Computational biologist Dr Murray Cox, of the Institute of Molecular BioSciences, worked with colleagues in the United States on a paper published this month in Nature Genetics.
The paper disputes the logical theory that more data is desirable when studying aspects of the human genome. It builds on earlier research the team carried out into the history of human populations. In that study Dr Cox and his colleagues investigated polygyny – men having multiple wives – in human genetic data.
“We sequenced 40 regions of the human genome that were a long way from genes,” Dr Cox says. “We found evidence of polygyny in all six global populations that we studied.”
However, when other researchers used entire genome sequences to conduct the same test, the signal disappeared. “This caused us to stop and think, why is this happening?” he says. “We found that there was too much extraneous information clouding the results.”
Dr Cox says to study demography, a scientist needs to avoid instances of natural selection, which are often found in genes. “For example, the gene for haemoglobin carries oxygen in the blood, and it’s a big target for Malaria. If there is a mutation there that helps resist Malaria, that new variant is going to increase in the population. It’s going to change DNA variation for a long way around that gene, so that’s why we avoided genes.”
Picking sequences from the spaces between genes let the team target only the information they required.
“The human genome has tens of thousands of genes scattered throughout it. The regions we chose to sequence were a long way away from genes so that we’d pick up signals of history and not natural selection.”
Dr Cox says the findings should lead to a change in the way scientists approach this kind of research to achieve more accurate results.