Jordi Piñero,Damian R. Sowinski,Gourab Ghoshal,Adam Frank,Artemy Kolchinsky
Abstract:We investigate minimal replicator systems that are able to use information in a functional manner. Specifically, we consider a population of autocatalytic replicators in a flow reactor, subject to fluctuating environments. We derive operational bounds on replicators production in terms of information-theoretic quantities, reflecting contributions from environmental uncertainty, side information, and distribution mismatch. We also derive the optimal strategy, expressed as a function of both intrinsic replicator parameters and environmental statistics. We compare and contrast our findings with existing information-theoretic formalisms such as Kelly gambling. The results are illustrated on a model of real-world self-assembled molecular replicators. For this system, we demonstrate the benefit of internal memory in environments with temporal correlations, and we propose a plausible experimental setup for detecting the signature of functional information. We briefly discuss the role that information processing may play in guiding the evolution of prebiotic replicator networks.
What problem does this paper attempt to address?
The core problem that this paper attempts to solve is: **In a fluctuating environment, how can self - replicating systems use information to improve their functional performance (i.e., replication efficiency)?** Specifically, the author studied the population of autocatalytic replicators in a flow reactor, and these replicators were exposed to a fluctuating environment. The main objectives of the paper are:
1. **Quantify the relationship between information and function**: By introducing concepts in information theory, such as mutual information, entropy, and KL divergence, the author derived operational bounds on replicator yield. These bounds reflect the effects of environmental uncertainty, side information, and distribution mismatch on replicator yield.
2. **Determine the optimal strategy**: The author derived the optimal initial conditions (i.e., the initial proportion of replicators) to minimize distribution mismatch and maximize yield. The optimal strategy depends on the intrinsic properties of replicators (such as replication rate) and environmental statistical characteristics.
3. **Verify theoretical predictions**: By applying a practical self - assembling photocatalytic replicator system, the author demonstrated the role of internal memory as a source of side information in a temporally correlated environment and verified the consistency between theoretical predictions and numerical simulation results.
4. **Propose an experimental scheme**: The author proposed a possible experimental setup for detecting the characteristic signature of functional information (i.e., information that leads to an increase in yield) in a minimal replicator system.
### Formula summary
- **Definition of productivity**:
\[
P := \frac{1}{\tau} \int_0^\tau \phi X(t) \, dt
\]
where \(P\) is the time - averaged replicator outflow rate, with units of mass concentration/time.
- **Steady - state productivity**:
\[
P^* := \phi x_r^* = \phi \left( \mu - \frac{\phi}{\eta_r} \right)
\]
- **Production loss**:
\[
L := \int_0^\tau \phi (X^* - X(t)) \, dt
\]
- **Average productivity**:
\[
\langle P \rangle = \frac{1}{\tau} \sum_{\epsilon, y} p_{\epsilon, y} \tau_\epsilon P_{\epsilon, y}
\]
- **Information - theoretic bounds**:
\[
\langle P \rangle = \langle P^* \rangle - \gamma - \Omega C_{\pi, q}(R|Y)
\]
where,
\[
\gamma := \frac{1}{\tau} \sum_{\epsilon, y} p_{\epsilon, y} \frac{\phi_\epsilon}{\eta_r(\epsilon)} \ln \frac{X^*_\epsilon}{X(0|y)}
\]
\[
C_{\pi, q}(R|Y) = -\sum_{r, y} \pi_{r, y} \ln q_{r|y}
\]
- **Optimal strategy**:
\[
\pi_{r, y} := \frac{1}{\Omega \tau} \sum_{\epsilon: r(\epsilon) = r} p_{\epsilon, y} \frac{\phi_\epsilon}{\eta_r(\epsilon)}
\]
These formulas and derivations reveal how replicator systems can use information to optimize their yield in a fluctuating environment and provide a theoretical framework for understanding and predicting this phenomenon.