Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

A few questions about the field of computational neuroscience

  1. Nov 5, 2012 #1
    Hi, I am a physics & math double major but I am very interested in computational neuroscience. I am particularly very curious about how the field will potentially be around 10 years from now.

    I am extremely interested in things such as the blue brain project, the connectome project, and the human brain project.

    I have a few questions:

    1) How much programming experience would be needed to work in computational neuroscience? I have took a course on Fortran and I am now taking a course on Matlab. How about for something as potentially complex as the simulation of the full human brain (blue brain project)?

    2) Another interesting aspect would be the simulation of cognitive behavior (such as the human brain) in a virtual reality, what type of field will pursue this type of work?

    3) Neurorobotics. I like the idea of approaching robots from the aspect of our neurology. I suspect this field is computational neuroscience as well?

    4) I am also interested in brain interfaces. Would it be better to change my double major in physics and mathematics to accommodate for this possibility.
  2. jcsd
  3. Nov 7, 2012 #2
    156 views and no reply? :( Come on, someone please reply. This is my future we are talking about lol, it is very important to me. I'm seriously seriously considering changing my major, but I need to know some information first.
  4. Nov 9, 2012 #3
    The only person I know that went into computational neuroscience was a math undergrad. He had to go to grad school of course to specialize but he told me he models brain signals all day long. He was a whiz with fourier stuff so I can see why he chose that field.
  5. Nov 9, 2012 #4
    The best I can do is share my experience. I'm a physics undergrad as well so take this for what it's worth... I started out in bioengineering thinking that I wanted to do something with brain computer interfacing and neural engineering. I was (and still am) interested in interdisciplinary research so I joined a neurobiology lab that focused on modeling as well as the biology. I'm still in this lab and it's been a good experience so far, but I can tell you that I've found that it's not quite for me. I'm taking cell bio right now and it's helped me realize that you don't really need to take any biology to do interdisciplinary research in biology... You can simply look something up if you need to know it. I'm actually in the opposite position as I've switched my major from bioengineering to physics and I'm starting to find an interest in the earth sciences and geophysics especially.

    As far as your interests go, I would have to say that math/physics will certainly be good preparation for graduate study in computational neuroscience. I wouldn't suggest switching your major if modeling is what you're primarily interested in. But as for interfacing goes, most of that is done in engineering. Different branches of engineering can find a place in neuroengineering applications. Electrical/computer for the components, mechanical for the prosthetics, and bioengineering for the materials. There is obviously a lot of overlap between these and a physics major may be able to squeeze their way into the materials side, but I would say your best bet would be some type of engineering. It just depends on which aspect of neuroengineering research you find most interesting. Also, you may be able to combine your interests.. I've seen a couple of neuroengineering research groups that have some computational research going on.
  6. Nov 10, 2012 #5
    See that's the thing, I want to enter computational neuroscience but I am afraid of being in a narrow field of study where I end up not really understanding the brain at all.

    We share similar interests, I've also been really interested in brain computer interface and neural engineering.

    You don't need to do biology but your knowledge of the brain will be very minimal from what I hear. Someone on this forum likened it to having an equivalent knowledge of a bachelors compared to someone who did his PhD in neuroscience.

    How about for simulating the brain? I would love to work for something like the Blue Brain Project. Dr. Henry Markram is reverse engineering the brain down to the ion channels, check out his Ted Talk. I see huge potential for propelling our understanding of the brain and even for general AI and it makes me really excited. He is the reason that I became re-interested in entering computational neuroscience.

    Last edited by a moderator: Sep 25, 2014
  7. Nov 11, 2012 #6


    User Avatar
    Science Advisor

    If you're interested in the Blue Brain project, why don't you write to some of the scientists in it? Often, scientists are happy to answer questions from the general public (you should, of course, be courteous and non-anonymous). Here is a paper from the Blue Brain project with a link to the free full version. There is a contact email for the first author at the end of the paper. http://www.ncbi.nlm.nih.gov/pubmed/22991468
    Last edited: Nov 11, 2012
  8. Nov 11, 2012 #7
    Hey, thanks a lot. I'll print the article out and try to digest it as much as possible (after I'm done getting hammered by all of the exams coming up).

