These days, genomics has its fingers in all sorts of pies. It is being applied to the analysis of evidence from a crime scene, to better understanding infectious diseases and public health, and to the study of animal populations. As these tools become cheaper and more accessible, genomics is spreading beyond the confines of basic research.
"I think technology is driving what we can do [in public health]," says Nicole Dowling, an epidemiologist at the Centers for Disease Control and Prevention.
Her colleague, Alison Mawle, says that not all aspects of the genomics area have been tapped. "I think it would be fair to say that they are still getting going in infectious disease research; we've only scratched the surface," she says. "There could be tremendous potential down the road, but we're not there yet. It's not just genomics — we would say that proteomics and metabolomics all have the potential for giving a tremendous amount of information about infectious diseases."
In the shiny world of crime shows, testing for DNA takes only a commercial break before the culprit is spat out of a printer. Of course, it's a bit more complicated than that. But forensics is getting a helping hand from genomics. "Having the human genome done now, or at least a pretty good sequence, is very helpful because we know exactly where the genetic markers that we are using for forensic testing are," says John Butler at the National Institute of Standards and Technology. Forensic scientists focus their energies on 20 short tandem repeats — 13 sites, chosen by the FBI in 1997, are tested in the US and eight of those sites overlap with those tested in Europe. So far, Butler estimates that 15 million samples have been run with those markers around the world — the US alone has 8 million profiles stored in its databases. "Genomics has helped with understanding those markers better, where they are located and so on," Butler adds.
"They provide a common currency to be able to exchange information — there's about 175 forensic DNA labs in the United States," Butler says. Most labs, he adds, use commercially available kits that test those STRs, along with a sex-typing -marker. NIST has developed standards — markers that they have sequenced — for scientists to use in calibrating their measurements.
Prior to running any samples, many researchers first turn to quantitative PCR. Often, Butler says, there is a limited amount of material to work with from a crime scene and qPCR allows forensic scientists to determine how much they have to work with. NIST also has a reference set to calibrate quantification.
In addition, Butler's group at NIST has developed a multiplexed test to use in more complex cases when distinctions must be made among family members, such as in immigration or paternity testing, or to be able to identify victims of a mass disaster through their families. To do that, they developed a 26-plex test with 25 STRs from separate chromosomes and one to determine gender. Those new STRs are located on separate chromosomes and are in different regions from the 15 standard ones. "You can combine them and get 40 STRs," he says. "When you are testing relatives, you need to have more markers." Subsets of those 25, he says, are already in use in some paternity testing labs. He adds that the US Armed Forces test mitochondrial DNA, especially of remains from World War II, the Vietnam War, or the Korean War, as well as in cases in which the remains were burned, because it is there in far greater numbers than nuclear DNA.
SNPs, however, have not been adopted for widespread use in forensics — mainly because there is already so much STR data. "Part of the challenge is once you develop a database of 8 million profiles, switching to a different set of markers isn't trivial anymore. You'd have to go back and re-run them in order to have the ability to compare legacy data to new data," Butler says.
It's not only human DNA that may wind up at a crime scene. Many people own pets, and their fur or blood may also be in the mix. "In the US there's almost a one-to-one correlation between the number of households and dogs," says Sree Kanthaswamy at the University of California, Davis. "When you have dogs or cats, you are going to find a lot of cat hair, dog hair, dander, and all the other stuff, so if someone breaks into your house … you can use whatever piece of evidence from this person to link him to a crime scene or a suspect to a victim."
With funding from the National Institute of Justice, Kanthaswamy and his colleagues developed a canine database to understand canine populations. In it, Kanthaswamy includes the 10 most popular dog breeds, as well as mixes that don't have such a pedigree. Based on STR and micro-satellite testing, he can determine just what dog was there. "We can uniquely identify a dog; you never have a mix-up between this dog and any other dog," he says.
Currently, Kanthaswamy and his colleagues are working on a way to identify the species from which a sample comes, right at the crime scene. That way, crime scene investigators would know which lab to forward the sample to. "Only human labs can do human evidence, and non-human labs can do non-human evidence," he says. "If, onsite, you can identify this piece of hair — sometimes you can't identify things microscopically, you have to do it through DNA" — that will help forward the evidence to the proper lab for testing.
There's always talk of how forensic testing labs have long backlogs — a 2003 study funded by NIJ found that state and local crime labs had more than 57,000 unanalyzed DNA cases waiting around. Faster genomics tools may help there as well. "Certainly, there is a need for higher throughput," Butler says.
