Model for the distributions of k-mers in DNA sequences

Area of Science:

• Genomics and Evolutionary Biology
• Bioinformatics and Computational Biology

Background:

• DNA sequence composition varies across species and evolutionary timescales.
• K-mer frequency distributions are a fundamental characteristic of genomic sequences.
• Understanding these distributions can provide insights into evolutionary processes.

Purpose of the Study:

• To investigate evolutionary features by analyzing k-mer distributions in diverse organisms.
• To classify organisms into evolutionary groups based on their DNA k-mer patterns.
• To develop a model explaining observed k-mer distribution characteristics.

Main Methods:

• Analysis of DNA sequences from various organisms representing different evolutionary periods.
• Classification of organisms into three groups: (a) E. coli, T. pallidum; (b) yeast, zebrafish, A. thaliana, fruit fly; (c) mouse, chicken, human.
• Examination of the distribution of 6-mers (hexanucleotide sequences) within these groups.

Main Results:

• Organism groups exhibited distinct 6-mer distribution patterns: unimodal for group (a), unimodal with shifted peaks for group (b), and bimodal for group (c).
• The bimodal distribution observed in group (c) (mammals and birds) was particularly notable.
• A model based on cytosine-guanine (CG) content was introduced to explain the bimodal k-mer distributions.

Conclusions:

• K-mer distribution patterns in DNA sequences reflect evolutionary relationships and periods.
• The CG content model effectively explains the bimodal k-mer distributions found in more evolved organisms.
• This study provides a quantitative approach to understanding DNA sequence evolution.