So, let me go backward a little bit to the beginning, so to speak. And the beginning involves an idea that came actually from the originator of this idea, The Blue Brain project. So, before The Human Brain Project there was the Blue Brain Project, some seven, maybe a little bit more years ago. The idea was two-fold. The idea was to, to, to build a new platform for first databasing the brain. So take the data of the brain. Whatever we know about the brain, databasing the brain. As I said in I told you in the first lecture, when you speak about data about the brain, there are so many levels of data. We can speak about genes associated with brain circuits. We can speak about proteins, we can speak about synapses. About channels, about a, a whole neuron, a, a network, a behavior, all these are data about the brain. So, so, what, what should be a mechanism to systematically database the brain? I should say that we, as a society of neuroscientists did wrong. Along the years, because we did not develop a platform for collecting and preserving data. So much of the data that comes from particular labs, are lost in a sense. They are not completely lost because they are used to generate papers, discussions, and so on. So, so data is available in some sense, but not the data in it's self many times. So one cannot, for example, in Israel, replicate or go into the data of another lab in a direct way, and use it to understand something that comes from your lab, and so. And so we want to database the brain. That's one of the big missions of The Blue Brain Project, and The Human Brain Project, how to build a platform at all levels, from genes to behavior, at all levels that are accessible. For the general of public because this data is mostly interesting for, for neurologists, for medical doctors and for neuroscientists like myself. So this is the issue of the database. Which is not controversial issue. The more controversial aspect of The Blue Brain Project and later of The Human Brain Project. Is the issue of simulation, based on the data. So you already understand what simulation is, because I saw, I showed you the Hodgkin-Huxley, taking data, recording from neurons, from the giant axon of the squid, looking at the data, at their concentration of ions, and recording conductances and so forth. Using the data, to simulate, to build a model, of the spike. Not only writing equations, but simulating the equation to generate a spike, a simulated spike, comparing it to the recorded spike, and so on. Now we are talking about simulation of much larger circuits, many many cells. Not only one cell, not only one spike. And this is controversial, we'll discuss a little bit about that later. And the whole, the whole, the whole field so to speak is called simulation based research. We want to simulate in details, in very fine details, the system you study. And the system is the, a brain, piece of a brain, network. And the question is, can we simulate, is it reasonable to simulate, is it useful to simulate? What can you study? What can you learn via simulation? And so on. So the idea is not only to collect data, but use this data as Hodgkin-Huxley did for the spike, to use this data to generate a realistic simulation of the network you are interested in. Hoping that via this process of from data through simulation. You would get an understanding. So that's the second part of the, of the project of the Blue Brain project. The, the data systematic data. Organization, arrangement, and then simulation based on the data of a particular phenomena, and we shall discuss it later on. In some, some people call this reverse engineering, because you take the system, you decompose it to elements, to spikes, to neurons, to synapses and so on, you decompose it, you break it down to elements. This system already exists. The brain and so this is reverse engineering. Because it was already engineered. Now you reverse engineer by taking it into parts and then rebuild it in the computer. Simulate it in the computer. Copy the system in the computer and then simulate the system. Hoping that you would understand via simulation. So, this is really the philosophy, so to speak, behind the Blue Brain Project. Just to mention that this Blue Brain Project actually, as I said, started by Henry Markram, who you see here is now at EPFL in Lausanne. And he's now also the head of the Human Brain Project, that as I said, emerged from the Blue Brain Project. And I was involved in this, Javier DeFelipe in Madrid. And several other people, and mostly who are using, with the help of IBM, because IBM provided the powerful computer originally. Now, we have already the second generation of this particular computer. The Blue Gene IBM computer. This Blue Gene IBM computer, supercomputer. Very powerful computer. It was used also in other occasions. Eh, for example, eh, in, in, in the, in the Deep Blue game between Kasparov and the computer and also in the Blue-Gene. It's called Blue Gene because that this gene was also, the computer was also used for to sequence the human genome,and so the blue comes, the blue comes from IBM color, the original color of IBM computer was blue. Today it's black, and so the blue comes, from, from IBM. IBM is involved also in this project today, also in other projects like the. Human Genome Project, and also the Blue Brain has a connotation that we like. So this is your brain what develops or generates connotation