dodo
© Photo Researchers / SPLThe Dodo, Raphus cucullatus, is one of the most well-known extinct animals.

The recipe for making any creature is written in its DNA. So last November, when geneticists published the near-complete DNA sequence of the long-extinct woolly mammoth, there was much speculation about whether we could bring this behemoth back to life.

Creating a living, breathing creature from a genome sequence that exists only in a computer's memory is not possible right now. But someone someday is sure to try it, predicts Stephan Schuster, a molecular biologist at Pennsylvania State University, University Park, and a driving force behind the mammoth genome project.

So besides the mammoth, what other extinct beasts might we coax back to life? Well, it is only going to be possible with creatures for which we can retrieve a complete genome sequence. Without one, there is no chance. And usually when a creature dies, the DNA in any flesh left untouched is soon destroyed as it is attacked by sunshine and bacteria.

There are, however, some circumstances in which DNA can be preserved. If your specimen froze to death in an icy wasteland such as Siberia, or snuffed it in a dark cave or a really dry region, for instance, then the probability of finding some intact stretches of DNA is much higher.

Even in ideal conditions, though, no genetic information is likely to survive more than a million years - so dinosaurs are out - and only much younger remains are likely to yield good-quality DNA. "It's really only worth studying specimens that are less than 100,000 years old," says Schuster.

The genomes of several extinct species besides the mammoth are already being sequenced, but turning these into living creatures will not be easy (see "Revival recipe"). "It's hard to say that something will never ever be possible," says Svante Pääbo of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, "but it would require technologies so far removed from what we currently have that I cannot imagine how it would be done."

But then 50 years ago, who would have believed we would now be able to read the instructions for making humans, fix inherited diseases, clone mammals and be close to creating artificial life? Assuming that we will develop the necessary technology, we have selected 10 extinct creatures that might one day be resurrected. Our choice is based not just on feasibility, but also on each animal's "megafaunal charisma" - just how exciting the prospect of resurrecting these animals is.

Of course, bringing extinct creatures back to life raises a whole host of practical problems, such as where they will live, but let's not spoil the fun...

Sabre-toothed tiger
(Smilodon fatalis)

Extinct: ~10,000 years ago

DNA preservation: 3/5

Suitable surrogate: 3/5

This fabled beast with its extraordinary canines would be a sight to behold. There are some spectacularly preserved sabre-toothed specimens from the La Brea tar pits in Los Angeles, but the tar makes extracting DNA tricky, so nobody has been able to isolate decent sequences. However, there are also some permafrost-preserved specimens that might be a better source of DNA. If we could obtain a genome, a close living relative of the sabre-tooth, the African lion, should be a good egg donor and surrogate mother. Californians, beware!

Neanderthal
(Homo neanderthalensis)

Extinct: ~25,000 years ago

DNA preservation: 1/5

Suitable surrogate: 5/5

A draft sequence of the Neanderthal genome should be published sometime this year. "To have a reasonable-quality genome, say comparable to the chimpanzee, will then be another two years of work or so," says Svante Pääbo of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. While he and his colleagues hope the genome will offer unique insights into the differences between us humans and our mysterious cousins, there is speculation it could also be used to resurrect the Neanderthal. Because of our very close shared ancestry, humans would make ideal egg donors and surrogate mothers.

However, while Soviet scientists might once have tried to create a human-ape hybrid, today it is hard to imagine even the most crazed of mad scientists entering such taboo territory. "I find the idea of resurrecting the Neanderthal so ridiculous that any speculation on surrogate mothers is superfluous," says Pääbo. At most, researchers might replace some human genes with the Neanderthal versions in cells growing in a dish to see what the effect is, he says.

