By Yves Lherry In On March 9, 2015
Randomness during embryogenesis, between mathematics, physics and biology
Abstract: How biological randomness should be described ? In developmental biology this problem is currently strongly discussed. Fate choice of stems cells and cell lineage commitment is a highly variable phenomenon and it is often hard to predict into which cellular type a cell is going to differentiate. During the last years three main approaches have emerged to explain empirical data : a deterministic one, relying on the central dogma  ; a semi deterministic one, giving more flexibility to it and using the concept of attractors  ; and a random one, using the theories of Claude Bernard and Charles Darwin in a convergent ontophylogenesis and rejecting the central dogma as a unidirectional causal process . These three approaches seem to be contradictory but try to explain the same reality, thus there might be a way to unify them or to reduce the ones to the others.
Several formal descriptions of randomness in physics and in mathematics already exist. They however may not be directly applicable to biological objects which rely on other paradigms, including indetermination at different levels of observation (e.g. cellular and global ones) and may involve the superposition of multiple types of randomness.
The proposed project consists in an interdisciplinary exploratory work of conceptualization of the notion of randomness during embryogenesis, on a theoretical level. It takes place in the epistemological frame developed by Bailly and Longo .
It shall begin with a conceptual analysis of the current frameworks used by biologists to describe these phenomena. It will highlight the possible obstacles and give a deep understanding about them (for example the use of a discrete space to represent continuous phenomenon leads to unavoidable misconceptions). The natural continuation of this is a critical comparison with the concepts shown in physics (two major ones exist : deterministic unpredictability in classical systems and intrinsic randomness in quantum mechanics ) and in computer sciences (Martin-Löf gave a definition of random strings , as well as Chaitin, who developed a theory around it called the algorithmic information theory ). Coming out these two parts, it is reasonable to describe a proper biological randomness adapted to the framework of developmental biology.
This work at the interface between disciplines should give a lot of perspectives, beginning with experimental issues for corroborating this conceptual analysis. The possible medical applications could lead to a better understanding of the human embryogenesis and of the origin of cancer. A good description of biological randomness may also give some ideas to understand limit physical phenomena, particularly in the problem of unification of quantum and classical frameworks.
Supervisors: Nadine PEYRIERAS & Guiseppe LONGO
Lab: Unité de Recherche N&D Neurobiologie et Développement
Funding: Région Ile de France