through learning and plasticity and so you connect the color blue brain to other connotation sexual connotation. And but this was not of course the meaning. The meaning was to use the Blue Gene machine to simulate the brain. And just to mention something that I very, very briefly highlight. we are very lucky to be in a particular time, because we have new methods, both anatomical and electrophysiological methods. To collect data at a single cell level. Synaptic and cell level. So Brainbow I mentioned before a very beautiful a very aesthetic eh, method eh, whereby today. You can really design genetically mouse brain. so that the cells will have. Intrinsically they will have colors. And this makes anatomy much, much easier than this gray matter anatomy because suddenly you start to see a green cell and a red cell and a blue cell and so on with this jungle of cells. And this blue, red and so forth anatomy, this colorful anatomy, this Brainbow anatomy In principle may help you to make some sense of this anatomical jungle because now the, the trees, the trees in the jungle have, have colors and this may help to find out whether the green axon is approaching yes or no the blue, then right yes or no. And this, this helps you just to direct yourself so to speak. Within, within this network. So this, this new anatomical and this is only one of them. Is really extremely helpful for getting data base that is much more systematic and complete. Right of these to the Ramon Y Cajal and Golgi original data which was very sparse as I mentioned in the second talk. We also have the Connectomics approach, which I just mentioned last, in the last, in the last talk. Because this Connectomics which is another controversial project, a very controversial project, because it is very, very labor intensive becuase it is very, very hard to collect this data and reconstruct. Full, detailed network at the level of synapse and connection, this is just one example from the work of Sebastian Seung and his colleagues, whereby they take a piece of cortex, in this case a very small piece of cortex. 20 micrometers, a very small piece of cortex and reconstruct it fully, completely using e m slices. This is the term connectomics here. And you can see a full network. These are of course fake colors. This is not rainbow. This is now e m so, you don't have colors for the cells when you slice them in e m. But then you can reconstruct them. Using artificial, dyes or colors in order to stain each cell after you reconstruct it, and then you get a full picture, including who is touching who synaptically, who is making synaptic contact, you can see blood vessel and so forth. So this Connectomics gives us really a whole, road map. If you want to call it, for a piece of a brain, hopefully for the whole brain. Of course they are, the question was and still is, so what do you do with this? You get all this gigadata or I don't know what, petadata, numbers of, of, of huge number, of, of detail well what do you do with it. So I just showed you last time. That you can use this amazing data to really solve particular questions. For example, orientation and direction and selectivity in the retina. I just showed you in the last lecture that you can now connect structure to function after you have an idea of what could be the function and how does this function is implemented in, let's say retina, you reconstruct the retina to really, cross check where the sum of your ideas, could be implemented by a particular system. So I think the [INAUDIBLE] is now getting its share Not just slicing and slicing and reconstructing, but asking questions, using this new technology. And the other technologies that I also mentioned, and that's related to the BAM, to the Brain Activity Map Project of President Obama, is this, system, whereby you use this new relatively new optics, and also the capability to record, from many, many cells simultaneously, still not millions or billions but maybe thousands, from a particular brain region, and so you can now, hopefully, connect, particular, activity of many, many cells using this In this case, calcium-sensitive dyes. There are may be other, other intrinsic dyes that you may develop that are not only calcium-sensitive, but other sensitive dyes to brain activity. But eventually you can see individual cells firing spikes via this calcium imaging when the system so to speak receives a relevant input, a bars orientation in different, direction, looking at when this particular region likes to behave, likes to respond to this particular input and so, and so. So both this advances in anatomy, huge new advances in ato, anatomically. Techniques, tools and also these optical tools, and the Obama project is to improve these optical tools and bring you 2 million cells, not only 2,000. Might provide, may provide, hopefully provide new understanding about how this circuit. Anatomy plus physiology, generates of this bigger function. [SOUND] Ok, so, so this is what I was saying, and, the whole idea was to really try to take the living behaving brain, while the the brain performs specific function, whatever the function might be, is it auditory, or vision, or motor, somatosensory. Doesn't matter. A function. And a function is not something easy to, to, to, to, to, to exactly define but if you think about orientation selectivity, you may ask yourself how does this network generate this capability, particularly.
