Scientific highlights 2019

Research focus area – Topological methods for dynamical data
Application – Materials for the future

Research focus area – Multi-scale visualization
Application – Molecular dynamics simulations for conformation analysis

Inviwo is a software framework for rapid prototyping visualizations. It builds the basis for the development of novel visualization research and teaching

Turbulent flow analysis plays a crucial role in many domains e.g. design of fuel-efficient cars and is an active research area primarily addressed through direct numerical simulations (DNS) of the Navier-Stokes equations.

Recently, leadership supercomputers are becoming increasingly heterogeneous. For instance, the two fastest supercomputers in the world, Summit and Sierra, are both equipped with Nvidia V100 GPUs for accelerating workloads.

Floating-point operations are indispensable for many scientific applications. Their precision formats can significantly impact the power, energy consumption, memory footprint, performance, and accuracy

When we study turbulent chaotic flows in complicated geometries our experience could not be sufficient to suggest where a more refined mesh is required.

A series of presentations were given over a span of two days to identify the state-of-the-art of large- eddy simulations (LES)

Scaling individual atomistic simulations of bio-molecules is strong scaling problem, as the molecules have fixed size.

Reasoning about cause and effect is an important aspect of human intelligence, and would thus constitute a valuable part of a medical diagnostics system.

SeRC researchers have assessed the prediction capabilities of long-short-term memory (LSTM) neural networks

We published an article entitled “Artificial intelligence for diagnosis and grading of prostate cancer in biopsies: a population-based, diagnostic study” in Lancet Oncology (https://doi.org/10.1016/S1470-2045(19)30738-7; impact factor=35).

Ontologies standardize terminology in a domain and are a basis for semantically enriching data, integration of data from different databases, and reasoning over the data. They deal with the big data issues of Variety, Variability, and Veracity.

An open-source program named VeloxChem has been developed for the calculation of electronic real and complex linear response functions at the levels of Hartree–Fock and Kohn– Sham density functional theories.

Our spiking neural network attractor memory model accounting for a synaptic theory of working memory has been validated using experimental data to explain the most recent observations about bursty nature of neuronal spiking as well as gamma- and beta-band oscillations of field potentials.

A new macroscopic computational model of brain oscillations in resting-state fMRI data was developed and evaluated. Results show that the effect of local lesions can be captured and studied by making regional changes in oscillation dynamics.

We have demonstrated that molecular-level simulations can inform subcellular models of synaptic plasticity. Several computational tools, from molecular dynamics and Brownian dynamics simulations to bioinformatics approaches, were combined to constrain a kinetic model of the adenylyl cyclase type 5 (AC5)-dependent signaling system in the striatum.