I'm very grateful for this work. However, saying that this contribution alone brings the tipping-point closer is rather pompous.
First of all, since gene expressions are such upstream information, knowing it won't give us the whole picture (or even come close to it.)
It's like knowing the code of a program, which has syntax you have never seen before, and you have no idea what the compiler does, not to mention the machine code describes an astronomical non-linear computation.
Second of all, gene expression dependents on age, environment, behavior, and probably things we don't even know. I haven't look at the data, but I assume it's just the mouse's brain in one instant. Otherwise it will be a VERY resource intensive job. Not to mention that you have to freeze the brain in order to slice it and map the gene (unless they got better technology now), so even if you get a different instant of the brain it won't be the same mouse.
However, a step towards the goal is a step towards the goal. And that's what this project was. Great work.
The Allen Brain Atlas, as currently realized, is likely to be very useful for disease research & basic biology. But, explaining how the brain works by profiling gene expression... I don't think so. Certainly not sufficient, maybe not necessary.
Generaly, though, applying high throughput imaging methods to different problems in neuroscience is going to be very fruitful. So in this sense, maybe we're near a tipping point. I don't really think quite yet, though; maybe in a decade or two.
There are plenty of microscopic techniques to distinguish inhibitory from excitatory cells. Computers can tell the difference if given appropriate microscopy images as input and the software to process those images. Not too hard.
But these are dynamic objects. Google 'rebound excitation' for some counterintuitive effects of inhibitory input. Hard to dissect how circuits of these objects work.
You're right, that was a pretty crummy suggestion. I looked at the wikipedia IPSP page and didn't like it much, either. A little deeper googling gave me this:
'Resting potential' = voltage between inside and outside of the neuron in the absence of any excitatory or inhibitory input. It is generated constitutively by ion pumps in the cell.
When cell receives inhibitory input, its voltage drops and it become 'hyperpolarized'. This causes a certain type of calcium ion channel, the 'T-type', to gain the capability of opening (it stops being 'inactivated'). When the inhibitory input ceases, the cell's constitutive ion pumps return it to its resting potential. As the voltage of the cell increases, some T-type channels begin to open, leading to an influx of calcium ions (there are more calcium ions outside the cell than inside), which in turn leads to the cell's voltage to increase further. Now the cell's voltage shoots above its resting potential, and voltage-dependent sodium channels begin to open. Sodium ions rush in, and a spike is generated.
And the payoff to me for trying to a little bit better job of answering your question is that I discovered what looks to be a pretty good resource:
Does anyone know how to get into diamond mechanosynthesis as a hobby (computational chemistry methods of course, not lab work)? I'm a pure math student, but the applications of that field are so vast (molecular nanotech?).
Executive summary: lottery winner writes self-congratulatory article with over-promising title for non-technical rag on lottery winner's funding an anatomical and genetic database of the nervous system -- the data least likely to yield a "tipping point" in how brains fundamentally work.
Off-topic, but your repeated use of the phrase "lottery winner" annoys me. Every self-made billionaire is a lottery winner in that none of them got where they are without luck. But none of them got there through luck alone. To characterize their success as pure luck indicates either ignorance or jealousy.
First of all, since gene expressions are such upstream information, knowing it won't give us the whole picture (or even come close to it.)
It's like knowing the code of a program, which has syntax you have never seen before, and you have no idea what the compiler does, not to mention the machine code describes an astronomical non-linear computation.
Second of all, gene expression dependents on age, environment, behavior, and probably things we don't even know. I haven't look at the data, but I assume it's just the mouse's brain in one instant. Otherwise it will be a VERY resource intensive job. Not to mention that you have to freeze the brain in order to slice it and map the gene (unless they got better technology now), so even if you get a different instant of the brain it won't be the same mouse.
However, a step towards the goal is a step towards the goal. And that's what this project was. Great work.