
She was referring to a different proposal to fund connectome research, but the point still stands. Why would a connectome researcher refuse a billion dollar budget increase? Well, the money has to come from somewhere. In light of recent sequestration threats, it doesn’t seem like the US government is going to be increasing the NIH budget any time soon. If we manage to avoid the fiscal cliff and the budget stays flat, the 3 billion dollars going towards BAM is going to be coming out of the pocket of other scientists.
Dr. Leslie Vosshall, a behavioral neuroscientist at The Rockefeller University, tweeted about the issue last week:
$300 million for Brain Activity Map Project. With flat NIH budget: 750 investigator-initiated RO1s lost per year nyti.ms/XkeczY
— Leslie Vosshall (@pollyp1) February 18, 2013
The R01 is one of the most popular NIH grants for biology researchers in America. By Dr. Vosshall’s estimate, every 300 million dollars that goes to fund a year of BAM would take away funding from 750 American labs. The funneling of government funds into individual massive projects like BAM is a worrisome trend. BAM is not the first massive government research initiative.
Professor Michael Eisen served as an advisor for ENCODE, a recently completed effort to analyze so called “junk DNA” (parts of the human genome that don’t contain genes). Reflecting on the multi-million dollar project in a recent blog post, Dr. Eisen wrote:
American biology research achieved greatness because we encouraged individual scientists to pursue the questions that intrigued them and the NIH, NSF and other agencies gave them the resources to do so.
Dr. Eisen argues that this greatness is threatened by an increasing emphasis on “Big Science” instead of investigator-driven research. In addition to depriving individual labs of funding, Big Science projects like ENCODE or BAM dominate the research landscape for years, establishing a current of trendiness and hype that is hard to swim against. Researchers are often pressured by funding agencies to base their work on the huge data sets that such Big Science projects generate.
The Takeaway: BAM and hype
Here on Behind the Buzz, I keep a lookout for media hype that gets in the way of real science communication. And when it comes to hype, Big Science proponents are often the worst offenders. The increasing sensationalism over the recent BAM announcement is just a taste of what’s to come. And after hype comes inevitable disappointment. The resulting damage to the public trust may be worse than any technological failures or financial costs. Again, we can learn from ENCODE. In recent weeks the scientific community has been in an uproar over an extremely critical paper from Dr. Dan Grauer which trashes the ENCODE project. Proponents and detractors have both turned to hype, derailing the scientific discourse.
To avoid another round of media backlash policy makers need to seriously reconsider the BAM project. This highly speculative, technologically ambitious, and hugely expensive project is the last thing American science needs today.
*Since I initially wrote this post the New York Times has published a more measured article in which they acknowledge that BAM will not be mapping the human brain any time soon.
These views are the work of individual authors, do not necessarily represent the views and opinions of The Rockefeller University, and are not approved or endorsed by The Rockefeller University.