new genetics database announced

[Guest guest]

Database lists human genes - International effort creates resource for

geneticists.

April 20 2004 HELEN PEARSON (Nature News)

Researchers have compiled a comprehensive catalogue of over 21,000 human

genes: as many as three-quarters of the total number of genes thought to

be in our genome.

Experts say the catalogue, called the Human Full-length

Complementary-DNA Annotation Invitational Database (H-Invitational

Database), will help geneticists identify what each gene does in the

human body, including their contribution to certain diseases.

" It is a tremendous resource, " says Len Pennacchio who studies human

genomics at Lawrence Berkeley National Laboratory in California.

Although researchers have the human genome sequence largely in hand,

they still struggle to work out which sections are genes, the stretches

of code that provide instructions for making proteins, and which are

simply filler. Computer programs that try to predict which bits

constitute genes are notoriously unreliable.

Details of the H-Invitational Database, which was created by an

international team led by Takashi Gojobori of the Center for Information

Biology and the DNA Data Bank of Japan in Mishima, are reported in the

Public Library of Science Biology1.

Analysis of the gene set has already thrown up some interesting

findings. It seems, for example, that the sequences at the beginning and

end of genes tend to be longer than those in the middle, although no one

yet knows why. " There's something funky going on at the front and back

of genes, " says Ewan Birney of the European Bioinformatics Institute in

Hinxton, Britain.

Work in progress

The team made the catalogue by pooling six existing collections of

molecules called complementary DNAs (cDNAs). These molecules can only be

made from active genes.

From a list of over 41,000 cDNAs, the researchers eliminated duplicates

and suspect molecules. About 120 collaborators met in Tokyo in the

summer of 2002 to do much of this sifting.

The resulting database contains over 21,000 genes, along with

information about where they are active in the body, predictions of the

proteins they make, and information about the individual variations in

genes that help to distinguish one person from another.

Researchers can collect information from the database, or they can

actually order many of the cDNAs by post.

The project has taken three years so far. But it is very much a work in

progress, the researchers note. " It will read differently in six

months, " Pennacchio says.

References

Imanishi, T. et al. PLoS Biol, e162, doi:10.1371/journal.pbio.0020162

(2004).