The maximum information principle and the evolution of nucleotide sequences

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Biology

Background:

• Nucleotide sequences exhibit non-random base distributions.
• Understanding these distributions is key to deciphering genetic information and evolutionary processes.

Purpose of the Study:

• To deduce probability distributions of bases in nucleotide sequences using the maximum information principle.
• To investigate the influence of selective constraints on these distributions.

Main Methods:

• Maximizing entropy under constraints (Markovian entropy, G+C content) representing random mutation and selection.
• Developing two formulations based on different selective constraints.
• Comparing theoretical distributions with experimental data.

Main Results:

• Theoretical distributions showed deviations less than 10% from experimental data for most sequences.
• Lagrange multipliers systematically varied across species.
• A correlation was established between selective constraints and evolutionary patterns.

Conclusions:

• The maximum information principle provides a robust framework for predicting nucleotide sequence composition.
• Selective constraints play a significant role in shaping genome evolution.
• Systematic variations in evolutionary pressures can be quantitatively inferred from sequence data.