Is free-will an illusion?

America’s BRAIN initiative
The observer corps
What is the way best to study the brain? Big labs or small?
The Economist
Nov 14th 2015 | Argonne, Illinois

IF YOU want to build an atom bomb, land men on the Moon or work out the exact order of the 3 billion chemical “letters” in the DNA of the human genome, then Big Science, a large-scale project backed by a budget in the billions, will do it for you. But will it also do it for the brain, the understanding of which is, perhaps, the biggest scientific challenge of all? That is a question with particular salience now, as the Brain Research through Advancing Innovative Neurotechnologies (BRAIN) initiative gets going in earnest.

At the moment this endeavour, announced by Barack Obama in April 2013, is definitely small science—despite its proposed budget of $4.8 billion over the next decade. According to Rafael Yuste of Columbia University, who was one of its instigators, some 125 laboratories around America have signed up to dip their bread in the gravy, so that gravy is very thinly spread. This reflects the way neuroscience has been done until now, but Dr Yuste would like the future to be different. Genetics was similarly fragmented before the Human Genome Project began, but the bargain imposed on researchers by that project was that the process of sequencing would be centralised into a few laboratories, where it could be industrialised. This, as Dr Yuste and several of his co-sponsors of the original idea of BRAIN write this month in Neuron, is the model they think neuroscience should now adopt.

BRAIN’s specific remit, they point out, is to develop new technological approaches for the study of brains. For that purpose, they propose creating a handful of “brain observatories” which would design and build the equipment, and then make it available to teams of outside researchers, much as academic astronomers book (and pay for) time on large telescopes. In this way, the observatories could concentrate on the engineering: refining the equipment by experience and arriving at designs that might then be commercialised.

Manhattan-project transfer

Dr Yuste and his colleagues identify four types of technology that BRAIN needs. Three are improvements of existing methods. The fourth is largely terra incognita.

The first of the improvements is better computing power. Even the task of mapping a mouse brain will require 500 petabytes of data storage. A petabyte is 1m gigabytes. For comparison, finding the Higgs boson required about 200 petabytes. A human brain is vastly more complex than a mouse’s. It has around 86 billion neurons, compared with 71m in a mouse. And the wiring that links these neurons (cell protrusions called axons) is reckoned to be about 100,000km long.

The second improvement that is required of existing technology involves brain-scanning. At the moment, this has a resolution measured at best in millimetres, which is fine for its original purpose—clinical diagnosis—but useless for understanding how brains work at anything more than a superficial level. Higher-resolution scanners would permit researchers to look at small, functional groups of neurons, such as the neuronal columns of which brain cortexes are composed.

The third improvement Dr Yuste desires is to instruments that can study the connections between individual neurons, a field known as connectomics. At the moment, this can be done only to dead brains, and the tools usually employed are electron microscopes. As with many other fields, the more closely you want to look, the bigger the machines you need to do it. Dr Yuste and his colleagues would also like to overlay on the three-dimensional electron-micrographs of connectomics information from other machines, such as high-resolution fluorescent microscopes, which can detect the molecules present in particular parts of neurons.

That leads to the unknown land. This is the designing of nanoscale devices (ie, instruments whose dimensions are measured in billionths of a metre) which can do something approaching connectomics on living brains, by studying the activities of hundreds or thousands of interconnected neurons simultaneously. Only in this way can the true nature of the brain’s circuitry be understood.

Asking what these nanoscale devices will look like tends to generate a lot of handwaving involving words like “quantum dots” and “nanodiamonds”—both types of tiny crystal that might act as neuron-scale probes—but with little clear idea of how these crystal probes would be controlled and listened to. In their case, then, there is an argument for letting a thousand flowers bloom in laboratories large and small. But for the other devices on Dr Yuste’s shopping list, observatories do seem a good idea. And, as luck would have it, America has a network of institutions that might be pressed into service as such.