    I've found a bunch of free articles by clicking on Dr. Henry Markram's name on the link you provided. Here is his page http://www.ncbi.nlm.nih.gov/pubmed?term=Markram H[Author]&cauthor=true&cauthor_uid=22991468

    I'll email the author you mentioned, I also found Dr. H. Markram's email so that will do as well. ^.^

    I've also found this page that contains all of the publications directly related to the Blue Brain Project. Awesome xD http://jahia-prod.epfl.ch/site/bluebrain/op/edit/page-52755.html

    One more contact here http://bluebrain.epfl.ch/page-52758-en.html
  9. Nov 30, 2012 #8


    User Avatar
    Gold Member

    1) Brain Corp, directed by Eugene Izhikevich, for example, takes the spectrum of programmers form C++ to Python to MATLAB:
    http://braincorporation.com/index.php/category/opportunities/ [Broken]

    If you are a computational neuroscientist, I think your value would be more algorithmic. You'd be prototyping algorithms that programmers (if they like them) might turn into more efficient C++ code to incorporate into a larger project.

    2) This is more cognitive sciences, there's some overlap, but this is generally performed by people somewhere between neuro and psych.

    3) Yes; this I think, is the end point of the company I noted in 1)

    4) depends on if you want the software or hardware side of this. Hardware side would be more computer/electrical engineering. Software side, I think, fits comp neuro.
    Last edited by a moderator: May 6, 2017
  10. Dec 2, 2012 #9
    Thanks for the reply. Brain corporation looks like an institution I would apply for after I graduate.

    From what I hear, the programming that you do in computational neuroscience only requires you to know the basics (the rest you can learn on your own). The most time intensive part is the nonlinear modeling. Correct? Because of this I am double majoring in physics (BA) and biomathematics (BS). I've taken classes in Fortran, TRUE Basic, and Matlab. I might supplement that with a C++ class as well.
    Last edited by a moderator: May 6, 2017
  11. Dec 2, 2012 #10


    User Avatar
    Gold Member

    Both take a lot of time. Any good programming starts with paper and pen. Depending on the kinds of analysis you do, programming may help you automate analysis too.

    All I've ever used is matlab. I've poked sound with C and Python; think I'm going to invest more time in python.
  12. Dec 2, 2012 #11
    You research under computational neuroscience?
  13. Dec 2, 2012 #12


    User Avatar
    Gold Member

    My master's thesis (that I just completed) was on the very theoretical side. Though I used a physiologically valid model of a single neuron, and a typical coupling in nature (diffusive coupling... so a reaction-diffusion system) I don't know how physiological the coupling parameter is or the topology of the network.

    But it yields results interesting in dynamical systems.

    The project I'll be working on for my PhD looks like it will be modeling a real network, similarly diffusively coupled (but probably far more complex topology) using physiological parameters from an experimental collaborator. In both my last and future position, it seems the advisor doesn't care what programming language you use as long as you're comfortable with it and can reproduce results they expect with their already known systems.
  14. Dec 2, 2012 #13


    User Avatar
    Gold Member

    I also had a job in scientific programming while I was getting my degree. It seems to me, that people hardly ever see each other's code or exchange it (in scientific programming). I did one project with another employee and we knew the importance of annotation and transparency, but 90% of the code I write is cryptic fasthacks that I'll never use again. As I modify and clean up the 10% that I continue to use long term, I make them more robust and transparent.

    I think the point is that scientific advisors don't program (even if they had to as a grad); they get the final product (words and plots) from their student and they don't want to see the guts; they get to just think about scientific problems (and handle politics.. poor bastards).

    I think I might prefer programming to politics.
  15. Dec 2, 2012 #14
    Sounds interesting. Can you link me to your thesis?

    When you applied to graduate school, under what major did you apply in? Perhaps biophysics?
  16. Dec 3, 2012 #15


    User Avatar
    Gold Member

    It's not online yet, I don't have my own webpage. I should probably work on that.

    I did it through the Interdisciplinary Department. My university had a neuroscience program (in biology department) and a complex systems group (in physics department). I took advisors and classes from each and read the comp neuro literature myself.
  17. Dec 3, 2012 #16


    User Avatar
    Gold Member

    Here's one of my main references to give you an idea:


    That paper is much more formal, mathematically, than my thesis was. My paper was more computationally "experimental". I ran trials/simulations, explored the phase space, characterized behaviors, collected statistics, calculated typical complexity characteristics.

    My analysis looked more like this:


    only on a neural network instead of a slime mold network.
  18. Dec 3, 2012 #17
    So then what is your original major and what are you planning to do after graduate school?
  19. Dec 3, 2012 #18
    I don't have access to the article and I don't really understand the graphs so I can't really comment here.
  20. Dec 3, 2012 #19


    User Avatar
    Gold Member

    My undergrad was in physics.
  21. Dec 3, 2012 #20
    I really shouldn't be having an off-topic conversation on this topic because it is against the forum rules so I'll stop here.

    Thanks for your help, I'll message you if I have any other question.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook