Olivia Eriksson

I am a reasercher working at Science for Life Laboratory on modelling signaling pathways of synapses in the brain. In particular I am interested in how to combine data driven methodology with more classical hypothesis driven modelling. An important aspect of this is uncertainty quantification and sensitivit analysis.

I am also coordinating the Swedish e-Science Research Center (SeRC).

Mypublications can be found here.

MSc thesis projects

Model extension and model selection of a synaptic plasticity model:

We are interested in further extending a recently developed biochemical model describing a mechanism of synaptic plasticity with new information about interacting molecules. This could be done in several ways corresponding to different hypothesis and model variants. We want to investigate which model variant performs best by first fitting the variants to experimental data using the R software package UQSA and then implementing and performing statistical model selection. The experimental data has been obtained through an ongoing collaboration and additional measurements from this or literature will be integrated as part of this master thesis project.

Required expertise (some of the following): mathematical modeling, parameter estimation, machine learning, programming skills (preferably in R), interest in biochemistry and neurobiology

Collaborators: Lisa Bast (KI), Andrei Kramer(KI) and Federica Milinanni (KTH)

Start date:As early as possible

Structural Identifiability for models of the brain

We are interested in assessing the structural identifiability of mathematical models, i.e. the ability to estimate parameters for a given mathematical model using measured output variables.

Within the scope of this thesis, we are interested in three mathematical models that have already been developed. These include small- and large-scale models; one model describes adult neurogenesis (production of neurons in the mouse brain) and the other two models a molecular mechanism for synaptic plasticity (AKAR4 and AKAP79 dynamics, respectively). The thesis project consists in analyzing the three models with two different structural identifiability algorithms. For this purpose, at least one of the three models and the two structural identifiability algorithms will be implemented in the programming language R. The structural identifiability algorithms will be used to determine if, for a given measurement output variable, all parameters are identifiable. For non-identifiable parameters/models, we aim to determine how such models can be made identifiable.

Required expertise(some of the following): mathematical modeling, parameter estimation, programming skills (preferably in R), interest in biochemistry and neurobiology

Collaborators: Lisa Bast (KI) and Andrei Kramer (KI)

Start date: As early as possible