Genome science goes well beyond deciphering the genetic blue prints of humans and a few model organisms. A quick peek through current issues of journals covering evolution, behavior, or ecology shows the explosion of research that includes comparisons of gene sequences. This research field - comparative genomics - recently reached a new phase in its evolution with the building of a dedicated infrastructure and the creation of a new research institute at the American Museum of Natural History (AMNH) in New York City.

AMNH President Ellen Futter announced the establishment of the Institute for Comparative Genomics during a May 2001 press conference preceding the opening of the museum's Genomic Revolution exhibit. AMNH officials then hosted a behind-the-scenes tour of the newly expanded facilities and described the development of their genomics research program covering nonhuman life forms.

Comparative Genomics: Building on a Strong Research Foundation at AMNH

Michael Novacek, provost for research and curator of paleontology, began the tour by noting that AMNH has a 130-year-old tradition of research in evolution and comparative biology. Such studies demand highly organized storage and retrieval systems for both specimens and data. Therefore, the comprehensive new program for gathering and analyzing vast quantities of molecular-genetic data builds naturally upon a well-established and conceptually similar research structure.

As Novacek mentioned, one key part of AMNH's genomics research involves constructing the "Tree of Life," the map of evolutionary relationships among the 1.5 million known species, as well as those species yet to be found. "This map will be analogous to the periodic table of elements; it will teach us how life is processed and how evolution works," Novacek said.

Collecting the DNA-Sequence Data

Currently, comparative genomics involves analyzing a limited number of genes per specimen rather than full-genome sequencing. A typical strategy used for many ongoing projects includes sequencing portions of five to ten different genes per organism and then comparing the DNA sequences of each gene among the individuals, explained Mark Siddall, assistant curator, Division of Invertebrate Zoology. The DNA selected for comparison includes the gene for ribosomal RNA and several genes coding for specific proteins. For studies of eukaryotes, this analysis looks at both nuclear and mitochondrial genes.

Siddall outlined this strategy while standing between two of the many workhorse machines that fill the 10,000-square-foot space devoted to molecular biology labs at AMNH. To one side of him sat a thermocycler, a device that amplifies selected DNA fragments within the total genomic DNA, 384 samples at a time. On his other side sat a robot that cleans the amplified DNA in preparation for sequence analysis, 96 samples at a time in just 15 minutes. According to Siddall, the laboratory is equipped to generate sequence data on about one megabase (one million nucleotide bases of DNA) per day.

Given the extensive DNA preparation and sequencing capabilities of AMNH, it comes as no surprise that AMNH has also built an impressive new storage facility to keep all the samples and has greatly expanded the computer facilities to handle the immense quantities of data generated by the new research projects. Indeed, part of the new computer facility is devoted to keeping records of all the research samples.

Very Cold Storage

AA man called Frost, naturally, oversaw the construction of AMNH's vast new frozen sample facility. Frost is Darrel Frost, associate dean of science for collections and associate curator, Division of Vertebrate Zoology. In the tour of the facility, which is managed by the molecular systematics laboratory's co-director, Rob DeSalle, Frost gave one of several examples of AMNH's penchant for combining low and high technology to solve problems. As he explained, most research institutions store frozen tissue specimens in ultracold (-80 °C) freezers. The freezers run on electricity, which makes them vulnerable to power outages.

AMNH researchers didn't want to risk power outages that could decimate their specimen collections. Therefore, rather than buying lots of new freezers, they opted to use liquid-nitrogen storage dewars - essentially giant thermos bottles - for storing the specimens. According to Frost, liquid nitrogen chills the samples to -150 °C and, therefore, gives better protection than ultracold freezers against sample breakdown.

These giant thermos bottles, which hold up to 70,000 small vials per container, currently give the museum a capacity to store about 400,000 samples. Frost expects that within a few years they will actually have over a million frozen samples, many collected by AMNH staff members and others coming from outside individuals.

As Frost explained, the fully automated system includes bar codes for each sample vial and computer-driven sample racks. Researchers use notebook computers with bar-code readers on a wireless network to retrieve and deposit samples.

The Homemade Supercomputer

The museum's most impressive demonstration of sophisticated technology, built from simple, inexpensive parts, is the new supercomputer. Ward Wheeler, codirector of the molecular laboratories and curator of the Division of Invertebrate Zoology, who directs the computer facility, led us through the small, noisy room that houses the cluster supercomputer, actually a set of several hundred small computers linked in parallel.

Speaking above the hum of all the fans needed to cool the room, Wheeler explained, "To save money, we built the computer ourselves from parts: 2000 DIMS of RAM, a few hundred Pentiums, a few hundred Ethernet cards. We got all the boxes and, up in my office, we assembled each unit. The most expensive thing we got was the box of RAM that was worth over $200,000. All sorts of people - grad students, postdocs, whoever - would come by and build the units. It takes just 15 to 20 minutes, much easier than building a kid's toy."

Wheeler noted proudly that "this computer is one of the ten fastest in the world when it's running at peak performance. Even by traditional supercomputer criteria, it's one of the few hundred fastest in the world!"

This supercomputer enables the researchers who compare DNA sequences to analyze the exponentially increasing amount of data that is now being generated. Speaking about the data, Wheeler commented, "Going from trying to compare 100 to 101 sequences is not one percent more complicated; it goes up by a factor of more than 200." Scientists describe this situation as being "combinatorially explosive," he said.

Wheeler and his team use software that he wrote to conduct phylogenetic analyses. "We assemble DNA data, morphologic data, and fossil taxa data, because 99 percent of the world's diversity is extinct, and so we must include the fossils to get some measure of rates of evolution."

As Darrel Frost explained, the researchers have developed a number of logical tools for using DNA-sequence comparisons to help construct evolutionary trees. Such tools draw on strategies used for morphological studies. "For example," said Frost, "why do we think that goats are more closely related to cows than either one of them is to foxes? They both have cloven hooves. Existing data say that the evolution of cloven hooves happened once in mammals, so animals with cloven hooves must be related to each other. Similarly, the evolution of jaws required complex changes and seems to have happened just once, so jaw-containing animals have a common ancestor. With sequencing, you can use the same logic to learn about evolutionary relationships."

The museum's scientists are incorporating genomics into many of their research projects. In fact, as Frost commented, "Newer questions can be more sophisticated because of the sheer amount of [DNA-sequence] data we can gather."

One such example is seen in the research on leeches done by Mark Siddall and his collaborators. Siddall is fascinated by the blood-feeding aspect of leech behavior, and he wants to learn about the evolution of anticoagulants that these animals make. Siddall said, "All leeches have proteins in their salivary glands that prevent blood from clotting so they don't turn into little bricks. Do different leeches do this in different ways? Are there different biomedically interesting compounds in leeches? Is there something we can learn from figuring out the family tree of these leeches that will give new hypotheses and lead us in new directions?"

With the new facilities for comparative genomic research, Siddall and his colleagues can begin tackling such questions with much greater ease than ever before. And with the advent of the new Institute for Comparative Genomics, AMNH will be able to prepare future generations of students for learning about the tree of life.