Unifying pattern encoding and regenerative remodeling in a connectionist cellular network.
Hammelman, Jennifer L.
2016
- Regeneration is a complex process where an organism must programmatically 1) determine what structures are missing and 2) instruct cell proliferation, differentiation, and migration toward the development of complex structures. In contrast to simple wound healing, regeneration of an organ such as a limb is a difficult inverse problem [1]. Bioelectrical communication between non-neural cells is ... read moreknown to have significant control over the development of complex structures, such as inducing ectopic eye development in frog embryogenesis [2]. We propose that the bioelectrical networks that have demonstrated control of patterning in non-neural cells may be better understood through novel models developing from mathematical models of cognition, in particular artificial neural networks (ANN) and reservoir computing methods. In this body of work, we formalize a reservoir computing model of planarian regeneration using positional wound as ANN input and positional tissue information as output and discover that stochastic noise and Hebbian rewiring during training is important for generalization of the response and robustness to perturbation and recapitulate de novo experimental results from planarian regeneration experiments. In closing, we suggest future hypotheses to be tested using this model that can be validated at the bench to improve our understanding of the bioelectrical control of patterning during planarian regeneration.read less
- ID:
- qv33s783b
- Component ID:
- tufts:sd.0000412
- To Cite:
- TARC Citation Guide EndNote
- Usage:
- Detailed Rights
