What it means to be human

Approximately six per cent of human and chimp genes are unique to those species, report scientists from the University of Bristol and three other institutions.

The new estimate takes into account something that other measures of genetic difference do not – the genes that are no longer there. The research is reported in the inaugural issue of Public Library of Science ONE (Dec. 2006). The team studied "gene families" that are shared by humans, common chimpanzees (Pan troglodytes), mice, rats and dogs. Gene families are sets of genes in every organism's genome that are similar (or identical) because they share a common origin.

"After surveying gene families that are common to both humans and chimps, we observed in the human genome a significant increase in the duplication of various genes, including some that influence brain functions. This may provide new information about what it means to be human," says Nello Cristianini, Professor of Artificial Intelligence at Bristol University, who statistically analysed the data with specially created software tools.

The results support mounting evidence that the simple duplication and loss of genes has played a bigger role in our evolution than changes within single genes.

Cristianini and his research partners examined 110,000 genes in 9,990 gene families that are shared by humans, common chimpanzees, mice, rats and dogs. They found that 5,622, or 56 per cent, of the gene families they studied from these five species have grown or shrunk in the number of genes per gene family.

The researchers paid special attention to gene number changes between humans and chimps. Using a new statistical method developed by Tijl De Bie, University of Bristol, and Cristianini, the international team inferred humans have gained 689 genes (through the duplication of existing genes) and lost 86 genes since diverging from their most recent common ancestor with chimps. Including the 729 genes chimps appear to have lost since their divergence, the total gene differences between humans and chimps was estimated to be about 6 percent. The team included computational biologists from the University of Indiana and University of California, Berkeley.

The results do not negate the commonly reported 1.5 percent nucleotide-by-nucleotide difference between humans and chimps. But they do illustrate there isn't a single, standard estimate of variation that incorporates all the ways humans, chimps and other animals can be genetically different from each other.

Any measure of genetic difference between humans and chimps must therefore incorporate both variation at the nucleotide level among coding genes and large-scale differences in the structure of human and chimp genomes. Cristianini commented, "So the question biologists now face is not which measure is correct but rather which sets of differences have been more important in human evolution."

The finding complements reports by University of Colorado and University of Michigan researchers in the journals Science and PLoS Biology earlier this year, in which researchers showed that both gains and losses of individual genes have contributed to human divergence from chimpanzees and other primates.