Back
Science

Sean Brady's computational platform advances antibiotic discovery from soil DNA

View source

Key Findings

  • Brady's lab extracts DNA directly from soil samples, bypassing the need to culture bacteria.
  • Recent advances enable sequencing of large DNA fragments, allowing reconstruction of hundreds of complete genomes from a single sample.
  • Computational algorithms predict chemical structures of potential drug molecules from gene sequences.
  • Predicted molecules are synthesized and tested in the lab.

"This pipeline is moving toward full automation: from soil input to candidate antibiotic output."

Discoveries

Malacidins: a class of antibiotics active against MRSA and other drug-resistant pathogens.

Cilagicin: targets two molecules needed for bacterial cell wall synthesis, making resistance difficult.

Two additional candidate drugs have been identified in the latest study; their mechanisms have yet to be fully characterized.

AI and Automation

  • AI models improve prediction of useful molecules from gene clusters.
  • Larger datasets from newly sequenced genomes enhance algorithm accuracy.
  • The pipeline is moving toward full automation: from soil input to candidate antibiotic output.

Potential Applications

The platform can be applied to any natural environment:

  • Rainforests
  • Deserts
  • Ocean sediments
  • The human gut

Expected to yield not only antibiotics but also antifungals, anti-cancer drugs, and antivirals.