Where I've Been and Where I'm Going

I promise there's a reason I haven't been posting to this blog as of late, and it isn't at all related to an excuse about my cat eating my blog drafts.

First, on the high school lab demonstration front, I've been stalled waiting for the chemicals to do the electrophoresis portion of the lab. We recently received the enzymes we needed from New England Biolabs, but haven't received the buffer chemicals to do any sort of qualitative analysis of the effects of the enzymes on our DNA samples. So, there hasn't been any work to be posted.

Second, on the front of neuroscience, I've recently decided that I am going to need a whole slurry of advanced math knowledge that I currently don't posses. As I have said before, my college math exposure is severely limited, and I've been acquiring the skills necessary to do advanced calculus and work with differential equations. To that end, I've been running through the Khan Academy programs on calculus and differential equations, as well as several resources I've found on the web, including the MIT OpenCourseware 18.01,18.02 and 18.03 video lectures. The "Paul's Math Notes" resource has also proved invaluable, available at http://tutorial.math.lamar.edu/Classes/. Note that anything I state incorrectly about math in the future should in no way reflect the quality of material at any of these sites, only my inability to absorb the information in the small amount of time I have allotted myself.

Also, I've been doing research on the structure of the neocortex in mammals, and our understanding of the processes their in. I've especially been focusing on the enzyme cascades involved in g linked protein receptors, as so far I've seen no evidence that these massive enzyme reactions are taken into account in the mathematical models of the brain I've seen so far (though I've only been looking at single neuron models, though most higher level models seem to be simplifications of the single cell models).

Over all, it seems that there is a lack of communication across the intra-discipline facets of neuroscience. The biologists aren't talking the physiologists, and everybody has their own mathematician, while the information theory guys sit in a corner and play with idealized circuits that no more approximate actual neurons than a light switch approximates a modern cpu. Maybe I'm wrong, but it seems to me like a good deal of the knowledge needed to understand the brain exists, but has never been combined (with the synthesis of new relations and destruction of dead end supposition that that would imply).

I find myself wondering why this hasn't happened, and almost am unable to believe that it isn't some gross oversight of recent research on my part. But if it isn't an oversight on my part, I wonder why this coming together of the various researchers in a self proclaimed inter-discipline science has never seemed to happen.

For example, I have an affinity for the serotonin g linked protein receptor cascades. I think that the interaction of the raphe nucleus (the main serotonin producer in the brain), which connects to almost every other part of the brain, including the neocortex, could be a very interesting site to study the down stream effects of these enzyme cascades on overall neuron firing. Yet nobody seems to have talked about it except in reference to sleep cycles.

The same goes for every other thought I have about the brain as a whole. Everybody seems to be doing studies of these small problems, creating points of understanding, but nobody seems to be drawing the thick connecting lines that are going to outline the actual functioning of the brain as a whole. Perhaps this is a sophomoric line of reasoning, and maybe it is that they want to be certain of their conclusions before they integrate that understanding into the larger image of the brain, but it seems to me like there has to be a connecting of assertions at some point, and i think connecting these small studies would inform them just as putting a part of a machine into a simulation of the machine and having the machine either work or not work would inform the process of creating the part.

It doesn't help that there aren't many killer applications of a whole brain understanding, at least that the few corporations involved in brain research are willing to see. Most applications that involve making money seem to be centered around developing an almost autistic understanding of a single process in the brain (memory for alzheimer's and dementia research, mood for the psycho-pharmaceutical companies, etc.), and university researchers seem almost afraid of stating any conclusions about the brain in general for fear of being wrong.

So, I've been doing a lot of thought experiments and studying, and not a lot of anything that you would want to read about on a blog. Most of the time I find myself wandering around in a daze, puzzling over the implications of a certain assertion about the functioning of the brain, trying to picture giant logic charts in my head. When I'm not doing that  either have my head shoved hermetically into a piece of graph paper doing a problem, or reading/watching on the computer until my eyes or head hurt, whichever comes first. It doesn't help that my hours at my day job have recently been cut back, so my ability to buy equipment has been decidedly hampered.

But I promise you this, I am still most definitely interested and active in the topics of this blog, I'm not dead, and haven't abandoned this site as so many people do with websites after a month or two. I've just found myself without anything to publish, or so I thought until I look above this cursor and see a huge screed laid out (though arguably more of it seems to be about what I haven't figured out than what I have figured out, but there's something about wisdom that comes to mind when I see that). I'll continue updating as I come upon things worth blogging about, but can't promise there will be a heaping of blog posts for whoever reads this every week, or every month for that matter. Put me on your update list and keep an eye out, but holding your breath may prove unhealthy.