Short-faced bear
(Arctodus simus)

Extinct: ~11,000 years ago

DNA preservation: 3/5

Suitable surrogate: 2/5

This towering beast would dwarf the world's largest living land carnivore, the polar bear. The short-faced bear may have been a third taller than the polar bear when standing upright, and it weighed up to a tonne. Recovering its DNA should be possible as there are specimens encased in permafrost. The short-faced's closest living relative is the spectacled bear of South America. The two species parted evolutionary company only around 5 million years ago, but unfortunately, at just a tenth the body mass of the short-faced bear, the spectacled bear is unlikely to be a particularly good surrogate.

Tasmanian tiger
(Thylacinus cynocephalus)

Extinct: 1936

DNA preservation: 4/5

Suitable surrogate: 1/5

The last Tasmanian tiger or thylacine - an individual that has become known as Benjamin - died in Hobart Zoo in 1936. The existence of various preserved tissues less than a century old means geneticists should be able to get good-quality DNA and produce a complete sequence of the thylacine genome before too long. When it comes to resurrection, marsupials like the thylacine might be easier than most other mammals. Pregnancy in marsupials typically lasts just weeks, and a simple placenta forms only briefly, meaning there might be less risk of incompatibility between an embryo and a surrogate mother of another species. For the thylacine, the surrogate would be the Tasmanian devil. After birth, the fetus could be raised on milk in an artificial pouch.

Glyptodon
(Doedicurus clavicaudatus)

Extinct: ~11,000 years ago

DNA preservation: 2/5

Suitable surrogate: 1/5

The Volkswagen Beetle-sized "colossal" armadillo, with its spiky, club-like tail, once rumbled across the South American countryside, and some might fancy seeing it do so again. Because there are no frozen glyptodons, obtaining usable DNA will depend on finding well-preserved remains in a cool, dry cave. Beyond that, there is an even bigger problem: the most suitable species to act as a host for a developing glyptodon embryo would be the far smaller 30-kilogram "giant" armadillo. The difference in size means it would struggle to carry its extinct relative to term.

Woolly rhinoceros
(Coelodonta antiquitatis)

Extinct: ~10,000 years ago

DNA preservation: 4/5

Suitable surrogate: 5/5

Resurrecting the woolly rhino has lots going for it. As with the mammoth, there are plenty of specimens preserved in permafrost, and the availability of hair, horns and hooves is a big plus. These tissues can be cleaned up with shampoo and bleach to remove contaminant DNA from microbes and fungi before using enzymes to release an abundance of near-pure rhino DNA. This makes it likely that geneticists will publish the complete genome of this hirsute beast before long. However, although the woolly rhino has close living relatives that might make suitable surrogates, all contemporary rhino species are themselves on the brink of extinction. As long as this remains the case, resurrecting a woolly rhino is unlikely to be a top priority.

Dodo
(Raphus cucullatus)

Extinct: ~AD 1690

DNA preservation: 1/5

Suitable surrogate: 3/5

In 2002, geneticists at the University of Oxford got permission to cut into the world's best-preserved dodo specimen, a foot bone - complete with skin and feathers - held under lock and key at the university's Museum of Natural History. "It was one of the scariest things I've had to do," recalls Beth Shapiro, an ancient DNA specialist now at Pennsylvania State University. This yielded minute fragments of dodo mitochondrial DNA but nothing more. Since then, no other specimen has yielded even a whiff of dodo DNA, but there is still hope that some will one day be found. "We're still looking," says Shapiro. If one turns up and a genome sequence could be produced from it, it would then be down to pigeons to help bring their famous cousin back from the dead.

Giant ground sloth
(Megatherium americanum)

Extinct: ~8000 years ago

DNA preservation: 2/5

Suitable surrogate: 1/5

This giant stood around 6 metres tall and is estimated to have weighed a whopping 4 tonnes. The sloth's relatively recent extinction means that several specimens have been found with hair, an excellent source of DNA. So are we likely to see the giant sloth genome published? "Absolutely," says Hendrik Poinar of MacMaster University in Canada, who has extracted giant ground sloth DNA from fossilised dung deposited some 30,000 years ago. The difficulty for anyone intent on resurrection would be the lack of a suitable surrogate. Its closest living relative - the three-toed tree sloth - is tiny by comparison. It might be able to provide eggs with which to create a giant ground sloth embryo, but the fetus would quickly outgrow its surrogate mother.

