New CMU algorithm can help thwart antibiotic resistance
Antibiotic resistance represents a modern medical challenge as big-gun drugs become less effective, and even ineffective, against deadly bacterial infections.
In response, Carnegie Mellon University, the University of California San Diego and St. Petersburg State University in Russia combined forces to develop a biological search engine — an algorithm named VarQuest — to expand the arsenal of nonresistant antibiotics with new discoveries.
The algorithm, described in the January online edition of Nature Microbiology, combed through and organized large scientific databases to identify 10 times more variants of “peptidic natural products” — antimicrobial and even anticancer compounds known as PNPs — than previously had been discovered.
Microbes throughout nature produce such antibiotic compounds to kill competing organisms. PNPs, in particular, have been used to create the most powerful antibiotics in medicine, including such drugs as vancomycin and daptomycin. But their effectiveness has diminished due to resistance, which involves bacteria changing or mutating to survive exposure to such drugs.
Hose in Mohimani, formerly of UCSD and now assistant professor of computational biology at CMU’s School of Computer Science, said VarQuest’s mass discovery of antibiotic variants provides many potential candidates for drug development. Pavel A. Pevzner of UCSD’s Department of Computer Science and Engineeringco-led the study.
As it turns out, more than 200 researchers over the years have produced “a gigantic amount of data — a billion possible data points — that could serve as molecular fingerprints of potentially new antibiotic medications,” said Mr. Mohimani, who holds a Ph.D. in computational biology.
But those variants remained well hidden in the data.
With VarQuest, the team was able to pre-process databases to enable the Google-like search that discovered 1,000 variants of the 100 known PNP compounds. Such variants can involve differences of only a few extra atoms of hydrogen and carbon — enough to thwart bacterial resistance, he said.
The major natural antibiotic compounds were discovered long ago and represented “the low-hanging fruit.” Finding new ones has become more costly and time-consuming.
“Five or six years ago,” Mr. Mohimani said, “researchers in the community of antibiotic discovery realized they have to coordinate their efforts to make sure that they are not doing redundant work, and at the same time, sharing valuable data collected from different parts of the earth.”
As a result, the new social network of scientists shares databases that include “fingerprints of molecules” of variants that the new algorithm found — a process that would have taken hundreds of years of computation to discover under previous methods.
Consider the entire Library of Congress disorganized with books randomly stacked and no way to find any specific one. In that context, VarQuest essentially organized and indexed the entire library in a matter of hours.
In a word of caution, Mr. Mohimani said, turning these variants into safe, effective drugs still can take a decade or longer, given the high costs of drug development and testing, and U.S. Food and Drug Administrationrequirements.
Still, VarQuest “is going to play the same role in the field of natural products that a Google search would play for users.”
“This is something we have been investigating for a decade,” he said, comparing it with finding needles in haystacks. “What you need is to organize the haystack in such a nice way that by looking at the big picture, you can identify the needle in it.”
Anna Edlund, an assistant professor in the genomic medicine group at the J. Craig Venter Institute in La Jolla, Calif., was not involved with the study but said it represents “a singlepush-button approach” to discovering an order of magnitudeof new PNPs.
“This study demonstrates that the novel VarQuest tool can be applied to bring new light onto the hidden universe of microbially produced PNPs,” she said. “The tool provides us with novel templates for the discovery of therapeutics such as antibiotics.”