An informational transition in conditioned Markov chains: Applied to genetics and evolution

Area of Science:

• Population genetics
• Evolutionary dynamics
• Theoretical biology

Background:

• Population states are often known at discrete time points, governed by stochastic processes like Markov chains.
• Understanding intermediate population behavior between observations is crucial for evolutionary inference.
• The Wright-Fisher model is a fundamental tool for studying genetic drift and allele frequency changes.

Purpose of the Study:

• To investigate the validity of assuming population behavior is constrained by observed states at two time points.
• To analyze the population dynamics between two known observation times.
• To determine the conditions under which intermediate population states are independent of initial and final observations.

Main Methods:

• Theoretical analysis of population dynamics governed by Markov chains.
• Derivation of a characteristic time interval based on population properties.
• Mathematical modeling of allele frequency changes over time.

Main Results:

• The assumption that observed states constrain all intermediate population behavior has limited validity.
• When the time interval exceeds a characteristic value, intermediate population dynamics show reduced or no dependence on observed states.
• An equilibrium-like distribution applies to the population for a range of intermediate times beyond this characteristic interval.

Conclusions:

• Knowledge of population states at two time points does not always allow for robust inference of intermediate states.
• A characteristic time interval dictates the extent to which past observations constrain future population behavior.
• The findings impact interpretations of evolutionary history and genetic inferences from limited temporal data.