An information theoretic approach to quantifying evolutionary functional trends

The accumulation of genomic and proteomic data of many organisms presents an opportunity to analyze entire phylogenetic trees in a systematic, quantified manner. The Universal Tree of Life, constructed by genomic data, provides an evolutionary context for proteomic data of individual organisms. Using proteomic information mapped to biological functions, we survey the Tree of Life for evolutionary trends, where trends are low p-value linear regressions of relative protein diversity against an evolutionary timeline. These trends provide quantified perspectives across evolution that would be otherwise difficult to discern from the large amount of available bioinformatic data. We analyzed 22,762 functions across 167 proteomes for approximately 3.8 million information content calculations. Of particular interest is the rise of protein-protein interactions along the path of mammalian evolution. Of general interest are the numerous uncovered trends that open the door to future queries of focused scope.