Generic structure of a PhD thesis in neuroscience (and perhaps mathemathical and natural sciences in general)

If you are wondering how your PhD thesis should be structured, here is a 

line of thoughts that helped me building mine and that is helping me right now 

in writing the content of the different parts.

Just think of a poster presentation in an important meeting. You are 

standing  there, a bunch of visitors with different backgrounds show-up at your poster asking you to present your work.

Professor - Hi, I'm Prof. X. I heard you just obtained your PhD degree. 

Congratulations! Now tell me what you have done.

You - Hello Sir, I'm <name>, I work in <institute>, with <name> and 

<name>. My work has been to develop a model of <something> to explain <evidences> and 

Give hints about <other problems>. We used <method> and tried <other thing>, 

and we found that <best> is the best way to do it.

Other professor - Oh wow! Interesting, I worked on the exact same field! 

Why did you choose this problem?

Y - Because <this and that>.

O - But isn't this useless?

Y - No because if you look at <fact> you might wonder whether it has a

relationship with <other fact> and this work might help understanding this

issue.

O - I see. But I heard <people> solving <similar problem> with <methods 

Different  from your>. Isn't also Michael <last name> in <lab in the US> doing the 

same?

Y - True, we might have tried that, but their method doesn't fullfill <some

requirement> and can't explain <some cool aspect>.

As you have seen, we just went through our Titlepage (introducing yourself 

And your institution), the abstract (a short, sounding summary) and the 

Introduction. The content of the Introduction chapter is needed to justify your work and identify it in light of other labs' research. As you see, it's not a summary of what you have done.

P - Ok, we got it, so how did you solve <problem> and what is your model

about?

Y - Ok, here is what I have done. First I developed <tools> by using <some

software> to interface to <some hardware>. Then I...

P - Wait a sec! Wouldn't it been easier to just use <this>?

Y - No for <this and that reason>. This new tool can be used also for 

<this and that>.

P - Alright, go on.

As you see, in the Methods you should always provide a little discussion 

for each of the important aspects of your research. We will skip the other 

parts of the thesis because I think you got the point.

Y - That's it!

P - Ok, interesting! We got the story. And... so what?

Y - Well, first of all our work shows that while <this> has revealed to be

true, <that> has revealed unimportant and <this other> is not true. It's 

like <analogy>, right? That's why probably the work of <authors> I told you in 

the beginning didn't lead to <some result>. Isn't this exciting?

Annoying professor - Oh, right, and would your model work in <other 

condition>?

Y - Well.. We tried something similar. we tried doing <this>, which we 

didn't put in the results, and the preliminary results suggest that it <does/doesnt> 

work.

A tall professor - If it does, will it be better than my model?

Y - Certain <yes/no>... We compared <this measure> with your model and the 

result <is/is not> in favor of our model.

Kind professor - Cool, so what do you think, is it better to follow now 

<this line of research> or <this other>?

Y - From my experience, both have interesting aspects...

This is the Discussion. It's a... discussion. People want to understand 

what's the benefit of your research, what are the insights they can gain on the 

field  from your research.

P - Ok guys, I have to catch a train soon, so now please summarize your 

work so I can remember you in the future.

Y - Alright. The field is <field>. The main issue of existing models is

<issue>. We thought "Well, why don't we <action> in <this way>". The 

Results are <results>. In the future it would be interesting to go in <some 

direction> to improve <aspect>.

And here we go, our Conclusion chapter is done! But before leaving the 

meeting, we need the most important chapter: Acknowledgments. Let's say the poster

session is over. You meet a friend and he proposes to have a drink.

Your friend - Do you want a drink? Let's leave this boring room and have a

drink on the balcony.

Y - Alright!

Your friend - So, how do you feel now that is over?

Y - Relieved but also excited, there is a lot to do!

Your friend - What do you think was the best part of your stay in

<institution>?

Y - Well I think <person> and <colleague> have been very important for all 

the drinks and stuff. The fun I had doing <this funny activity> helped me in

releasing  the tension. But also the inspirational conversations I had with

<friend> and <professor> have been very important for this work. It's been

difficult to always deal with my <girl/boy>friend and explain her that 

playing <sport> once a week is not wasting time for the thesis but it is a way to

improve productivity on the coming week. In conclusion, I think I should be

thankful to my family.

Friend - ...why??

Y - I don't know, but this is how acknowledgments always end.

- F. Stefanini

Basel Declaration

Basic re­search with an­i­mals as ex­per­i­men­tal sub­jects is es­sen­tial if we hope to un­der­stand the brain. The eth­i­cal stan­dards to which we hold our­selves and by which we con­duct those ex­per­i­ments are a mea­sure of our hu­man­ity.