Critical mass

The National Laboratories, are, in many ways, Big Science personified. A lot of them started life as atomic-weapons establishments, so they know how to run large projects that go on for years. They also house five of the world’s ten fastest supercomputers. One, indeed, is ready and eager to go. Argonne, near Chicago, has been hiring out one of its instruments, an X-ray generator called the Advanced Photon Source, to biologists for several years. Last year, 2,000 of the 5,000 experiments conducted using it were biological, and Argonne has recently hired a neuroscientist specifically to run a project that will employ the photon source to X-ray mouse brains, with a view to mapping them at high resolution. It is also buying a top-of-the-range electron microscope, with which researchers will be able to zoom in on areas of the brain the photon source suggests are particularly interesting, and a flashy, new computer that will help analyse the results of all this. Argonne’s director, Peter Littlewood, thus hopes to create a model which other national laboratories thinking of developing brain observatories might follow.

Whether they do will depend a lot on the say-so of Ernest Moniz, America’s energy secretary, who is the labs’ collective supremo. He and Francis Collins, head of the country’s National Institutes of Health (and thus, in effect, BRAIN’s treasurer), have met several times to chew the matter over. Some in Congress think the energy department should be minding its own business. Dr Moniz, an astute politician, understands this. But he is also a scientist to his fingertips, for he was once head of the physics department at the Massachusetts Institute of Technology. That he will turn down the chance to be part of what may become the scientific adventure of the 21st century seems unlikely.

Yeah, well, good luck with that, eh? In the meantime, I'm gunna just keep on doin what I always do. It's a lot easier, and don't cost ****. I just say I KNOW things about the brain, and that what I know is SCIENCE, by God, see?

Pretty simple, actually.

Indeed, the brain's complexity places it far beyond anything else we ever studied, and thus we must avoid simplistic, reductionist analysis grids.

The brain also consumes about 20% of our energy but accounts for only 2% of our weight. Small piece of meat, big energy intake, umonguous connectivity, and almost "magical" output (the mind etc.)... That makes for an amazing scientific challenge.

.

Same with the question of how it came to be in the first place, eh, Ollie? Strictly random mutations, "directed" by "natural selection," don't take me too far with that question. Aint sayin it's God, or Intelligent Design, but it aint RM and NS that I can put all my faith in, that's for sure. I look at that proposition as being "simplistic" to the point of being wilfully stupid.

Rome wasn't built in a day and nor was the human brain. Evolution has been working on it since the nematods, in a process called "cephalization" and it's one of the rare long term trends of evolution.

The basic idea of being an animal is to be able to move from place to place in search of food, or to avoid becoming food. Mobility defines the animal kingdom. To do that well, you need to know where you are and where you're going, eg where food and predators are, and thus you need senses to collect data and a data processing system. Even amybias do it.

Exactly.

Sorry, the English term is "amoebas". These are the most complex of the unicellular animals. One would think they cannot have a nervous system (nerves are made of many cells called neurons), and yet they kinda do: they rely on intra-cellular network of microfilaments to collect information and to move around by changing shape. These unicellular dudes can move around, and THEREFORE they need to "know" or to "sense" what's around.

Once you realize that life has been refining data management systems since 4 billion years, the emergence of the brain doesn't seem so mysterious anymore.

Its functioning... that's another matter. We don't even know what neurons actually do, beyond "manage and transmit information" (i.e. beyond the most basic principle). So yeah, "big science" and 50 billion dollars may be what it takes to crack it, as the Economist argued. Or maybe not.... :-]

Yeah, I knew what you meant.



Maybe not to you. We know it happened somehow, but how? Natural selection? Totally random mutations?



I have trouble seeing these kinds of claims as being more than hand-waving dismissals of the difficult underlying questions.

I mean how is it that "Evolution" can make, strictly by sheer accident, something of such immense complexity that we have no present capacity to even comprehend it, let alone "create" it?

