SALT LAKE CITY — Despite huge strides in medicine, doctors today are still regularly confounded with what seems to be a simple task: diagnosing infectious diseases.
Whether it's a child who shows up at the hospital with severe pneumonia or a chemotherapy patient battling an unusual infection, doctors still have very few tools at their disposal to quickly and accurately identify the pathogen causing the infection.
Now ultrafast software from developers at the University of Utah can scan the genetic material in a patient’s bodily fluid and search a database of millions of bacteria, virus and fungi for a match — all within minutes.
Taxonomer may be a "game changer" for how doctors diagnose patients, according to co-developer Mark Yandell, a professor of human genetics at the University of Utah and a co-director of the USTAR Center for Genetic Discovery.
How it works
The software builds on already-existing technology that sequences the DNA and RNA in a patient's bodily fluid. Researchers can then upload that data to Taxonomer over the internet. Within a minute, Taxonomer matches the gene sequences to known viruses, bacteria and fungi, displaying an inventory of all pathogens in the sample.
"With Taxonomer, you can look at every known pathogen all at once, without any prior hypothesis as to what's causing the disease," Yandell said.
The software can also detect which genes are being turned on to produce certain antibodies or proteins — a clear signal of which pathogens the immune system is trying to attack.
"Instead of this cat-and-mouse game where physicians say, 'Do they have X?' and you run the test for X, or 'Well, do they have Y?' and et cetera, what you get is everything all at once," Yandell said. "Even stuff you would have never guessed they would have had."
Taxonomer is currently limited by the speed of genome sequencing technology. Right now, the turnaround time for a lab to sequence a genetic sample is about 24 hours, Yandell said. In the next few years, advances in technology could cut that time down to hours.
Combined with the power of Taxonomer, a patient with a respiratory infection in the ER could get a conclusive diagnosis of their infection from a blood draw within five to six hours, according to Yandell.
U. professor of pediatrics Carrie Byington said the tool may be particularly important for treating children, who face more life-threatening infections early in life.
“As a clinician, this gives you a better idea when we identify a pathogen whether it is really the cause of the disease,” Byington said in a statement.
She added: "I believe this is a paradigm shift in how we diagnose people. It is why I wanted to be involved.”
Put to the test
Using real-world sample data collected by the Centers for Disease Control and Prevention and University of Utah hospitals, researchers put Taxonomer to the test in a study published Thursday in the journal Genome Biology.
Researchers found that some patients who appeared to be infected with Ebola were actually infected with more exotic bacteria.
On Thursday, the Bill & Melinda Gates Foundation awarded co-developer Dr. Robert Schlaberg, a medical director at ARUP Laboratories, a $100,000 grant to test Taxonomer in "resource-limited areas" around the world with an eye toward reducing childhood mortality. Almost 5 million children under the age of 5 die each year from infectious diseases worldwide.
"Believe it or not, a lot of Second or Third World countries have sequencing machines," Yandell said. "What they don’t have is the computer expertise to analyze the data that comes off those machines."
All are based on technology that sequences DNA and RNA in bodily fluids and matches them to certain pathogens.
What's new is that those computations usually take months to analyze, even on advanced computers, due to the "hideously complex" nature of sequencing genetic material in human bodily fluid, Yandell said.
So-called “big data” — the kind of computations used to monitor emails flowing across the country or cellphone activity of millions of people — has driven computer science forward to allow the same approaches to be applied to genomic data, according to Yandell.
"This is about as big as big data gets," he said.
Scientists at the University of Utah, USTAR Center for Genetic Discovery, ARUP Laboratories and IDbyDNA Inc. worked for three years on the ultrafast, web-based tool.
Yandell estimated the cost of using Taxonomer at around $2,000 per test. He predicts that will drop quickly as the technology becomes more popular, just as the cost to sequence a genome fell from $100 million in 2001 to about $1,300 today.
He believes these ultrafast computing technologies have far broader implications than whole human genome sequencing because human genetic diseases are relatively rare. Infections, however, are not.
"Ultimately these technologies are at least going to match the impact of human genome sequencing on human health," Yandell said. "At least."