Software predicts bacteria's response to new drugs

Researchers have developed a novel computer software which identifies genetic changes that allow bacteria to develop resistance to new experimental drugs.
Duke University researchers used the software to predict a constantly-evolving infectious bacterium's countermoves to one of these new drugs ahead of time, before the drug is even tested on patients.
The team used their programme to identify the genetic changes that will allow methicillin-resistant Staphylococcus aureus, or MRSA, to develop resistance to a class of new experimental drugs that show promise against the deadly bug.
When the researchers treated live bacteria with the new drug, two of the genetic changes actually arose, just as their algorithm predicted.

"This gives us a window into the future to see what bacteria will do to evade drugs that we design before a drug is deployed," said co-author Bruce Donald, a professor of computer science and biochemistry at Duke.
Developing pre-emptive strategies while the drugs are still in the design phase will give scientists a head start on the next line of compounds that will be effective despite the germ's resistance mutations.

"If we can somehow predict how bacteria might respond to a particular drug ahead of time, we can change the drug, or plan for the next one, or rule out therapies that are unlikely to remain effective for long," said Duke graduate student Pablo Gainza-Cirauqui, who co-authored the research paper.

Until now, scientists trying to predict the genetic changes that would enable a bacterium to evade a particular drug have had to look up possible mutations from "libraries" of resistance mutations that have been observed previously.
But this approach falls short when it comes to anticipating how bacteria will adapt to new drugs.
To overcome this problem, a research team led by Donald at Duke and Amy Anderson at the University of Connecticut used a protein design algorithm they developed, called OSPREY, to identify DNA sequence changes in the bacteria that would enable the resulting protein to block the drug from binding, while still performing its normal work within the cell.

The team focused on a new class of experimental drugs that work by binding and inhibiting a bacterial enzyme called dihydrofolate reductase (DHFR), which plays an essential role in building DNA and other processes.
The drugs, called propargyl-linked antifolates, show promise as a treatment for MRSA infections but have yet to be tested in humans.
From a ranked list of possible mutations, the researchers zeroed in on four tiny differences, known as single nucleotide polymorphisms, or SNPs, that would theoretically confer resistance.
Though none of the mutations they identified had been reported previously, experiments with live bacteria in the lab showed their predictions were right.

The study appears in the journal PNAS.