Dear students, you will find in this page an overview of a variety of possible topics for student final projects (BEP and MEPs, research masters), and internships, currently available at the Neurodynamics branch of the Neurocomputing Lab (NCL).
Most final year BSc and MSc projects consist of an internship, literature study and a final project. The latter two should as much as possible be linked together (it helps, though it is not a strict requirement). Have a look into the project descriptions below, and we can then plan a conversation to discuss specifics.
If we jointly decide to pursue one of these, we will plan both the literature study and the specific milestones of the final project. Projects are either related to ongoing research in the NCL, or exploratory research lines. Projects can be modified based on your interests and background. You will find that this document shows projects for MSc final thesis assignments. However, it will be indicated which projects can be used for BEPs. BEP students: don’t be afraid to tell us that you like a project even if it is “only” for MEP!
Note: We are always looking for eager students to work with, so please feel welcome to propose your own ideas for a project!
NCL: The Frontier Between Computational Neuroscience and Neural Computation
Most of our efforts relate to the porting of biology to “in silico” models. We traverse levels of description and implement neural networks both at a biophysical and at simplified computational levels. At the NCL we are interested in the secrets that biology has wrought which endow the brain with high level function, and particularly sensorimotor function (motor behavior). Our particular focus is in the reconstruction and analysis of the olivocerebellar networks, concerned with sensorimotor behavior.
For the difference between computational neuroscience and neural computation watch this video.
The projects below center on the structure, function and dynamics of the networks of the cerebellar system. A somewhat dated, but still relevant talk (by Mario for the Belgium Society for Neuroscience) can be found here.
If you look into the sample of theses in the lab, you will notice that we have a wide variety of topics, for instance, regarding neurodynamics of recurrent neural networks. If you have an idea about your thesis and you think it may fit within our themes, feel free to talk to us.
Analytical skills are required for most projects. Also essential is some background on biophysics of neurons and the origin of action potentials (such as the Hodgkin-Huxley model), ordinary differential equations, linear algebra and a reasonable knowledge of statistics. Your programming skills can range from basic to advanced, and you can expect to exercise them in the course of these projects.
- (MEP) Unsupervised models of cerebellar plasticity (PF-PC plasticity in the cerebellar loop)
- The olivocerebellar system consists of multiple feedback and feedforward loops. One of the main loops that we are interested in consists of the “Parallel Fiber (PF) – Purkinje cell (PC) – Deep Cerebellar Nuclei (DCN) – Inferior Olive (IO) – Purkinje cell” loop. Most theories of cerebellar function are based on the coincidence of PF-PC signals with the activity arriving via the climbing fiber (CF) from the IO, as these signals modulate the PF-PC synapse. Hence, there is particular interest in investigating plasticity at the PF-PC synapse. We start by looking at plasticity with a simplified loop (no excitatory/inhibitory interneurons and Golgi and basket cells). This closed loop model of the cerebellar circuit will consist of a mixture of spiking neurons and conductance based neurons (Hodgkin-Huxley), with a model of synaptic plasticity to study how incoming signals change the cerebellar circuit.
- (BEP) When does a neural circuit resonate?
- Part of the IO population in the cerebellar loop is coupled. This means that they share information and are affected by the response of each other. While each PC receives information from multiple PF, and one IO projects onto multiple PCs, each PC only receives information from one IO. The PCs project to multiple DCN which in turn project to multiple IO cells. This intricate combination of connections means that an input on one or multiple PCs will affect another PC through both the PC – DCN – IO – PC connection and the IO – IO coupling. Hence, we want to investigate in which scenarios (or for which inputs) does the circuit resonate.
- (MEP) Mixing and matching: how to create a homeostatic neural network?
- Neurons alter both their intrinsic and synaptic properties with plasticity mechanisms such as long term potentiation (LTP) and depression (LTD). These alterations in the neuron’s properties lead to changes in the neuronal response (i.e. firing frequency). Homeostatic networks are able to maintain a target level of neuronal response. The question is, how do we get there?
- (MEP) Local Field Potential measurements of Inferior Olivary networks (reproducing biological behavior)
- A key step in all computational and theoretical neuroscience is model validation with experimental data. While our models are defined at the cellular level, we don’t have experimental electrophysiological data from large sets of individual cells in the inferior olive. We do have local field potential measures of these brain regions and we’d like to use these to validate and optimize our large scale networks to these experimental data. Your task will be to recover local field potential information from large scale simulations and compare these with published experimental data.
- (MEP/BEP) Reconstruction and analysis of realistic olivocerebellar networks
- Currently we have small (~1000 cell) biologically realistic networks of the inferior olive and cerebellum. We are currently expanding these networks to the scale of realistic mouse networks to recreate the mouse olivo-cerebellar loop of motor control to study motor control and learning bottom-up, while at the same time increasing the level of morphological model detail. This will require efficient use of a large amount of computational resources. The neuronal clustering behaviour in these large-scale simulations will then be analyzed either by in-house developed statistical methods or something you come up with yourselves during the project.
