Job ID: 106460

PhD project – Higher-order interactions in human brain networks supporting causal learning

Position: Ph.D. Student

Deadline: 2 April 2023

Employment Start Date: 2 October 2023

Contract Length: 3 years

City: Marseille

Country: France

Institution: Aix Marseille Université

Department: Institut de Neurosciences de la Timone (INT)


The NeuroSchool PhD Program of Aix-Marseille University (France) has launched its annual calls for PhD scholarships for students with a master’s degree in a non-French university.

The following project is one of the 14 proposed projects. Not all proposed projects will be funded, check our website for details.

State of the art. A central hypothesis in neuroscience posits that cognitive functions arise from the coordinated activity of neural populations distributed over large-scale brain networks. Goal-directed learning, which supports the acquisition of causal relations between our behaviors and their consequences, is no exception, and it is thought to emerge from interactions between neural populations distributed over the associative fronto-striatal circuit and limbic “reward” system. Although central, this hypothesis has never been fully tested, yet. Indeed, progress has been limited by the lack of approaches for studying brain interactions beyond pairwise relations, the so-called higher order interactions (HOIs).

Objectives. The aim of the proposed PhD thesis in computational neuroscience is twofold. The first objective is to develop new functional connectivity (FC) approaches to analyze HOIs and their relation with cognitive processes. The second aim is to investigate the role of HOIs in cortical brain networks supporting goal-directed causal learning.

Methods. To achieve the first goal, we will exploit recent advances in information theory and network science. We will test novel measures that allow the characterisation of statistical interdependencies within multiplets of three and more neural signals, such as the O-information, and measures of redundant and synergistic interactions, based on the Partial Information Decomposition framework. Recent network science approaches will also be used to represent and analyze the structure and dynamics of the observed HOIs. To achieve the second goal, we will analyze a unique dataset collected from the BraiNets team that includes cortical high-gamma activities (HGA, from 50-150 Hz) estimated from MEG and intracranial SEEG data collected from human participants performing a causal action-outcome learning task. Model-based analyses of HGA will be used to map learning computations predicted by Bayesian or reinforcement learning models onto cortico-cortical HOIs.

Expected results. We expect to provide a novel FC framework for the analysis of HOIs in cognitive brain networks. We expect to find HOIs in brain circuits mediating goal-directed learning including the parietal, temporal, dorsolateral and dorsomedial prefrontal cortex and orbitofrontal cortical systems. We expect to reveal choice- and outcome-related learning signals (e.g., reward prediction errors) to be mapped onto higher-order cortico-cortical interactions.

Feasibility. The theoretical foundations for the analysis of HOIs are present in the literature and the brain data has already been collected. Although the project has an explorative nature, feasibility is good.

Expected candidate profile. We search for a candidate with a strong interest in computational neuroscience, the study of complex systems and cognitive neuroscience questions related to the neural bases of learning. Prior experience in network neuroscience and Python is a plus.