Gene regulatory networks versus Neural networks and Bipartite Graphs: what we can learn on cellular (re)programming 

Statistical Physics and Complexity Group meeting

Gene regulatory networks versus Neural networks and Bipartite Graphs: what we can learn on cellular (re)programming 

  • Event time: 10:00am until 11:00am
  • Event date: 23rd June 2021
  • Speaker: (King's College London)
  • Location: Online - see email.

Event details

Cell differentiation is one of the most fascinating areas of developmental biology. This was long thought to be an irreversible process, however it has been shown recently that it is possible to reprogramme fully differentiated cells into a state which strongly resembles embryonic stem cells, via the introduction of a few transcription factors. This opens up exciting perspective, however, no universally accepted theory exists that explains the phenomena. The purpose of this work is to drive forward our understanding of cell reprogramming and programming by using tools from statistical mechanics.  

In the first part of the talk we present a model for gene expression dynamics inspired by neural networks. Cell types are modelled as hierarchically organized dynamical attractors of the gene expression dynamics and reprogramming is rationalised as triggering transitions between attractors laying at different levels of the hierarchy. We found two mechanisms for such switching, induced by noise and direct perturbations, which offer interesting perspectives on reprogramming experiments.  

In the second part of the talk, the mechanism for the effective interactions arising between genes, is studied by means of a directed bipartite graph model, that integrates the genome and transcriptome into a single regulatory network, evolving according to the AND logic dynamics. By adapting percolation theory to directed bipartite graphs, evolving according to the AND logic dynamics, we are able to determine the necessary conditions, in the network parameter space, under which sparse bipartite networks can support a multiplicity of stable gene expression patterns, under noisy conditions, as required in stable cell types. In particular, the analysis reveals the possibility of a bi-stability region, where the extensive percolating cluster is or is not resilient to perturbations, and it provides valuable insights for the interpretation of gene knock-out experiments. 

[1]   Percolation on the gene regulatory, Giuseppe Torrisi, Reimer Kühn, Alessia Annibale, J. Stat. Mech. (2020) 083501  

[2]   Percolation in Gene Regulatory Networks and its role in sustaining life, R Hannam, R, Kühn, A Annibale, J. Phys. A: Math. Theor. 52 334002 (2019) 

[3]   Cell reprogramming modelled as transitions in a hierarchy of cell cycles, R Hannam, A Annibale, R Kühn J. Phys. A: Math. Theor. 50 425601 (2017)