The mystery of human memory

Computational Biology is an area in which mathematical techniques, and in particular computer simulation techniques, are used to address complex biological problems. In close collaboration with biologists and neurophysiologists Professor Anders Lansner and his research group are trying to solve the mystery of human memory.

Mathematics and computer models have long been used in physics and chemistry, but have now also made its way into biological research. This is a new scientific trend and the eight research groups at the Department of Computational Biology (CB) are very much a part of that process.

Professor Anders Lansner leads the research group Lansner´s lab Computational Neuroscience and Neurocomputing at CB.

"There is clearly a paradigm shift in biology where the quantitative aspect is becoming more and more pronounced. Computer science and applied mathematics will be immensely important in biological research in the future. We see only the beginning,” he says.

The theme of the research in his group is to simulate memory and perception in the cerebral cortex and other parts of the brain of mammals. A prerequisite for the success of the simulations is the availability of powerful computers, and the new supercomputer at PDC was well wished-for. With a supercomputer like that it is possible to simulate hundreds of millions of neurons and many billions of synapses, approaching the cortex size of a small mammal.

"We work with large-scale simulations and have always thought that the computers were too weak. Nowadays we don’t feel that anymore, it's very nice because then we can let our imagination materialize much more,” says Anders Lansner.

He first began studying the mind and memory when he came to KTH in the 80s and then became deeply engaged in modelling of locomotor systems with neuroscientists at Karolinska Institute for twenty years.
"Yes, and we have returned to this area [mind and memory] of research over the past 5-6 years. Now we have much more detailed models and more biological data, and can also mimic the oscillations in the cerebral cortex at different states”, says Anders Lansner.

The research in his group focus on using network-level dynamics to find out which properties of neurons and synapses really matter to specific cognitive functions like the macroscopic phenomenon of memory, for instance the recall from memory. Their work is amongst others performed in collaboration with neurophysiologists at the Karolinska Institute’s Department of Neuroscience and within the Stockholm Brain Institute (SBI) where Anders Lansner is director of the Computational and Modelling platform.

One current aim of the research is to understand what mechanisms at the neural level are associated with particular diseases. Future practical applications of research findings are better understanding of cortex-localized diseases like Alzheimer's and schizophrenia as a basis for developing more efficient therapy including drugs. But with increased knowledge of how the brain works, researchers can also touch on issues adjacent to the ones philosophers are engaged in; who am I, what is consciousness and how does our thought processes work?

"Exactly, and these are questions we begin to get closer to having answers to. We can see if a human test subject discovers a partially hidden stimulus since then they show certain type of oscillations in the cerebral cortex. And these oscillations we can model, and thereby see exactly what caused it.”

The work is done within several European collaborative projects, including the projects FACETS and BrainScaleS, as well as with funding from SSF, VR and Vinnova via SBI.  The European projects encompass everything from cortex neuroscientists right through to designers of neuromorphic hardware. In between the computational neuroscientists do supercomputer brain modelling. “The idea is to integrate all of these competencies so as to build a very high-capacity simulator in hardware to replicate biological neural systems”, tells Anders Lansner.

He adds: "The EU funding has been very important in recent years for the research on brain functions like memory and perception. Furthermore, this research is linked to the development of future computing and hardware. As an example, the Human Brain Project, a proposal for a “blue-sky” moon landing project funded by the Future Emerging Technologies within EU and in which also IBM and Cray are included, aims at in 10 years being able to simulate the human brain all-out on the next-next generation of supercomputers.”

Top page top