- (MEP/BEP) Measuring synchrony and dynamical cluster formation
- The neurons of the inferior olive (IO) form a coupled nonlinear oscillator network. This means we will observe various spatio-temporal patterns and different synchronization behaviour depending on cell and network parameters. Your task is to build different network models of the IO, simulate those and come up with methods to quantify synchronization behaviour over time within the simulated output. Localized synchronization and desynchronization is thought to lead to complex spikes which indirectly control muscle movement.
- (MEP) The influence of plasticity on cluster formation in the Inferior Olive
- As said before, the neurons of the inferior olivary nucleus can be modelled as coupled nonlinear oscillators and the synchronization between neurons is thought to play an important role in motor control feedback. Apart from various parts of the brain that are able to specifically modulate this synchronization behaviour via glomerular synapses, the inferior olive itself is also able to change the connection strength of dendro-dendritic gap junctions over time via various plasticity methods. This conductance modulation will change how cells synchronize and is expected to have a key role in motor learning and motor feedback tuning. Your task will be to quantify the effect of various models of plasticity, including spike-timing dependent plasticity (STDP), on dynamical cluster formation in the inferior olive and then to connect this to motor control and learning.
Cellular/Sub Cellular Level
- (MEP/BEP) Homeostasis: How do neurons change to stay the same?
- One of the fundamental differences between artificial neural networks and biological neural networks is that in the latter the neurons adjusts its own intrinsic properties to maintain functionality. Neurons maintain complex regulatory protein expression (i.e., ion channels) as a function of their activity. The molecular pathways involved in this regulation are only now beginning to be unveiled. This project studies the processes maintaining equilibrium either from a mathematical or a biological perspective. We model with a variety of neuronal types (Purkinje neurons, Inferior Olivary cells), and simulate them at various levels of detail.
- (MEP/BEP) Influence of temperature on neuronal dynamics
- The Inferior olive’s dynamics depend on the balanced interaction of a multitude of voltage dependent currents. More specifically, this balance comes forth from the timed activation and inactivation of various ion channels and may vary as a result of environmental factors. Amongst these environmental factors is Temperature. Temperature influences the conductance, activation and inactivation of all ion channels. As different ion channels generally have different temperature sensitivities, this poses a challenge for maintaining stable oscilla-tory behaviour over extended temperature ranges. Your task will be to research and simulate the temperature (in)variance of the inferior olive.
- (MEP) Intrinsic cascades of ion channels expression
- (MEP/BEP) Chloride channel influence on excitability
- Recent research has shown the existence of another repolarising current in the inferior olive neurons: calcium activated chloride channels (CaCC). Experimental evidence shows that CaCC have large influence on the sub as well as suprathreshold behaviour of inferior olive neurons. We have already added these channels to the our model cells. Your task will be to analyze the behavior of these channels in the context of an existing model of the inferior olive and examine its effects, at the cellular and network levels.
- (MEP) Calcium diffusion in morphologically detailed inferior olivary cells
- – Calcium diffusion has been theorised as one of the main drivers behind the inferior olive’s dynamics. Many of the channels present in the inferior olive show functional coupling between calcium channels and calcium concentration activated channels. Moreover, the calcium concentration itself is further influenced by buffers, diffusion and pump mechanisms. Detailed modeling is preferred considering the large influence calcium concentration has on a neuron’s behavior. However, most models use simplified ‘phenomenological’ calcium concentration models which group all aforementioned biophysical influences together into one parameter as it makes tuning much easier. Your task will be to update an existing phenomenological calcium concentration model to include the effects of diffusion. This project would be conducted with the STEPS simulator and in collaboration with the computational neuroscience laboratory of Erik De Schutter in OIST, Japan.
- (MEP) Modeling biophysical plasticity mechanisms of gap junctions
- The synaptic arrangement of gap junctions of the inferior olive are unique: each gap junction exists in a glomerulus which receives convergent synaptic terminals of both excitatory and inhibitory sources. Gap junctions undergo plasticity with the expression of new protein (Cx36), which is dependent on a complex molecular expression network, many factors of which are known. We intend to model the complex molecular interactions in a reconstruction of a glomerulus, where protein diffusion and reactions guide the plasticity process. This project would be conducted with the STEPS simulator and in collaboration with the computational neuroscience laboratory of Erik De Schutter in OIST, Japan and the Lab of Marcel de Jeu, who has conducted in-vitro gap junction plasticity experiments.
