Connectionists: PhD position on computational modelling of neural network development

[roman bauer]

 Dear all,

A fully funded PhD position in Computational Neuroscience is available at
the University of Cyprus in collaboration with the University of Surrey
(UK), titled “Brain Neuronal Networks Development via Multiscale
Agent-based Modelling”. More details are available here. The closing date
for applications is May 31st, 2024.


*Project summary*

Here, we will demonstrate an innovative computational approach to model and
emulate biological neural networks (NNs): rather than directly modelling a
brain with all its complexity, we will model NN development from a single
precursor cell. In other words, by leveraging the same approach that nature
uses to build brains, we aim to reproduce challenging neural complexities.
Inspired by the biological brain, we will make use of developmental rules
that are encoded in a gene-type manner. For instance, a rule could be that
a given neural stem cell should divide under certain conditions while they
could migrate under others. Similarly, neural connections develop according
to well-conserved mechanisms that are adaptive to individual neurons’
activities and wiring patterns.

To inform the computational modelling, we will make use of data from
experimental studies as well as synthetic, simulated data. Well-established
characteristics of biological NNs will be used to inform realistic NNs,
both structurally as well as functionally. We will make use of innovative
machine learning techniques to match the in-silico NNs with specific
organisms. In the early stages, we will employ synthetically generated NNs
as a test scenario, and iteratively increase biological correspondence. We
will make use of the agent-based modelling software BioDynaMo (
www.biodynamo.org), an open-source software we have been actively
developing for almost a decade. Notably, this project builds on previous
work of the supervisory team, including for instance the simulation of a
spatially embedded, functional, and biologically realistic neural network
that self-organized from a single precursor cell (link).


*Training*

The studentship is highly interdisciplinary, because it is at the interface
of computer science and neuroscience. Machine learning will constitute a
core component of this studentship. This work will benefit from the
prestigious BioDynaMo collaboration (www.biodynamo.org). Tight involvement
with the collaboration will support the student in obtaining training in
the usage of modern computational biology methodologies. Moreover, the
student will benefit from weekly seminars and daily interactions with
computational and experimental researchers, as well as the research groups
of the supervisors (https://in-silico-modelling.ucy.ac.cy/ and
www.combynelab.com).


*Supervisory team*

The project will be supervised by Dr Roman Bauer (r.bauer at surrey.ac.uk) and
Prof. Vasileios Vavourakis (vavourakis.vasileios at ucy.ac.cy). Interested
candidates are encouraged to contact either one of the supervisors in case
of questions.

Best regards,

Roman








*Roman Bauer, Ph.D.Senior LecturerBioDynaMo SpokespersonComputer Science
Research CentreUniversity of SurreyGuildford, UK*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20240410/a5a136ba/attachment.html>