What Have I Been Doing?

Just what Have I been doing lately? I certainly wasn't posting on this blog, as is evident from the time stamp on my last post. I've been working on the high school science class demonstration I outlined in my last post, and doing a LOT of reading on neuron imaging, current injection and computational models (both theoretical and practical).

On the front of science demonstrations, I've made a lot of progress as far as pinning down protocols, and now am mainly just waiting for supplies to arrive so I can start testing the techniques and protocols I've conglomerated from resources online and the helpful posts of fellow biohackers on the Diybio google group. I've posted a first tested draft of a human DNA extraction protocol on this blog for the world to start perusing (if the world ever manages to find my blog), and it is workable if not incredibly coarse and still slightly unquantified (ran out of time to test the optimal amount of meat tenderizer protease for the mixture, and will probably need to see results in electrophoresis before I can really determine what ratio and conditions are optimal or even make a difference).

I spent five hours yesterday in the high school lab working on different scales of extraction and trying to coax the DNA I extracted into "spooling" onto a bamboo shoot (as is talked about here, here and just about everywhere else that uses a protocol like the one mine is based off). My conclusion is that either my stirring/enzyme use has somehow not yielded the structural form of DNA that their experiment yields (shearing or DNA structural proteins still coiling the DNA), or that human and animal DNA behaves differently at a macro level. My specialty isn't genetics so I consulted the diybio group, and their answer is "we haven't ever tried 'spooling' DNA". My conclusion is that spooling human DNA most likely isn't possible, and am going to use a centrifuge to pelletize the DNA im trying to extract.

So, yes, I bought my first piece of lab equipment yesterday. A Dremel 300 from a big box store for $49 US, and a Dremelfuge Rotor designed by Cathal Garvey and sold by Shapeways for a total of $58 US with shipping to the USA (not sure where it's coming from). It should arrive by 11/9. It promises to be a good investment, being able to develop (relatively safely; disclaimer: I'm not responsible if you get shot in the leg or eye by an errant centrifuge tube) just over 50,000Gs worth of force. That's as good as a high grade commercial centrifuge I'm told, and I'm excited to see how it works, and will post pictures, text and maybe even a video of it when it comes in.

In the realm of neuroscience I've been splitting my time between continuing to read about research methods in the realm of neuron imaging and current injection (reading what a neuron does, and making it do what I want it to do respectively), and the mathematical models that exist to describe both a single neuron and the computational nature of many neurons acting in concert.

In the realm of neuron imaging I have gone from reading about recording electrodes full circle back to the realm of fluorescence imaging techniques, this time in reference to voltage sensitive dye techniques, which are less complicated than the voltage sensitive proteins I had read about and possible mentioned earlier in the course of this blog.

Also, I'm trying to pin down and wrap my head around the nature of and the specifications of an electrode capable of injecting a current into a single neuron along the lines of those used in the Dynamic Gap technique, which I have been reading about quite enthusiastically. It sounds like the perfect technique for refining and modifying the single neuron model that I think badly needs attention.

And what better segue could I have asked for to talk about computational neuroscience, aka the art and math of figuring out what these neurons we're observing are doing to information? I've been mostly focused on the single neuron models such as Intercept-Fire, the Hodgekin-Huxley model, and the stochastic models derived from the H-H model. In order to look at system level models I'm going to have to have a good grasp on these first. I'm most definitely only doing research at this time, and to be honest am struggling with the math involved, as my last Calculus class was in my senior year of high school, and even then we didn't get very far in the book. So, right now I'm wrestling with the single neuron model, trying to work my way up, fighting on another front with the math involved with any of this. The single greatest resource I've found so far is a computational neuroscience textbook that I randomly found posted on a website in the .ch domain. I haven't had the time to figure out if it's a legally published copy or not, but it has most definitely been a god send for this scientist.

I apologize for the lack of links in the neuroscience part of this post, and will do a link update in the next few days, but I'm too lazy on my last day off to track down all the articles I've been reading. I did at least bother to find the link to the neuroscience textbook I've been reading, as it's sitting in the tab next to this one on my browser. You'll be hearing a lot more from me, and possibly even seeing me, in the coming weeks, as I ramp up my efforts on this science demo, and continue working on my neuroscience pursuits. Now:


Stand Back, I'm Going to Try Science.