AUG 11, 2020 7:48 AM PDT

Supervillain Alert! New Tech IDs Drug-Resistant Superbugs.

WRITTEN BY: Tara Fernandez

The human body is made up of around 10 trillion cells. Fascinatingly, we have 10 times more bacterial cells on our bodies which make up the normal microflora. We are also exposed to a vast number of microbes with pathogenic potential that can on occasion break past immune defenses and wreak havoc. When infection strikes, patients usually provide diagnostic labs with a biological sample for them to determine the identity of the offending bacteria. However, it’s not easy for microbiologists to spot a “supervillain” lurking in the sample. These are antibiotic-resistant superbugs -- pathogens that infect 2 million people in the U.S. and kill at least 23,000 every year.

A team of biomedical engineers at Duke University has invented a new platform for distinguishing the bad guys from the supervillains, using artificial intelligence. Their technology uses machine learning algorithms to pick up subtle differences in the way antibiotic-resistant bacteria grow under laboratory conditions. This would allow microbiologists to flag superbugs in a fraction of the time it currently takes, with a cost-effective and less labor-intensive procedure. Ultimately, this could revolutionize patient outcomes for those infected with these nasty invaders.

Before this breakthrough, diagnostic labs relied heavily on next-generation sequencing to “read” the genes from pathogenic bacteria in search of signature mutations that are known to shield bacteria from the killing effects of antibiotics. However, these techniques are not universally available and it is not always clear how these gene signatures correlate with symptoms or optimal treatment strategies.

Linchong You, the lead author of the study published recently in the Proceedings of the National Academy of Sciences, looked to bacterial growth dynamics instead of genetics for clues.

“We thought that the slight variance in the genes between strains of bacteria might have a subtle effect on their metabolism,” said You.

“But because bacterial growth is exponential, that subtle effect could be amplified enough for us to take advantage of it. To me, that notion is somewhat intuitive, but I was surprised at how well it actually worked.”

The team validated their AI-powered platform using more than 200 strains of bacterial pathogens, many of which were clinically-relevant variants of E. coli. The growth dynamic profiles of these bacteria alone were able to predict a strain’s resistance to antimicrobial drugs  around 75 percent of the time.

According to You, this result is on par or better than standard techniques, including those that rely on genetic sequencing data, adding “And this was just a proof of principle. We believe that with higher-resolution data of the growth dynamics, we could do an even better job in the long term.”

Future plans involve refining the system further such that results can be obtained in under 12 hours.


Sources: Duke Pratt School of Engineering, PNAS.

About the Author
  • Tara Fernandez has a PhD in Cell Biology and has spent over a decade uncovering the molecular basis of diseases ranging from skin cancer to obesity and diabetes. She currently works on developing and marketing disruptive new technologies in the biotechnology industry. Her areas of interest include innovation in molecular diagnostics, cell therapies, and immunology. She actively participates in various science communication and public engagement initiatives to promote STEM in the community.
You May Also Like
JUL 29, 2021
Genetics & Genomics
Obesity May Not Always Lead to Disease
JUL 29, 2021
Obesity May Not Always Lead to Disease
Some gene variants might be protecting people from the negative health effects of obesity.
SEP 09, 2021
Clinical & Molecular DX
A Drug Test for Synthetic Cannabis Use
SEP 09, 2021
A Drug Test for Synthetic Cannabis Use
Designer drugs are synthetic analogs of prohibited substances such as cocaine and LSD made in clandestine laboratories a ...
SEP 14, 2021
Clinical & Molecular DX
HIV Self-Test App Proves Promising
SEP 14, 2021
HIV Self-Test App Proves Promising
A new app allows users to self-test for HIV, which has proven to help positive patients get access to medical care and c ...
OCT 07, 2021
Drug Discovery & Development
Extended approval for a drug to treat children living with HIV
OCT 07, 2021
Extended approval for a drug to treat children living with HIV
On Monday October 4th, ViiV Healthcare announced that it had submitted an application to the United States Food and Drug ...
OCT 14, 2021
Clinical & Molecular DX
Cancer-Sniffing Worms: The Future of Diagnostics?
OCT 14, 2021
Cancer-Sniffing Worms: The Future of Diagnostics?
Researchers have unlocked a way of sniffing out cancer in patients with very-early-stage pancreatic cancer. This time, i ...
NOV 09, 2021
Clinical & Molecular DX
Device Scavenges Body Heat, Allows Crowds to Gather Safely
NOV 09, 2021
Device Scavenges Body Heat, Allows Crowds to Gather Safely
If efforts to maintain public safety amidst the ongoing pandemic have taught us one thing, it’s that we need bette ...
Loading Comments...