Lineage-specific genes are those exclusively found in one species or group of related species. Their study can shed light into the mechanisms of novel gene formation and into lineage-specific adaptations. We have investigated the mechanisms of formation of primate-specific genes and have found that many of them are likely to have originated de novo from genomic non-coding regions. Recently evolved genes appear to be very plastic and typically display high evolutionary rates. Furthermore, they are enriched in low-complexity sequences. We are currently performing research to further unravel how the age of a gene relates with its mode of evolution.
Low-complexity sequences, including homopolymeric tracts and other short amino acid tandem repeats, are extremely abundant in eukaryotic proteins. These sequences may expand or contract rapidly by the action of replication slippage and/or recombination. We are performing several analysis to learn about the role of natural selection in shaping the LCR content in several vertebrate genomes. We are also investigating the impact of other kinds of short insertions and deletions in the evolution of mammalian proteins.
Gene duplication is an important motor of protein functional diversification. We have investigated the changes in expression patterns of recent duplicated mammalian genes and observed that loss of expression domains is more common than gain of novel expression patterns. We are currently using large gene duplicate sets to investigate how the sequences of initially redundant gene copies progressively diverge and which are the implications for protein function.
We have developed a method to identify genes that show accelerated or decelerated evolution in particular branches of the mammalian phylogeny indicating a shift in the selective regime. We are currently investigating the relative contribution of negative and positive selection in causing lineage-specific deviations in the evolutionary rates. We have examined the effect of protein isoform selection in genome-wide scans of positive selection and shown that selecting isoforms of similar length reduces the fraction of misaligned positions and false positives in tests of selection. The selection of isoforms with the minimum length difference can be performed using our software PALO.