Software speeds up search for new antibiotics

Antibiotic resistance is increasing worldwide. At the same time, it is taking longer to find new antibiotics. ‘The lowhanging fruit has already been plucked,’ says Bioinformatics researcher Marnix Medema. Most antibiotics are based on substances that bacteria and fungi produce naturally to protect themselves, mainly from other microorganisms.

The software shows at a glance which bacteria and genes have the potential to produce as yet undiscovered molecules.

Marnix Medema, researcher at Bioinformatics

While many antibiotics have already been discovered, they only make up a small proportion of all the potential substances. For the DNA of bacteria shows that they are capable of producing many substances that are as yet unknown. But finding the most valuable substances is like looking for a needle in a haystack: until recently, researchers could only discover new antibiotics by testing bacteria one by one in the lab.

Gene clustering

Now Medema and his Mexican and American colleagues have developed software that goes through that haystack at lightning speed. They presented the software in the journal of Nature Chemical Biology. The software, called BiG-SCAPE and CORASON, automates the process of genome mining. First the DNA is determined for various bacteria and entered into the computer. Then the software searches for gene clusters that are known to be involved in the production of antibiotics. It compares the clusters with one another and groups them so that they can be searched effectively to find new genes with the potential to make antibiotics. Medema: ‘The software shows at a glance which bacteria and genes have the potential to produce as yet undiscovered molecules.’

Super quick analysis

‘In the past, such computer analyses could only be performed on the genes of one individual bacterium,’ says Medema. ‘Our programme can analyse thousands of bacteria in one go.’ That was quite a challenge because it requires an awful lot of processing power to compare those millions of genes. That would normally take a computer months but smart programming has brought it down to less than one day.

Other applications

Medema says the software has other applications too, such as searching for new anti-cancer drugs, nutritional supplements or crop protection agents. The software is available free of charge.

Also read:

• Humans are main source of antibiotic resistance

  Science – 22 February 2018 11:20-The idea that people mostly get infected by antibiotic-resistant bacteria by eating meat is mistaken. A Wageningen study shows that most infections…

• ‘There are still new antibiotics to discover’

  Science – 21 February 2018 12:00-American Researchers have discovered a new class of antibiotics encoded in the DNA of the soil microbiome. These malacidines could potentially be used…

• Taming the antibiotics monster

  Science – 14 February 2019 12:28-Ten years ago, the Dutch livestock sector was a big contributor to the emergence of antibiotic-resistant bacteria, as a big consumer of antibiotics….