For all the genuine "explanation" it gives, I could just as well substitute the word "God" for "Evolution." Many do treat "Evolution" (whatever that really means) as being just about as omniscient and omnipotent as God. I aint converted to that faith yet, my own damn self, eh? Certainly not if all evolutionary novelty, structure, order, and suitability depends strictly on "random mutations."

I admire life. Biology studies bring more than their fair their share of awe. I remember my jaw dropping on quote a few occasions: In biochemistry, photosynthesis, the ATP (eng. TPA?) and its uses and derivations, the DNA-RNA-ribosome-protein decoding system of course (that's a beauty) or nitrogen fixation, including in legumes (peas, beans and cie), whose roots harbor symbiotic bacterias that fix the air nitrogen for them. But also the way ants live, flowers and how they make sex, the very genetic principle of sex: mixing her and my genes in a new cell/embryo/individual... Immunity systems, coagulation, the capacity of an organism to repair itself (within limits)...

Does the human brain make the list? Yes... Am i more in awe in front of it than for the rest of the list? No.

It's mostly about sex, for me.

Heat that, homey!

Seriously, it's not about just mutations. That's not the only way to create new genetic information. Sex does it more reliably. To cut a long story short, when you make a baby, you mix half of your chromosomes with half of your partner's the create a totally new whole genome. The various biological "technologies" through which that happens permit a very high degree of interpenetration after even one more heneration. Already in your grand-kids. Your DNA and hers can end up on the same chromosome, mixed up sometimes even inside a gene: Part of the gene - ie the code for one protein- comes from her, part from you, or anyone else in the ancestry.

The possibilities of tampering with the message are enormous, AND BUILT IN THE SYSTEM. Sex is a great way to evolve.

Heh, "evolve" from what to what? Only individuals of the same species can breed with each other. My Pappy had black hair, my Mama blonde. For some reason I just came out with black hair, not hair with every color of the rainbow in it. Go figure, eh?

Recombining EXISTING genes does not create "new" genes. It does not "create new traits" that didn't exist before. It does not create a brain, or any other new organ, body part, etc.

PS: I mean to say "Hear that, homey!" (not heat that).

From what to what is not really important Here. The Point is that sex is a way to recombine and change the genome constantly. The reason it exists pretty much is each and every species, as opposed or in addition to cloning which is far easier to make happen biologically than sex, is certainly that species need to change in order to keep up surviving. Evolution is built in the system. It's not optional. Even species that seem to not have changed forever morphologically are still recombining their genes constantly.

It's true that such a model seems to preclude the emmergence of entirely new organs. But as it turns out, such case are very rare. The first brain appeared among nematods and more convincingly among earthworms. Earthworms have many small brains along their body, all connected, with a slightly bigger one in the head. Every brain coming after that is just an amplification or a viariation on the same theme.

So yes, new organs pose a challenge, (eg mammal's placenta / pregnancy system) but those cases are in fact very rare. It's not like the human brain appeared suddenly out of nowhere.

I know you’re gay, but I’m fine about that… Live and let live, ya know?

Well, Ollie, one theory says that the type of variation which leads to NEW traits, new organs, etc. is due to "random mutation." That's the model I've been questioning.

Yes, and that's unconvincing. But it's important to remember that such events seems to be extremely rare.

OK, we agree on that. And, if I understood you correctly, we agree that genetic recombination resulting from sexual intercourse can't explain it either. What "does" explain it? Well, different theories of evolution have different (at least partial) answers to that question. I won't try to go into them all. But I will note this:

"Natural selection" creates NOTHING. It can only act (to the extent it acts) on what has already been created. It can't "explain" evolution at all if the question is "how does all this variation come about?"

Most of the natural variation CAN be explained by mutation, recombination and selection. But whatever the explanation for the few cases where I find that explanation unconvincing, i bet you it's entirely natural.

Most of the natural variation CAN be explained by mutation, recombination and selection

You add in "selection." Do you disagree with what I said?:



I guess maybe I DID misunderstand what you were saying about recombination and mutation. Are you referring to "random mutation" or some other kind of mutation.?