Indecision 2

The previous Indecision post, in case you're wondering, dates back almost 4 years, demonstrating if nothing else that my vacillatory tendencies are no passing fancy -- more of a lifelong vocation. I'd dither professionally if only I could decide which of the competing accreditory bodies to join.

In any case, the particular matter under consideration at the moment merits more than merely quotidian indecisiveness, given that it involves committing at least the next three years of my life to some largely unfamiliar project in some largely unfamiliar domain. In theory the deadline for this decision was today; I have ducked that, although it's not clear how long I will be able to procrastinate. In any case, for the record -- and perhaps in the faint hope that writing this down may help crystallise my thoughts -- the PhD projects in contention are sketched below. Bear in mind that I have a pretty slender grasp of this stuff at the moment, so these descriptions are likely to be wrong in every important respect.

• Interplay between mechanical and biological mechanisms during cell cortex assembly, supervised by Guillaume Charras and someone else TBD, located in the shiny new labs of the London Centre for Nanotechnology.

  The contractile cortex is an outer region of cytoskeleton -- the internal protein scaffolding that provides much of a cell's structure -- that attaches to the cell membrane. It can exert force on the membrane -- directly via attachment or indirectly by inducing pressure changes in the cytoplasm -- and thereby significantly affect the cell's shape. The structural bulk of the cortex is a mass of actin filaments, but there are many other proteins involved, in both functional and regulatory roles, many of which are not well understood.

  The purpose of this project is to investigate how the cortex is assembled and regulated, using a particular cellular phenomenon known as blebbing as its model. Blebs are transient blisters in the membrane where it detaches from the cortex and is expanded by pressure differences into the extracellular space. Although they are associated with some important cellular processes, including apoptosis and cytokinesis, and may sometimes be used for cell movement, blebs are probably not of huge importance in themselves. However, the expanding bleb contains no cortex, just cytosol. A cortex is then assembled and attached inside it, finally contracting to bring the membrane back to its normal position. The bleb thus provides a window through which cortical assembly can be viewed, using a range of cellular and molecular biological techniques in combination with optical, fluorescence and electron microscopy.

• Feedback control of transmission at individual nerve cell synapses, supervised by Trevor Smart, Guy Moss and Yuri Korchev, based in UCL's Pharmacology Dept, at least in part in the lab where my summer project nominally took place.

  The standard model of neural interactions is pretty much uni-directional: upstream neurons fire, causing release of packets of neurotransmitter at the synapses. Neurotransmitter molecules bind to receptors in the post-synaptic membrane of the downstream cell, opening up channels that adjust the ion balance in the cell, increasing or decreasing its likelihood of firing. The network of connected neurons can contain cycles, allowing for feedback mechanisms on a larger scale, but locally the information is considered to flow only one way.

  As it turns out, things aren't quite that clean and clear-cut, and there are circumstances where neurotransmitters are released from the body and even dendrites of the downstream cell that can affect the pre-synaptic behaviour, for example stimulating the release of inhibitory neurotransmitters. Not very much is known about such processes, including how significant they are in modulating brain function.

  The particular system of concern here comes from the mammalian cerebellum, looking at the interactions between nerves known as interneurons and the Purkinje cells they deliver signals to. Some theoretical work would be required to model how the released neurotransmitters diffuse through the rather complicated spaces between the cells -- which are full of all sorts of other junk just to make life more difficult. This would be married in some not entirely clear way to electrophysiological measurements taken via 'smart' (that is, Scanning Ion Conductance Microscopy-based) patch clamping and fluorescence confocal imaging to locate particular structures of interest.

• Combining NIRS with MRI for functional activation imaging in newborn infants, supervised by Clare Elwell, Jem Hedben and someone TBD, located in the Biomedical Optics Research Lab in Medical Physics, probably with some data acquisition in the Neonatal ICU at the Hammersmith Hospital.

  This project relates to the near-infrared spectroscopic imaging handwavingly discussed here (on Tuesday and in the comments). Putting aside the challenges of reconstruction, both NIRS and MRI provide ways of looking at what's going on in a baby's brain. NIRS is relatively portable, low impact and functionally important -- since the thing it measures, blood oxygenation, is crucial to health. MRI has amazing spatial resolution in general, but relatively poor time resolution, and its functional imaging -- which basically depends on looking at the difference between frames -- is actually coarser than that of NIRS.

  One component of the project, then, would be to collect data with both techniques simultaneously and find ways of meshing them together to get the best possible information out. This can then be applied to particular physiological problems concerning brain blood supply in infants. For instance, a number of studies suggest that babies show a markedly different oxygenation response to functional activation than adults -- basically, their oxygenation levels go down -- and this is not properly understood. The combination of techniques is also expected in the long run to find many clinical applications.

On the spectrum of CoMPLEX projects, all of these probably fall towards the more experimental rather than the theoretical end. They'll all sooner or later involve some kind of physical modelling and probably rather a lot of data processing, but none of them is at all abstract. From my current perspective this seems very appealing -- the opportunity to 'get my hands dirty' doing (in some crass and superficial way) 'real' science.

The trouble is, when it comes to the day to day business of real science, I haven't the first fucking clue what I'm talking about. It's all some romanticised fantasy with only the most distant and tenuous link to actual lab work. On the basis of my limited exposure so far I have no doubt that sheer novelty value will carry me over the first few weeks of playing with cells and high tech equipment and long hours at the microscope. I'm equally certain that a good deal of the laboratory experience is basically donkey work, requiring concentration and attention to detail but little intellectual engagement.

That, of course, is no different from many other activities we dedicate ourselves to, and I've been pretty successful at doing donkey work in my time. But without concrete experience of these particular donkeys, it's hard to know how I'll find it. Which makes it rather difficult to make a sensible decision.

Beyond that, I don't know enough about any of the subjects individually to be able to tell whether I'm likely to develop the kind of passion for them that would keep me going. Perhaps that's not a meaningful consideration. It's usually possible to find interest in any sufficiently-complicated subject when you're immersed in it -- I managed to get queasily fascinated by bloody financial derivatives, for instance, and indeed can still summon up some residual enthusiasm even now. So perhaps it's just a decision about where I'd rather see that passion grow.

The three topics are all pretty different. They all, I think, have something going for them. I just don't have any sensible criteria by which to compare them. And so, as usual, I am stuck in a quandary.

What do you think?
Posted by matt at September 14, 2007 06:36 PM