UH Biocomputation Group - Benjamin Torben-Nielsen

Neuronal computation: dendrites at work

2016-11-10T10:15:13+00:00

Brain dynamics emerge from the collective and orchestrated activity of single neurons. The main characteristics of neurons are their morphologically elaborate structures to receive and integrate inputs (i.e., the dendrites) and communicate their signal to other neurons (i.e., the axons). Because dendrites receive, integrate and transform inputs into relevant output they can be considered as the functional workhorses of the brain.

In this presentation, I’ll outline the problematic relation between dendrite structure and function in neurons. Then I’ll show how highly non-trivial processing can take place in neurons due to the spatial extend of dendrites. Lastly, I will present my current work on distilling the essence of dendritic computations using simplified models.

Date: 11/11/2016
Time: 16:00
Location: LB252

From morphology to network-level activity patterns: dendritic arrangement and clustering among inferior olive neurons

2016-06-11T10:21:52+01:00

Marylka Uusisaari from the Erasmus Medical Centre, Rotterdam joins us for a special journal club session.

Anatomical understanding of the neuronal circuitry, comparable to electronic chip blueprint underlies theories and models of computational capabilities of various brain structures. The fundamental and essential nature of this knowledge is recently gaining visibility in the form of large-scale connectomic mapping initiatives. Their work aims (among others) to define the communication pathways between neurons, most commonly delineated in terms of axonal termination spaces and their overlaps with the target neurons’ dendritic fields.

The axo-dendritic or axo-somatic chemical synaptic connection is not, however, the only mechanism for fast interneuronal communication, as gap junctions linking neighbouring neurons provide the means for electrical signal propagation and synchronisation of spiking activity. A prime example of a structure where this mechanism plays a key role in shaping network activity is the inferior olive (IO), a nucleus in the brainstem integrating multimodal sensorimotor input and providing the climbing fibre input to the cerebellum. Intriguingly, there seem to be no local chemical synapses within the IO; instead, the coherence of inter-olivary network relies entirely on gap junctional communication. Thus, it is of key interest to define the anatomical arrangement of IO cells in respect to each other, as the amount of electrical coupling between individual IO cells - defining the emerging spatio-temporal patterns of olivo-cerebellar activity - must depend on the extent of dendritic overlap.

It is generally assumed that the IO cells are spherical neurons interspersed homogeneously throughout the nucleus, with the strength of electrical coupling decreasing with increasing inter-somata distance. However, this assumption has not been rigorously examined until now; and indeed, early anatomical works (Sotelo et al., 1974) described the olive as formed of segregated clusters of olivary cells.

To gain insight into the possible inhomogeneity and anisotropy present on anatomical level in the IO, we employed a novel “sparse viral labelling” technique that preserves the flexibility of genetically targetable staining but results in a sparse, Golgi-stain-like labelling of neurons. This method allows detailed reconstruction of large number of neurons in thick brain sections and thereby quantitative assessment of their dendritic morphology in respect to the boundaries of the IO as well as to the neighbouring cohort of IO cells.

Examining a large number of reconstructed cells as well as the overall arrangement of thousands of IO cell bodies revealed that while closely positioned IO cells’ dendritic fields may overlap to a great extent, the inter-somatic distance is not necessarily indicative of overlap. In contrast, IO cells can show strong avoidance regarding their neighbouring cells’ dendritic fields, suggesting that the functional clustering of IO as well as their axonal activity (i.e., the climbing fibre) is defined by the IO cell dendrite arrangement. Such non-uniform neuronal arrangement calls for re-evaluation of our hypotheses regarding the origins of cerebellar complex spike synchrony.

Date: 17/06/2016
Time: 16:00
Location: LB252

PhD studentship in Computational Neuroscience

2015-12-21T16:45:55+00:00

This position has been filled.

We welcome applications for a funded PhD position in the Biocomputation Group at the University of Hertfordshire. The successful applicant will work on a project related to modeling the structure and physiological function of the inferior olive. The inferior olive is an important part of the olivo-cerebellar circuitry as its axons, the climbing fibers, play a central role in all theories of cerebellar learning. Further, the inferior olive has been postulated to act as a clock for the brain. In this project, we aim to use an accurate structural model of the inferior olive to investigate, in silico, recent hypotheses related to the generation and maintenance of timing signals. An integral part of this project involves the further development of the NeuroMac software to generate structural models of the inferior olive. Please refer to the publication list on the website for recent publications.

Who are we looking for?

Candidates should have excellent programming skills (preferably also in Python) and familiarity with parallel and high-performance computing and visualization toolkits (such as VTK) is appreciated.
Candidates should possess great curiosity.
Candidates should be able to formulate their own questions about the circuitry and dynamics in the olivo-cerebellar system to direct their research efforts.

Knowledge of neuroscience is a plus but the eagerness to learn about the brain and how to use scientific approaches (such as computational neuroscience and neuroinformatics) is considered pivotal.

Successful candidates are eligible for a research studentship award from the University (approximately GBP 13,800 per annum bursary plus the payment of the standard UK student fees). Applicants from outside the UK or EU are eligible, but will have to pay half of the overseas fees out of their bursary.

Research in Computer Science at the University of Hertfordshire has been recognized as excellent by the latest Research Excellent Framework Assessment, with 50% of the research submitted being rated as world leading or internationally excellent. The Science and Technology Research Institute provides a very stimulating environment, offering a large number of specialized and interdisciplinary seminars as well as general training and researcher development opportunities. The University of Hertfordshire is situated in Hatfield, in the green belt just north of London.

The student will be supervised by Dr. Ben Torben-Nielsen (b.torben-nielsen at herts.ac.uk) and Dr. Volker Steuber (v.steuber at herts.ac.uk) to whom informal enquiries can be sent. Application forms can be obtained from:

Mrs Lorraine Nicholls,
Research Student Administrator,
STRI,
University of Hertfordshire,
College Lane,
Hatfield, Herts,
AL10 9AB,
Tel: +44 01707 286083,
l.nicholls @ herts.ac.uk.

The short-listing process will begin on 28 December 2015.

Reconstruction and simulation of neocortical microcircuitry

2015-11-05T09:38:27+00:00

In 2013, the Human Brain Project was awarded one out of two European FET flagship grants worth 1.2 billion Euros over the next ten years. Recently, the HBP’s predecessor, the Blue Brain Project published a first-draft model of a small piece of rat cortex . A staggering 13 companion papers were published at roughly the same time disclosing details about the model construction pipeline, the algorithms used and dedicated software to built and disseminate the model. The reported simulations indicate a high degree of accuracy in replicating in vitro and in vivo (rat) brain dynamics.

I would like to highlight some parts of the model and selected results. Then, I will discuss what this type of results can tell us. Specifically, I want to address incredible flexibility of the brain (because a simulation without several crucial components can still replicate brain dynamics) or, our grotesque misunderstanding of the brain (because a simulation without crucial components appears to replicate brain dynamics).

Date: 06/11/2015
Time: 16:00
Location: LB252