Store:QLMes16

9. Epigenetic evolution within theory of open quantum systems

In paper (Asano et al., 2012b), a general model of the epigenetic evolution unifying neo-Darwinian with neo-Lamarckian approaches was created in the framework of theory of open quantum systems. The process of evolution is represented in the form of adaptive dynamics given by the quantum(-like) master equation describing the dynamics of the information state of epigenome in the process of interaction with surrounding environment. This model of the epigenetic evolution expresses the probabilities for observations which can be done on epigenomes of cells; this (quantum-like) model does not give a detailed description of cellular processes. The quantum operational approach provides a possibility to describe by one model all known types of cellular epigenetic inheritance.

To give some hint about the model, we consider one gene, say $g$. This is the system $S$ in Section 8.1. It interacts with the surrounding environment  $\varepsilon$ a cell containing this gene and other cells that send signals to this concrete cell and through it to the gene $g$. As a consequence of this interaction some epigenetic mutation $\mu$ in the gene $g$ can happen. It would change the level of the $g$-expression.

For the moment, we ignore that there are other genes. In this oversimplified model, the mutation can be described within the two dimensional state space, complex Hilbert space ${\mathcal {H}}_{epi}$ (qubit space). States of $g$ without and with mutation are represented by the orthogonal basis $|0\rangle$,$|1\rangle$; these vectors express possible epigenetic changes of the fixed type $\mu$.

A pure quantum information state has the form of superposition$|\psi \rangle _{epi}=c_{0}|0\rangle +c_{1}|1\rangle$.

Now, we turn to the general scheme of Section 8.2 with the biological function $F$  expressing $\mu$-epimutation in one fixed gene. The quantum Markov dynamics (24) resolves uncertainty encoded in superposition $|\psi \rangle _{epi}$ (“modeling epimutations as decoherence”). The classical statistical mixture , ${\rho }_{steady}$see (30), is approached. Its diagonal elements $p_{0},p_{1}$give the probabilities of the events: “no $\mu$-epimutation” and “$\mu$-epimutation”. These probabilities are interpreted statistically: in a large population of cells,  $M$ cells,$M\gg 1$ , the number of cells with $\mu$-epimutation is $N_{m}\approx p_{1}M$. This $\mu$-epimutation in a cell population would stabilize completely to the steady state only in the infinite time. Therefore in reality there are fluctuations (of decreasing amplitude) in any finite interval of time.

Finally, we point to the advantage of the quantum-like dynamics of interaction of genes with environment — dynamics’ linearity implying exponential speed up of the process of epigenetic evolution (Section 8.4).

10. Connecting electrochemical processes in neural networks with quantum informational processing

As was emphasized in introduction, quantum-like models are formal operational models describing information processing in biosystems. (in contrast to studies in quantum biology — the science about the genuine quantum physical processes in biosystems). Nevertheless, it is interesting to connect the structure quantum information processing in a biosystem with physical and chemical processes in it. This is a problem of high complexity. Paper (Khrennikov et al., 2018) presents an attempt to proceed in this direction for the human brain — the most complicated biosystem (and at the same time the most interesting for scientists). In the framework of quantum information theory, there was modeled information processing by brain’s neural networks. The quantum information formalization of the states of neural networks is coupled with the electrochemical processes in the brain. The key-point is representation of uncertainty generated by the action potential of a neuron as quantum(-like) superposition of the basic mental states corresponding to a neural code, see Fig. 1 for illustration.

Consider information processing by a single neuron; this is the system  $S$ (see Section 8.2). Its quantum information state corresponding to the neural code quiescent and firing, $0/1$, can be represented in the two dimensional complex ${\mathcal {H}}_{neuron}$ Hilbert space  (qubit space). At a concrete instant of time neuron’s state can be mathematically described by superposition of two states, labeled by  $|0\rangle$,$|1\rangle$: $|\psi _{neuron}\rangle =c_{0}|0\rangle +c_{1}|1\rangle$. It is assumed that these states are orthogonal and normalized, i.e., $\langle 0|1\rangle =0$ and$\langle \alpha |\alpha \rangle =1$, $\alpha =0,1$. The coordinates  $c_{0}$ and  $c_{1}$ with respect to the quiescent-firing basis are complex amplitudes representing potentialities for the neuron $S$  to be quiescent or firing. Superposition represents uncertainty in action potential, “to fire” or “not to fire”. This superposition is quantum information representation of physical, electrochemical uncertainty.

Let $F$ be some psychological (cognitive) function realized by this neuron. (Of course, this is oversimplification, considered, e.g., in the paradigm “grandmother neuron”; see Section 11.3 for modeling of $F$ based on a neural network). We assume that $F=0,1$ is dichotomous. Say $F$ represents some instinct, e.g., aggression: “attack” $=1$, “not attack” $=0$.

A psychological function can represent answering to some question (or class of questions), solving problems, performing tasks. Mathematically $F$ is represented by the Hermitian operator ${\widehat {F}}$  that is diagonal in the basis $|0\rangle$,$|1\rangle$. The neuron $S$ interacts with the surrounding electrochemical environment $\varepsilon$. This interaction generates the evolution of neuron’s state and realization of the psychological function $F$. We model dynamics with the quantum master equation (24). Decoherence transforms the pure state $|\psi _{neuron}\rangle$ into the classical statistical mixture (30), a steady state of this dynamics. This is resolution of the original electrochemical uncertainty in neuron’s action potential.

The diagonal elements of ${\widehat {\rho }}_{steady}$ give the probabilities with the statistical interpretation: in a large ensemble of neurons (individually) interacting with the same environment $\varepsilon$ , say  $M$ neurons,$M\gg 1$ , the number of neurons which take the decision  $F=1$ equals to the diagonal element $p_{1}$.

We also point to the advantage of the quantum-like dynamics of the interaction of a neuron with its environment — dynamics’ linearity implying exponential speed up of the process of neuron’s state evolution towards a “decision-matrix” given by a steady state (Section 8.4).