The Basel de­c­la­ra­tion is a pub­lic com­mit­ment to seek the high­est pos­si­ble stan­dards for an­i­mal re­search. Sci­en­tists must trans­par­ently com­mu­ni­cate their work to the gen­eral pub­lic, and state their sup­port for well-con­ducted basic re­search. I have signed the basel de­c­la­ra­tion, and lay out here how I see the re­la­tion­ship be­tween basic and clin­i­cal re­search, and be­tween mod­el­ling and ex­per­i­men­ta­tion in neu­ro­science.

Modelling and simulation

All of the analy­sis and sim­u­la­tions I per­form are fun­da­men­tally based on data col­lected from an­i­mal ex­per­i­ments. There is no other source for the anatom­i­cal and phys­i­o­log­i­cal mea­sure­ments nec­es­sary to make mod­els of the neo­cor­tex, and so all neural modellers have an ethical investment and obligation in the animal research their models are based on. If we want to un­der­stand the work­ings of the human mind and brain, of course mea­sure­ments from human brains would be ideal, and our close pri­mate rel­a­tives the next best choice. For good eth­i­cal rea­sons, we re­strict the type and num­ber of ex­per­i­ments that can be per­formed on hu­mans, non-hu­man pri­mates and other an­i­mals. As a re­sult, most data that un­der­lie cor­ti­cal mod­els come from mice, rats and cats. As a rule, the data come from basic re­search — re­search with no di­rect clin­i­cal ap­pli­ca­tion.

Every model is based on simplifying assumptions.  Any model that assumes something that is known to be false is destined for the trash.  But in many cases, experimental measurements required to build an accurate model are either highly variable, or not present.  Every model can only be based on current knowledge, which implies that modelling and experimental work exist in a co-dependent cycle: models can make predictions and suggest hypothesis, which must be verified experimentally; the experiments offer improved data with which to improve models.

Cor­ti­cal mod­els have been pre­sented as a tool for re­duc­ing the num­ber of an­i­mal ex­per­i­ments that must be per­formed. This may well prove to be true, but the re­quire­ments are im­mense and ar­du­ous:

• A complete description of the connections between neurons in the human brain; ideally with a complete understanding of the rules used to form those connections.
• A complete understanding of the physiology of every class of neuron in the brain; ideally with an molecular-level understanding of how that physiology arises.
• A complete understanding of the developmental process of the brain, including all of the molecular and activity-driven cues required to form an adult brain.
• All of the above, for every neurological condition that one would want to treat.

Clinical value of basic research

Basic research is necessary if we want to understand the mechanisms by which brains work, and exactly how they fail.  This understanding makes it easier to propose a clinical treatment that has a chance of success.  Without a basic understanding of the working brain, we cannot know why some interventions work and other do not.  It is also difficult to judge in advance what the clinical outcome of a basic research result will be.

Deep brain stim­u­la­tion is a clin­i­cal treat­ment for Parkin­son's dis­ease and chronic pain. The tech­nique in­volves im­plant­ing a num­ber of elec­trodes in the thal­a­mus of a pa­tient (a sub-cor­ti­cal nu­cleus in­volved in gat­ing sen­sory and motor pro­jec­tions), and de­liv­er­ing pe­ri­odic elec­tri­cal stim­uli to the neu­rons there. It is a re­mark­ably ef­fec­tive treat­ment that dra­mat­i­cally im­proves the qual­ity of life of af­fected pa­tients.

Deep brain stim­u­la­tion is an im­pres­sive med­ical ad­vance de­liv­ered by neu­ro­science, and was only re­cently ap­proved for reg­u­lar use in pa­tients. How­ever, the foun­da­tions of this clin­i­cal treat­ment are laid on deep strata of basic re­search, most of which was not originally performed to cure Parkinson’s disease.  The clinical value of basic research must be measured over the scale of hundreds of years, not quantified over a decade or less.

Subscribing to an ethical framework

Given that an­i­mal ex­per­i­ments must be per­formed for neu­ro­science to progress, and given the deep pub­lic sup­port for med­ical ad­vances in neu­ro­science, how should those ex­per­i­ments be de­signed and per­formed? Sign­ing the Basel de­c­la­ra­tion im­plies that we will seek and up­hold the high­est stan­dards of an­i­mal care and ex­per­i­men­tal de­sign, by min­imis­ing the num­ber of an­i­mals used and im­pos­ing as lit­tle dis­com­fort and suf­fer­ing as pos­si­ble.

We also have an oblig­a­tion to en­gage with so­ci­ety, by ac­tively pro­mot­ing good re­search and trans­par­ently ex­plain­ing our own. We should pro­mote pol­icy based on sound ex­per­i­men­tal ev­i­dence and not on opin­ion. But equally, we in­sist that new laws gov­ern­ing re­search must be based on facts and de­mo­c­ra­tic dis­cus­sion. Rad­i­cal groups that act out­side the legal and eth­i­cal frame­works that so­ci­ety has agreed upon must be cur­tailed. Dis­cus­sion of an­i­mal re­search in the media and po­lit­i­cally must be con­ducted in an im­par­tial man­ner as part of a bal­anced di­a­logue in­clud­ing re­searchers.

