Web Extras: Article

Computing Evolutionary Trees Using Ancient Molecules
By Emily Carlson
Posted May 28, 2008

The next time you bite into a chicken nugget, know that you’re eating a close relative of the once mighty Tyrannosaurus rex.

Paleontologists studying structural similarities among bones have long said that dinosaurs are more closely related to modern birds than to reptiles. Molecular biologists now have backed up this hypothesis using sophisticated computational tools to compare proteins from 21 living organisms to fragments of a 68 million-year-old collagen protein preserved in a T. rex femur.

Chris Organ

"The placement of T. rex with the chicken isn’t new," says Chris Organ, a Harvard University postdoctoral fellow who did part of the analysis. "But this work does demonstrate that it’s possible to salvage biomolecules from animals that died millions of years ago and retrieve meaningful evolutionary information."

To study the evolutionary links between organisms across time, researchers construct phylogenies—family trees that show varying degrees of relatedness. For centuries, these diagrams were based mainly on visible traits, such as bone structure, wings, or the presence of fur.

As more tools for studying DNA have become available, evolutionary trees have expanded to include genetic traits. Called molecular phylogenies, these charts enable scientists to track differences between and within species—from changes in flu virus strains to the migration patterns of humans—and, as shown most recently, use ancient molecules to confirm evolutionary hypotheses.

"It’s a golden age for biology, and what’s making this possible is technology," says Organ, who traded dinosaur digs to do work in a lab. Modern phylogenetics, he adds, relies heavily on computer-aided analysis and, without it, would be less powerful and informative.

Phylogeny showing evolutionary relationships of major vertebrate groups based on similarities in a protein.
Credit: Chris Organ. Click for larger image.

To construct their phylogenies, Organ and his colleagues start with a table of traits for different species. For the T. rex study, these traits included more than 2,000 amino acids of sequence data per protein. Computers search for similarities and use them to infer evolutionary relationships, resulting in millions of possible phylogenies. The researchers base their conclusions on the average of the possibilities.

While the resulting phylogenetic tree is built from the best available data, it isn’t necessarily conclusive, says Organ. As with any scientific inquiry, including new information could change the outcome.

Each tree, though, does offer new clues as to why certain traits, including diseases, have emerged. "The basic processes of evolution are inherently the same for all organisms," says Susan Haynes, a genetics expert at NIGMS. "Every study, whether it focuses on dinosaurs, reptiles, or mammals, gives us one more piece to the puzzle, revealing a more detailed picture of our evolutionary tree."

Back to Top