Coding sequence density estimation via topological pressure.

Area of Science:

• Genomic analysis
• Dynamical systems theory
• Bioinformatics

Background:

• Accurate estimation of coding sequence (CDS) density is crucial for genomic analysis.
• Existing methods may lack the precision needed for complex genomic structures.
• Ergodic theory offers novel mathematical frameworks for biological sequence analysis.

Purpose of the Study:

• To develop a new approach for coding sequence (CDS) density estimation using topological pressure.
• To apply this method for ab initio predictions of CDS density across different species.
• To utilize the derived nucleotide triplet weightings for distinguishing between genomic sequence types (intron/exon).

Main Methods:

• Developed a novel method based on topological pressure from ergodic theory.
• Incorporated 64 parameters representing nucleotide triplet weights.
• Trained parameters on human genome CDS density and applied to predict CDS density in Mus Musculus, Rhesus Macaque, and Drosophila Melanogaster.
• Defined a probability distribution using trained weights to differentiate intron and exon sequences.

Main Results:

• Achieved reasonable estimates for coarse-scale CDS density prediction across tested genomes.
• Demonstrated the ability to distinguish between human intron and exon sequences using the trained probability distribution.
• The method provides a new tool for analyzing genomic sequence composition.

Conclusions:

• Topological pressure offers a robust framework for CDS density estimation in genomic analysis.
• The approach successfully predicts coarse-scale CDS density and aids in sequence classification.
• The underlying theory of Thermodynamic Formalism provides a strong theoretical foundation.