Molecular evolution provides the theoretical foundation for interpreting sequence comparisons. Substitution models (JC69, K2P, GTR) formalize how DNA sequences change over time, accounting for multiple hits at the same site. The distinction between orthology (divergence by speciation) and paralogy (divergence by gene duplication) determines whether sequence similarity reflects shared species history or gene family expansion. The dN/dS ratio (nonsynonymous to synonymous substitution rate) identifies genes under purifying selection, neutral drift, or positive selection. These concepts underpin every comparative and phylogenetic analysis in genomics.
Calculate raw percent identity between two homologous sequences, then apply a Jukes-Cantor correction and compare the two distance estimates. The difference — which grows dramatically for divergent sequences — demonstrates why substitution models matter. Then compute dN/dS for a pair of orthologous coding sequences using an online tool.
Every time you compare two sequences and ask "are these related?" or "how has this gene changed?", you are implicitly relying on molecular evolution theory. This topic covers the core concepts that make sequence comparison biologically meaningful rather than purely computational.
Substitution models are mathematical descriptions of how DNA sequences change over time. The simplest, Jukes-Cantor (JC69), assumes all four nucleotides are equally frequent and all substitutions are equally likely. This gives a clean formula for converting observed percent differences into estimated evolutionary distance, correcting for the multiple-hit problem. More realistic models add complexity: Kimura's two-parameter model (K2P) distinguishes transitions from transversions (transitions are more common); the General Time-Reversible model (GTR) allows each of the six substitution types to have its own rate and accommodates unequal base frequencies. Choosing the right model matters because underparameterized models underestimate distances between divergent sequences, distorting phylogenetic trees and divergence time estimates.
The concepts of homology, orthology, and paralogy are central to comparative genomics. Two sequences are homologous if they share a common ancestor. Orthologs diverged by speciation — they are the "same" gene in different species. Paralogs diverged by gene duplication — they are "sibling" genes within a genome (or across genomes if the duplication preceded speciation). This distinction matters because orthologs tend to preserve function (the gene does the same job in mouse and human), while paralogs are more likely to have diverged in function (one copy may take on a new role). Incorrectly treating paralogs as orthologs leads to wrong functional predictions, which is why reciprocal best BLAST hits and more sophisticated orthology assignment tools (OrthoFinder, OMA) are critical in comparative studies.
The dN/dS ratio connects molecular evolution to natural selection in a quantifiable way. By comparing the rate of amino acid-changing substitutions (dN) to the rate of silent substitutions (dS) in protein-coding genes, you can infer the selective pressure acting on the protein. Most genes show dN/dS well below 1 because most amino acid changes are deleterious and removed by purifying selection. A ratio near 1 suggests the protein is evolving without constraint — perhaps a pseudogene or a gene that has become dispensable. A ratio above 1 is the signature of positive selection, meaning amino acid changes are being actively favored. Genome-wide dN/dS scans have identified genes under positive selection across many lineages, including genes involved in immunity, reproduction, and sensory perception.