UCSD's Kanthaswamy has also applied genomic tools to study the population structure of the macaques at the California National Primate Research Center. There are more than 5,000 rhesus monkeys there, including both rhesus macaques and long-tail macaques, which are closely -related. These monkeys are both imported into the US and bred in colonies. After humans, rhesus monkeys are the most genetically and geographically diverse primates. "We don't have these strict inbred lines, like we do with mice where the outcomes are very predictable when you compare two different mice," Kanthaswamy says. The genetic differences among the monkeys, which are tied to their geographic origins, affect how they respond to disease.
"There are big differences between how Indian and Chinese monkeys respond to AIDS experimentation," he adds. "When people include Chinese and Indian animals in their experiments, or hybrids, then they are adding another level of variation."
Using tools similar to those of forensic scientists' — 14 STRs — Kanthaswamy was able to get an idea of the genetic makeup of this colony of macaques and thus, a better idea of how they would respond to disease and aid the researchers studying the monkeys.
Already, genomics tools have played a role in the response to infectious disease. Soon after the novel H1N1 influenza pandemic emerged last year, researchers were hard at work studying its genome. Within days, the origin of all eight segments of the influenza genome had been determined. This wasn't the first time, however, that such tools have been used to study an ongoing disease outbreak. The CDC's Mawle says that this approach was used during both the SARS outbreak and a recent hanta-virus outbreak. "The difference, I think, with flu was the speed at which we were able to do that," she says. Because of that speed, the CDC was able to quickly devise and distribute RT-PCR -diagnostic tests for the pandemic flu under emergency use authorization from the Food and Drug Administration. "That was a first," Mawle adds. "The point to make here is that the speed of the current technology means that what might have taken days or weeks can be done — I wouldn't say hours — but can be done very quickly."
Penn State's Eddie Holmes adds that genome sequencing has become a standard approach. "Before it was extra," he says. "For H1N109, it became the default and the whole virus was available for the population to analyze ... in incredible detail very, very quickly at a genome-scale."
In the current pandemic, Mawle says, there has been a high number of children hospitalized; also during the 2003 regular influenza season, there was a higher-than-expected number of pediatric deaths. Many of these were children who had no risk factors for adverse outcome. "In terms of genomic research, that would be something that we would like to be able to predict," she says. Those people could be first in line for a vaccine.
Penn State's Holmes focuses on the pathogen side of the spectrum, particularly on the RNA virus dengue fever. Holmes says that RNA viruses are more likely to be emerging diseases since they have a high mutation rate and no error repair processes. They do, though, have tiny genomes — the average size of an RNA virus is about 10,000 nucleotides. "We can produce vast amounts of genomic data," he says. "It's -trivial. … It's a very different way of doing genomics."
And it has opened new windows into the evolution of viruses, he says. Holmes uses those sequences to construct phylogenetic trees to study the virus and track where it is circulating. "With dengue, what we've done is look at patterns of diversity worldwide," he says. "What we can see is in different locations, for example Southeast Asia or North America, there are different strains of dengue and turnover in these strains."
Also, Holmes can determine where those strains arose. "We look at trees and we work out where the most likely place is that the virus comes from," he says. For the seasonal flu, he says, every year it arises in Southeast Asia.
Other infectious disease research looks into why people respond differently to the same infection. In 1998, the public health genomics office at CDC established -HuGENet, an international effort to assess human genome variation in health and disease at the population level. "The goal of the network has been to continue to help to translate genetic research findings into opportunities for preventative medicine and public health, doing so by -advancing -synthesis and interpretation and dissemination of population-based data of human genetic variation," says the CDC's Dowling.
Recently, that same branch of the CDC created HuGE Navigator to house published, population-based, epidemiological studies of human genes. "The way this works is that studies are tracked and curated weekly from listings in PubMed, and it's done by a pretty sophisticated computer algorithm to capture them all," she says. Currently, there are more than 48,000 articles listed in HuGE Navigator, which include genetic variations, genetic diseases, and both gene-gene and gene-environment interactions.
The infectious disease community has caught on to using SNPs and genetic variants in their research, and some are looking to use more high-throughput tools. "With the rapidly advancing technology and the price coming down, there is certainly interest in looking beyond the SNPs and gene expression -profiles, to more sophisticated laboratory [tools]," Dowling says. "I think there is interest in the infectious disease community … to know more about interactions between the host genome and pathogen genomes."