LIFE SCIENCES

University of Cologne Researchers Use Jülich Supercomputers to Further Understand Schizophrenia

Principal Investigator:
Prof. Dr. Joseph Kambeitz

Affiliation:
University of Göttingen

Local Project ID:
brainsim

HPC Platform used:
JURECA at JSC

Date published:

Pipeline for the execution of the simulations. The pipeline for this work can be divided in two parts: Preprocessing and Simulation. For the preprocessing the team collected MRIs with information about structure and functioning and generated individualized structural and functional connectivity representations. For the simulations, it used the structural connectivity representation along with space of parameters to be explored, containing all possible parameters combinations within a specific range. Then, for each parameter combination, a simulation was performed, resulting in a simulated functional connectivity. Finally, for each simulation, the simulated functional connectivity was compared to the original functional connectivity and the parameter combination that generated the simulation mostly similar to the empirical data was then inferred as the actual parameters from a subject.

Using the JURECA supercomputer, a team of University of Cologne researchers led by Prof. Dr. Joseph Kambeitz is simulating biophysical processes in the brain in pursuit of better understanding what leads to schizophrenia in patients. 

Using the JURECA supercomputer, a team of University of Cologne researchers led by Prof. Dr, Joseph Kambeitz is simulating biophysical processes in the brain in pursuit of better understanding what leads to schizophrenia in patients. The team pointed out that, to date, much of the research into schizophrenia has been done experimentally, and while experiments have alluded to how changes in brain structure, function, and neurochemistry may play a role, none of these hypotheses have been broadly adopted as part of standard clinical practice.

The team has started to simulate brain functions through a series of mathematical representations—so-called “neural mass models.” The team used a computational neuro-platform called “The Virtual Brain,” and is in the midst of analyzing its simulation results. A publication is forthcoming.

For the full project report, click here.