Over the course of 3.7 billion years of protein evolution, protein segments mutated, duplicated, and recombined. We can syllogize about the evolutionary process and its early stages by studying its result: present-day proteins. Indeed, contemporary proteins hold traces of the “historical” mutations, duplications, and recombinations that shaped them. In particular, sequence similarities among proteins offer valuable clues, as these are a hallmark of shared ancestry; that is, random sampling of a specific sequence is such a low-probability event, that it is unlikely that two similar long sequence segments formed independently. Most of the observed homology among current-day proteins may reflect relatively recent events, but there is also hope of finding protein segments that are relics of primordial, ancestral peptides. As Eck and Dayhoff highlighted in their seminal 1966 paper, these may be in conserved and functionally critical parts of contemporary proteins, because “natural selection which severely inhibits any change to an (ancient) well-adapted system on which several other essential components depend”.<br/><br/>We develop computational tools to search for traces of ancient events and to study patterns that govern protein evolution. We define and use ‘themes’ – homologous segments that recur, or are ‘reused,’ in multiple proteins in the structural databases. While themes can be found at various scales, we focus on ones at the sub-domain-scale, as these are candidate traces of more ancient events. Particularly interesting are themes that are reused across dissimilar contexts, that is, where the rest of the protein domain differs. I will describe how we use these to reveal evolutionary patterns across the protein universe, connecting proteins from evolutionary lineages that were assumed to be separate. Among these are links between ancient lineages that date back to LUCA, and even one connecting the Rossmann and P-loop proteins. Our tools also identify more recent events, allowing us to trace the emergence of the beta-Trefoil proteins. Finally, I will describe future challenges in studying the protein universe from a global perspective, focusing on the promising potential of the new deep network technologies, offering a path to more accurate perspectives and means to incorporate the vast sequence databases into our view.

Evolution-CSI: Hunting for evidence in the protein universe

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.