91av

PHEW! Our genome wasn鈥檛 invaded by genes from bacteria.

Earlier this year, it was claimed that over a hundred bacterial genes had
somehow jumped into the genomes of vertebrates. But Michael Stanhope and his
team at GlaxoSmithKline in Pennsylvania think the claims were based on faulty
analysis. They say their findings should calm fears that genes from genetically
engineered foods or bacteria could easily jump to humans.

When the first analysis of the human genome appeared in February
(91av, 17 February, p 4),
researchers said they鈥檇 found 223 genes that
were common to humans and bacteria but absent from evolutionary links between
bacteria and vertebrates such as yeasts, worms and flies. They suggested that at
least 113 genes had jumped straight from bacteria into vertebrates.

But while bacteria regularly swap genes among themselves, Stanhope was
sceptical that there had been so many transfers to vertebrates. 鈥淥ne or two
would be very interesting,鈥� says Stanhope. 鈥淥ne hundred and thirteen? Incredibly
别虫迟谤补辞谤诲颈苍补谤测.鈥�

Stanhope鈥檚 team picked 28 of the genes and used computer tests to study their
evolution and relation to other genes more closely. The team also ran tests on a
database the original study had overlooked.

They found no evidence of bacterial genes suddenly crossing to humans.
Instead, they found that invertebrates such as flies and worms had similar
genes, showing that all animals inherited them from a common ancestor after all.
But there has been traffic the other way鈥攖hey found evidence that one gene
has jumped from humans to bacteria.

Other researchers have also cast doubt on the original bacterial transfer
study (Science, vol 292, p 190), highlighting problems with the
computational methods used.

鈥淚t鈥檚 a bit symptomatic of the bioinformatics field at the moment鈥攊t鈥檚
sometimes missing the biological perspective,鈥� says Stanhope. 鈥淲e need to keep
in mind that bringing order to genomic sequences is a biological issue.鈥