Job Offer: Postdoctoral Researcher in Machine Learning for Large-Scale Brain Network Models
Institute: Institut de Neurosciences des Systèmes, Aix-Marseille University, France Website: http://ins-amu.fr/
Summary:
The Theoretical Neuroscience Group (TNG), led by Viktor Jirsa, invites applications for a postdoctoral position within the context of the Virtual Brain Twin initiative, focused on personalized treatments for Psychiatric Disorders. This research program is coordinated by EBRAINS AISBL, a major player in the European digital neuroscience research infrastructure (https://www.ebrains.eu/). The core of this innovative project revolves around the Virtual Brain Twin platform, which leverages big data, multiscale modeling, and high-performance computing. It will be seamlessly integrated into the EBRAINS ecosystem.
As a postdoctoral researcher, the candidate will be responsible for developing statistical and machine learning methods tailored to construct parameter inference workflows for connectome-based large-scale brain network models. These models are part of the framework of The Virtual Brain (http://www.thevirtualbrain.org) and are applied to various brain imaging modalities, including EEG, SEEG, MEG, and fMRI. Specifically, the project will involve the application and evaluation of Bayesian estimation techniques for high-dimensional biophysical and phenomenological time-series models based on coupled ordinary differential equations (ODEs) and stochastic differential equations (SDEs) with nonlinear latent state-space variables. This research builds on prior successful applications, such as estimating brain excitability in personalized models and analyzing SEEG recordings of seizure propagation in epileptic patients (Wang et al, Sci Transl Med. 2023. DOI: 10.1126/scitranslmed.abp8982 & Jirsa et al., 2023. DOI: 10.1016/S1474-4422(23)00008-X ).
A key component of this project is inverting in-vitro and in-vivo datasets within a Bayesian framework for brain network models, providing posterior distributions of inferred parameters, and addressing issues related to model degeneracy, multimodality, and non-identifiability, using state-of-the-art Bayesian inference techniques. The successful candidate will join a dedicated team working to extend these methodologies to other paradigms, including stimulation and resting state.
Required Qualifications:
PhD (or equivalent) in computer science, physics, statistics, computational neuroscience, engineering, or related discipline.
Experience in one or more of the project research topics, including MCMC sampling, Simulation-based inference, and Dynamical Causal Modeling (preferably with related peer-reviewed publications).
Proficiency in programming with Python or a similar language, with expertise in scientific computing packages.
Proficiency in spoken and written English.
Desired Qualifications:
Experience in a research or academic environment, or a background in neuroscience.
Familiarity with Jax/Pytorch and/or Probabilistic Programming tools such as Stan.
Experience running parallelized large-scale simulations on supercomputers.
Expertise in data science, data visualization, and deep learning algorithms.
Qualifications:
Successful candidates should have a robust background in machine learning, particularly in Bayesian inference. Prior research experience in computational neuroscience, specifically related to networks and dynamic system theory, is a valuable asset. Familiarity with programming in Jax/Pytorch is preferred, and candidates should possess strong programming experience with languages such as Python, Julia, or C++.
About The Theoretical Neuroscience Group:
We are a multinational and interdisciplinary team dedicated to unraveling the spatiotemporal organization of large-scale brain networks. Our work encompasses mathematical and computational modeling of large-scale network dynamics, the analysis of human brain imaging data, and the development of neuroinformatics tools for studying large-scale brain networks, particularly in the context of brain disorders such as epilepsy, Alzheimer's disease, and multiple sclerosis, and healthy aging.
Terms of Salary and Employment:
(Include relevant information on salary, benefits, and other employment terms.)
The initial appointment is for one year, with the potential for extension up to three years.
Application Deadline:
The position will remain open until filled.
Starting Date:
January 2023
How to Apply:
Interested candidates are encouraged to submit their applications, including a cover letter, curriculum vitae, and contact information for references, to giovanna.RAMOS-QUEDA@univ-amu.fr and lisa.otten@univ-amu.fr. Please specify "Postdoc Application - Machine Learning for Brain Network Models" in the subject line.
We look forward to welcoming a highly motivated and qualified postdoctoral researcher to our team to advance our understanding of large-scale brain network models in the context of neuroscience and psychiatric disorder research.
UMR 1106 – Institut de Neurosciences des Systèmes INS Faculté de Médecine de la Timone
27 Bd Jean Moulin - 13385 Marseille cedex 05. France
Tel : + 33 (0) 4 91 32 42 21 ou 23 | Fax : + 33 (0) 4 91 78 99 14