SKM 2023 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

BP: Fachverband Biologische Physik

BP 16: Systems Biophysics

BP 16.5: Talk

Wednesday, March 29, 2023, 12:30–12:45, BAR 0106

Sensitivity of Boolean attractors to small network changes and implications for the inference of microbial interaction networks — •Jyoti Jyoti and Marc-Thorsten Hütt — Jacobs university, Bremen, Germany

Sensitivity of dynamics on graphs under small topological changes has been studied for diverse types of dynamics and in relation with a wide range of applications. Here we extend this direction of research by studying (and deciphering) the change of attractors in Boolean threshold dynamics under single and multiple edge switches. By evaluating differences in attractor sets, we can, with high accuracy, predict the structural change in the network. This is of high relevance for network inference, e.g., inferring microbial interactions from abundance patterns: Current approaches, where interaction networks are inferred from attractors [1], often fail to provide networks, which in turn can reproduce the initial attractor set. Evaluating the differences in attractor sets (initial set vs. the one produced by the inferred network) and estimating topological differences from them can pave the way towards better inference algorithms. We briefly discuss the implications of our findings for microbiome analyses.

[1] Claussen, J. C., Skiecevičienė, J., Wang, J., Rausch, P., Karlsen, T. H., Lieb, W., Baines, J. F., Franke, A., and Hütt, M.-Th. (2017). Boolean analysis reveals systematic interactions among low-abundance species in the human gut microbiome. PLoS Computational Biology, 13(6):e1005361.