- D.R. Muir

Hey! Small world!

            A major point of interest for INI-ots one and all is the local arrangement and connectivity of neurons in the neocortex.  But I'm working right now at Kinderspital Zurich, so I won't get into any of that nonsense.  What I'm going to talk about is the long-range connectivity of neurons.  And then I'll get a bit meta, but you don't have to read that part.

            Duncan Watts and Steven Strogatz (whose columns for the New York Times are a much better read than this) published their original paper on so-called “Small World Networks” in 1998.  Before I explain what they are, I'll explain where they are.  Everywhere.  For those who remain unconvinced, more compelling examples are below.  A small-world network is an attempt to mesh two common network types, those that are completely random, and those that are entirely ordered.  Many real-life networks tend to fall somewhere in between.  The brain is a good example of this: though most connectivity is local, long-range projections are what make quick computations possible.  Computations in a brain comprised of purely local connections would take so long as to make the incoming information irrelevant in the first place.  Lions as clumsy as cows would stumble after humans lumbering at a snail's pace, in a scene more reminiscent of Absinthe than Attenborough. 

            Watts and Strogatz's contribution was the idea that a certain amount of randomness in the organization of their network could decrease the distance between nodes without adding a noticeable detriment to the local clustering of the network.  The sense behind this is apparent: removal of a few short connections rarely adds more than one new link between neighboring nodes, while it shortens the number of links between distant clusters logarithmically, comparable to the advent of the freeway/motorway.  Whereas one originally had to take whichever local roads were available, being able to drive to a high-speed thoroughfare, even if its access point is in the wrong direction from the start, greatly decreases the time to get across town.  Even if it cuts off a few local streets (apologies to Arthur Dent). 

            To prove that this concept is not limited to the famed “Six Degrees of Separation” theory, they applied this to collaborations between film actors (Six Degrees of Kevin Bacon seems to be an overestimate; 3(.65) degrees would suffice on average), connectivity between generators, transformers, and substations on a power grid, based on the power lines connecting them, and, more notably, synapses and gap junctions in the flatworm C. elegans.  All showed the same principles, namely, a reasonably short average path length, with high amounts of clustering between neighbors.  Principles characteristic of small-world networks.  While we have yet to map every connection in the human brain (a bit more difficult with more than the 302 neurons of C. elegans), it is nevertheless extremely slick (sorry) that  a real-life neural network can be described by such a simple model.

Figure 1: The left-most graph features only local connectivity.  By exchanging one close connection for one random connection, the average path length decreases dramatically, but the local clustering remains high.  (Creative Commons Licensed from http://en.wikipedia.org/wiki/File:Watts-Strogatz-rewire.png)

            The gamma rhythm, which many, though not all, link to our ability to consider detached visual percepts (like your ability to recognize my feet as being associated with my face, even if the stuff in the middle is obscured) as being part of the same object, is too fast to be possible without such long-range connections.  Furthermore, small-world networks have been shown to be bistable, displaying both stable loops across their long-distance connections while also being capable of quiet behavior.  Memory consolidation and retrieval is often hypothesized to function using such networks; high network activity levels first imprint a memory in the synaptic strengths of a loop, and the network then lies quiescent until activity causes the memory to be recalled.  So this small-world stuff might be pretty big.

            I promised that this would get a bit meta, so I'll finish with one last example.  As a tongue-in-cheek tribute to the mathematician Paul Erdős, the Erdős number commemorates Erdős's penchant for collaboration by giving an Erdős number of 1 to all of his direct collaborators, a 2 to their collaborators, and so on.  While most scientists haven't worked on graph theory (Erdős and one of his collaborators, Alfréd Rényi, came up with a more random precursor to Watts and Strogatz's model), a very large and even more eclectic spread of scientists have an Erdős number, and it's average (per the Erdős number project website) is a mere 4.65, matching the short path length of a random graph, and the high clustering (after all, not many neuroscientists with Erdős numbers collaborate with the Erdős-endowed political scientists) of a tightly ordered graph.  Just 4.65 links between most scientists, and one of the forefathers of this concept itself.  A very small world indeed.

- G. Lee

Citations:

The original paper:

D.J. Watts and S.H. Strogatz.  Collective dynamics of 'small-world' networks. Nature 393, 440-442 (1998).

An introduction to the Temporal Binding hypothesis:

Wolf Singer (2007) Binding by synchrony. Scholarpedia, 2(12):1657.

And its detractors:

Shadlen, MN and Movshon, JA (1999). "Synchrony unbound: a critical evaluation of the temporal binding hypothesis." Neuron, 24: 67-77.

A quick introduction to bistability in small-world networks:

Cohen, Philip. Small world networks key to memory. New Scientist. 26 May 2004.

Steven Strogatz's Really Quite Well-Written Series, On the Elements of Math:

http://topics.nytimes.com/top/opinion/series/steven_strogatz_on_the_elements_of_math/index.html

The Erdős number project:

http://www.oakland.edu/enp