SIMBER

Simula ~ UC Berkeley ~ Berkeley Lab

Collaborators

IdentiPhy @ Simula Research Laboratory

IdentiPhy is a RCN (Research Council of Norway) funded project hosted by Simula’s Computational Physiology (ComPhy) department and led by Dr. Sam Wall. It shares major aspects of scientific scope with SIMBER and supports research to develop an integrated experimental and computational pipeline for improving preclinical to clinical translation in cardiac drug safety and efficacy screening. At the centre of this approach are the microphysiological assay systems (MPSs) being developed by the Healy Lab at UC Berkeley and a computational pipeline being developed by Dr. Wall’s group to discriminate (from MPS recordings) probable cardiac targets of novel compounds. This pipeline involves computational models of both single cardiac cells and microtissues, and applies a novel model framework developed by Professor Aslak Tveito, coined the “EMI” model (Extracellular-Membrane-Intracellular).

Department of Computational Physiology

Research Projects

  • Computational inversion from MPS recordings using traditional optimization algorithms
  • Machine learning for parameter estimation and computational inversion
  • The EMI framework for tissue-level models with an explicit representation of individual cells
  • MPS database for data management and analysis

Members

  • Sam Wall (Head of group)
  • Aslak Tveito
  • Karoline Horgmo Jæger
  • Henrik Finsberg
  • Åshild Telle
  • Nick Forsch

Publications

HPC Department @ Simula Research Laboratory

The Department of High Performance Computing (HPC) at Simula is host to the Norwegian eX3 national infrastructure for exascale computing. Leveraging this unique ensemble of hardware the department develops a range of approaches to optimizing code and numerical efficiency for modern applications in scientific computing, and machine learning and artificial intelligence. One such application has come in the form of the EMI model that is a core component of IdentiPhy. This model permits simulation of continuum electrodiffusion (and now mechanics) at user-nominated spatial and temporal resolution. The HPC group has been central in defining parallelization and numerical schemes that have made MPS-scale simulations tractable, approaching routine. As the biological complexity of the model continues to increase, these contributions will only become more important.

Department of High Performance Computing

Research Projects

  • eX3 infrastructure for exascale computing
  • MICROCARD: Numerical modeling of cardiac electrophysiology at the cellular scale
  • SparCity: Ecosystem of software, algorithms, and tools for sparse computations
  • EGNE: Enabling Graph Neural Networks at Exascale, project on extending deep learning methods to unstructured or irregular data

Members

  • Xing Cai (Head of department)
  • Johannes Langguth
  • James Trotter
  • Kristian Hustad
  • Luk Burchard

Publications

Healy Lab @ UC Berkeley

Led by Professor Kevin Healy, the Biomaterials and Tissue Engineering Laboratory is internationally recognized for developments in biological applications of materials science. Particularly in tissue regeneration and development, and more recently in microfluidic tissue organoid systems based on human induced pluripotent stem cell (hiPSC) technology. The latter research efforts have given rise to the cardiac MPS platform that forms the basis for interaction with IdentiPhy and ongoing collaboration between Simula and UC Berkeley.

Biomaterials and Tissue Engineering Laboratory

Research Projects

  • Polymer-based scaffolds for ECM mimicry
  • Microphysiological systems for drug screening and development

Members

  • Kevin Healy (Head of group)
  • Gabby Neiman
  • Lizzy Gill
  • Ishan Goswami
  • Brian Siemons

Publications

Scalable Solvers Group @ Lawrence Berkeley National Lab

The Scalable Solvers Group at LBNL, led by Dr. Xiaoye (Sherry) Li, has developed a number of internationally utilized code libraries for efficient solution of sparse linear systems, and a range of other relevant problems in scientific computing. In particular, the SuperLU and STRUMPACK libraries have seen broad adoption and application across domains of scientific computing. Upon beginning SIMBER the group has collaborated with Dr. Langguth at Simula HPC to develop improved algorithms for parallel LU decomposition of sparse matrices, and through SIMBER the intention is to extend this to broader collaboration with Simula HPC and IdentiPhy.

Scalable Solvers Group
Lawrence Berkeley National Lab

Research Projects

  • Sparse Direct Linear Solvers and Preconditioners
  • Sparse Eigenvalue solvers
  • Optimization and Machine Learning

Members

  • Xiaoye (Sherry) Li (Head of group)
  • Pieter Ghysels
  • Yang Liu
  • Chao Yang

Publications