Moa
(Dinornis robustus)

DNA preservation: 3/5

Suitable surrogate: 2/5

There is plenty of moa DNA to be found in well-preserved bones and even eggs in caves across New Zealand, so obtaining a moa genome should be doable. But which one? It would be tempting to go for the massive Dinornis robustus, which stood more than 3 metres tall, but starting with the more modestly sized Megalapteryx didinus might make more sense. Although only distantly related to ostriches, it might be possible to boot up the moa genome in an ostrich egg. As no bird has yet been cloned, however, perhaps the most feasible approach would be to engineer an ostrich embryo to be moa-like.

Irish elk
(Megaloceros giganteus)

Extinct: ~7700 years ago

DNA preservation: 3/5

Suitable surrogate: 2/5

Deer-hunting enthusiasts would give almost anything for a chance to stalk this Pleistocene giant, once found across Europe. A typical male Megaloceros stood more than 2 metres tall at the shoulder and sported antlers 4 metres wide. It is actually a deer rather than an elk, and its closest living relative is the much smaller fallow deer, the two species having parted evolutionary company around 10 million years ago. The gulf between the two species means it is hard to see how a complete genome could be converted into a living, breathing animal.

Giant beaver
(Castoroides ohioensis)

Extinct: ~10,000 years ago

DNA preservation: 2/5

Suitable surrogate: 1/5

There is fierce controversy over the reintroduction of normal beavers in some countries, so imagine how much fuss there would be over the reintroduction of the 2.5-metre-long giant beaver to North America. It's not too much to hope for a genome sequence of this massive rodent, says Hendrik Poinar, a geneticist at McMaster University in Hamilton, Canada. The capybara, which is about half the mass, would probably be the most suitable surrogate, though it might still be too distant a relative.

Gorilla
(Gorilla gorilla)

Extinct: Almost

DNA preservation: 5/5

Suitable surrogate: 5/5

The first species to be brought back from extinction will most likely be one that is alive today. Conservationists are freezing tissue samples from some threatened species, so clones could be created with the help of a closely related surrogate species if a suitable habitat becomes available. For gorillas, the surrogate would be the chimpanzee.

Revival recipe:

YOU WILL NEED:

Well-preserved DNA
Several billion DNA building blocks
A suitable surrogate species
Some seriously advanced technology

HERE'S WHAT TO DO...

1. Extract the DNA from your extinct species, sequence the fragments and assemble to obtain a complete genome.

REALITY CHECK: genome sequences from extinct animals are likely to be riddled with lethal errors.

2. Now take your DNA building blocks and recreate the DNA of your extinct beast, in the correct number of chromosomes.

REALITY CHECK: it is not yet possible to make such long DNA molecules from scratch, but we should be able to one day.

3. Package the chromosomes up into an artificial nucleus and pop it in an egg collected from your suitable surrogate species. This should then develop into an embryo, which will be a clone of a long-dead animal.

REALITY CHECK: finding compatible species, let alone extracting eggs from them, could be a huge problem. Plus, no one has yet managed to clone birds or reptiles.

4. Grow a baby animal from the embryo. For mammals, implant the embryo in the womb of a compatible surrogate mother. For a reptile or bird, incubate embryo using yet-to-be-developed techniques. For an amphibian or fish where fertilisation takes place outside the body, just sit back and watch.

REALITY CHECK: compatible surrogate mothers may not exist for many extinct mammals.

HOW TO CHEAT:

Rather than synthesising the entire genome from scratch, you could take the DNA of a closely related living species and modify it to be more like that of the extinct species you are aiming for.

REALITY CHECK: some living species have already been made superficially more like extinct ones, but with today's knowledge and technology they remain far from the real thing.