Example Published Theses
- Nanobiology (TUD)
- Romano Van Genderen (2018) – Synchronicity and patterning in the olivo-cerebellar and cerebello-cortical loops
- (MEP NB) Theodora Trandafir (2018) – Parameter space exploration of the subthreshold dynamics of morphologically detailed neurons of the inferior olive
- Biomedical Engineering (TUD)
- Piergiuseppe Liuzzi (2020) – Parkinsonian Resting Tremor: Source and Interaction with Movement
- Thijs Hoedemakers (2020) – Physiological Cerebellar Model in Motor Control
- Elias Mateo Fernandez Santoro (2019) – Encoding of Correlated Temporal Information in a Model Cerebellar Loop with Olivary Oscillations and Long-Term Plasticity
- Nanobiology (TUD)
- Arun Karim (2021) – A simple model of the Purkinje cell
- Rocher Smol (2021) – A biophysically detailed model of the IO cell including Calcium dynamics
- Mathijs Heidekamp (2020) – Model optimization of inferior olive cells.
- Friso Douma (2018) – Innovation and validation in the multicompartmental modeling of inferior olive neurons
- Staf Bauer (2018) – Analysis of simple and complex spikes in Purkinje cell mutants
- Ronald Janssen (2018) – Learning Pong
- Aoibhinn Reddington (2017): Cerebellar Involvement in Respiration
- Jeroen Metthorst (2017) – Plasticity in a Kuramoto Model
- Capstone (EUC)
- Daphne Cornelisse (2020) – Transients in Randomly Recurrent Neural Networks
- Naomi Hulst (2020) – Large scale parameter explorations of the IO
- Mario Negrello: email@example.com,
- Elias Fernandez: firstname.lastname@example.org
Selected Lab Publications
- Negrello, M., Warnaar, P., Romano, V., Owens, C., Lindeman, S., Iavarone, E., Spanke, J., Bosman, L., Zeeuw, C. (2019). Quasiperiodic rhythms of the inferior olive PLOS Computational Biology 15(5), e1006475. https://dx.doi.org/10.1371/journal.pcbi.1006475
- Vrieler, N., Loyola, S., Yarden-Rabinowitz, Y., Hoogendorp, J., Medvedev, N., Hoogland, T., Zeeuw, C., Schutter, E., Yarom, Y., Negrello, M., Torben-Nielsen, B., Uusisaari, M. (2019). Variability and directionality of inferior olive neuron dendrites revealed by detailed 3D characterization of an extensive morphological library. Brain structure & function 92(4), e52068 – 19. https://dx.doi.org/10.1007/s00429-019-01859-z
- Gruijl, J., Sokol, P., Negrello, M., Zeeuw, C. (2016). Calcium Dependent Gap Junction Plasticity: Modulation of Electrotonic Coupling in the Inferior Olive Glomerulus bioRxiv 17(), 072041. https://dx.doi.org/10.1101/072041
- Sudhakar, S., Hong, S., Raikov, I., Publio, R., Lang, C., Close, T., Guo, D., Negrello, M., Schutter, E. (2017). Spatiotemporal network coding of physiological mossy fiber inputs by the cerebellar granular layer PLoS Computational Biololgy 13(9), e1005754 – 35. https://dx.doi.org/10.1371/journal.pcbi.1005754
- Hong, S., Negrello, M., Junker, M., Smilgin, A., Thier, P., Schutter, E. (2016). Multiplexed coding by cerebellar Purkinje neurons. eLife 5(), 1234. https://dx.doi.org/10.7554/elife.13810
- Ju, C., Bosman, L., Hoogland, T., Velauthapillai, A., Murugesan, P., Warnaar, P., Genderen, R., Negrello, M., Zeeuw, C.(2019). Neurons of the inferior olive respond to broad classes of sensory input while subject to homeostatic control.The Journal of Physiology 597(9), 2483 – 2514. https://dx.doi.org/10.1113/jp277413
- Romano, V., Reddington, A., Cazzanelli, S., Negrello, M., Bosman, L., Zeeuw, C. (2019). Functional convergence of autonomic and sensorimotor processing in the lateral cerebellum bioRxiv
- Gruijl, J., Hoogland, T., Zeeuw, C. (2014). Behavioral correlates of complex spike synchrony in cerebellar microzones. Journal of Neuroscience 34(27), 8937 – 8947. https://dx.doi.org/10.1523/jneurosci.5064-13.2014
- Hoogland, T., Gruijl, J., Witter, L., Canto, C., Zeeuw, C. (2015). Role of Synchronous Activation of Cerebellar Purkinje Cell Ensembles in Multi-joint Movement Control Current Biology 25(9), 1 – 10. https://dx.doi.org/10.1016/j.cub.2015.03.009
- Warnaar, P., Couto, J., Negrello, M., Junker, M., Smilgin, A., Ignashchenkova, A., Giugliano, M., Thier, P., Schutter, E.(2015). Duration of Purkinje cell complex spikes increases with their firing frequency Frontiers in Cellular Neuroscience 9(), 1 – 30. https://dx.doi.org/10.3389/fncel.2015.00122
- Gruijl, J., Sokol, P., Negrello, M., Zeeuw, C. (2014). Modulation of electrotonic coupling in the inferior olive by inhibitory and excitatory inputs: integration in the glomerulus. Neuron 81(6), 1215 – 1217. https://dx.doi.org/10.1016/j.neuron.2014.03.009