With the advance of experimental techniques, molecular cell biology is turning into a quantitative science. Individual cellular processes that comprise the interplay of several molecular players (e.g. gene expression) can now be quantitatively characterized — enabling the use of mathematical models to dissect and understand them and to finally predict new biology based on them.
The main objective of systems biology is to gain a quantitative understanding of the dynamic interplay of molecular agents within a cell. Unlike classical mathematical biology, systems biology is data-driven and the questions addressed and level of abstraction reflect the currently available experimental means. The inverse problem of mapping experimental data to mathematical models is the quintessence of systems biology.
In our group we focus on two particular inverse problems, namely determining the regulatory network structure from data and dissecting the different contributions to cell-to-cell variability observed in single-cell measurements.
Many computational techniques in systems biology have their origin in computational chemistry, in particular classical chemical kinetics. However, the validity of those methods in an in vivo context can be questioned. To address this, we develop more fine-grained spatial simulation algorithms of reaction kinetics to generate a reference standard of the more abstract systems biology models. For this, we can leverage the local expertise in molecular dynamics and HPC at the Cognitive Computing & Computational Sciences department of the IBM Research – Zurich Laboratory.