Multi-scale simulations of synaptic plasticity

In this project, models of subcellular signalling cascades important for synaptic plasticity (see e.g. Nair et al, 2015) will be further developed and then challenged in co-simulations. We will explore how to extract phenomenological and simplified activity- dependent synaptic plasticity rules and neuromodulatory effects by considering multi-scale models of a neural network.

The finest scale model of subcellular cascades will be embedded in a meso-scale whole neuron model, which in turn can be coupled to a coarse scale large network model of spiking neurons. Some of these plasticity models can later be used in the network modelling of working memory in Brain network architecture and dynamics of short- and long-term memory. Efficient numerical software is available for simulating the individual scale models: the signalling cascades, the whole neuron and the network models. The big challenge is how to couple them to simulate the full multi-scale system.

This will be done within the framework of numerical heterogeneous multi-scale methods where the coupling is performed in an efficient,stable and systematic way. The coupling can be either sequential or concurrent, depending on the complexity of the data that needs to be transferred between scale levels. The work will be a collaboration between KI (model building) and KTH (numerical analysis) in interaction with one group in Uppsala as well.

It is a continuation of an earlier informal collaboration between the Brain-IT community and numerical analysis researchers at KTH. Proof of concepts are published (Brocke et al, 2016 and 2017; PhD thesis by Brocke is due autumn 2018). It also builds on the long experience with heterogeneous multi-scale methods in the numerical analysis group. The goal is to use tools such as MUSIC (Djurfeldt et al, 2010) to convey signals between the co-simulated systems during run time. New numerical methods may also be developed.