Tuesday, April 15th, 11:00 AM

NCBI Library B2, Bldg. 38A

 

Pavel Novichkov

 

Evolutionary and functional characteristics of ancestry-specific classes of eukaryotic genes 

 

 

Recently, an attempt was undertaken to revise the current catalog of human genes and to assess the extent of gene innovation during mammalian 
evolution. To this end, human genes were partitioned into those that had reliable orthologs in either mouse or dog, and those for which such 
orthologs could not be detected. The great majority of genes from the second category (orphans) were claimed to be artifacts of gene prediction,
 with the corollary that there was little gene innovation in mammalian lineages. These findings prompted us to systematically examine
 the evolutionary and functional features of eukaryotic genes of different evolutionary provenances. Genes from three eukaryotic lineages, 
namely, mammals, insects, and fungi, were classified into ancestry-specific classes, i.e., those genes whose origin could be reliably traced 
back to a particular node in the tree of life. Specifically, for humans, we divided the genes into primate-specific, mammal-specific, 
chordate-specific, metazoan-specific, “fungal-metazoan”, eukaryote-specific, and ancient (conserved in both eukaryotes and arcahea and/or bacteria) ones. 
For these gene classes, we compared the nucleotide sequence evolution rates, the strength of purifying selection 
(non-synonymous to synonymous ratios dn/ds) for orthologs from closely related species, or intraspecies SNPs), protein lengths, 
intron densities, and functional characteristics according to the GO classification.   For all of these properties, we observed consistent 
trends between the ancestry-specific gene classes: “recent” genes encoded smaller proteins, had lower intron density, evolved faster
 and were subject to weaker purifying selection than older genes. There was also a clear functional distinction, with the “young” genes 
encoding, mostly,  proteins involved in specialized functions such as defense and reproduction, and more ancient genes engaged 
in more central, e.g., metabolic roles. In the case of mammals, this trend was very prominent in comparisons of primate-specific, 
mammal-specific, and chordate-specific genes with more ancient ones, but leveled off from metazoans down. Fully consistent results
 were obtained with insect and fungal genes on the respective ancestry-specific classes. Thus, the apparent evolutionary age of 
a gene, traced as far back as the origin of chordates some 600 million years ago, seems to be a good predictor of a gene’s behavior
 in recent evolution, including that within human populations. On the more practical side, we showed that the distribution of nucleotide 
sequence evolution rates for human primate-specific genes could be presented as the sum of two distributions, a relatively slow-evolving 
one fitting the general trend and an anomalously slow-evolving one. We hypothesize that the first of these distributions represents bona 
fide genes involved in specialized functions, whereas the second one encompasses gene prediction artifacts.