Graphs in sequence spaces: a review of statistical geometry

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Genetics

Background:

• Comparative sequence analysis is crucial for understanding gene evolution.
• Existing methods may rely on specific evolutionary models.
• Novel approaches are needed to analyze sequence data without prior assumptions.

Purpose of the Study:

• To introduce and detail the method of statistical geometry for gene sequence analysis.
• To demonstrate its capability in assessing evolutionary relationships and substitution rates.
• To apply statistical geometry to diverse biological sequences, including tRNA, homeoboxes, and viral families.

Main Methods:

• Statistical geometry computes sequence set geometries (e.g., quartets) using vertical and horizontal information.
• It analyzes sequence data within a conceptual 'sequence space'.
• The method does not require a priori evolutionary models.

Main Results:

• Statistical geometry quantifies the 'tree-likeness' of sequence data.
• It effectively detects varying positional substitution rates within sequences.
• Applications provided an age assessment for the genetic code using tRNA sequences.
• Reliable kinship relationships were assigned for homeoboxes and viral families.
• A lower bound for the age of the common ancestor of human and simian immunodeficiency viruses was established.

Conclusions:

• Statistical geometry offers a model-agnostic approach to comparative sequence analysis.
• It provides robust insights into evolutionary history, genetic code age, and viral relatedness.
• This method enhances our understanding of molecular evolution and phylogenetic inference.