ARTFEED — Contemporary Art Intelligence

AlphaEvolve reduces DNA sequencing errors by 30% for PacBio

ai-technology · 2026-05-07

AlphaEvolve, a coding agent powered by Gemini from Google DeepMind, has enhanced DeepConsensus—a model developed by Google Research to rectify DNA sequencing inaccuracies—resulting in a 30% decrease in variant detection mistakes. This advancement allows PacBio scientists to analyze genetic information more precisely and cost-effectively. According to Aaron Wenger, Senior Director at PacBio, this solution significantly boosts the accuracy of their sequencing devices, which may lead to the identification of disease-causing mutations that were previously undetectable.

Key facts

  • AlphaEvolve is a Gemini-powered coding agent by Google DeepMind.
  • It improved DeepConsensus, a model for correcting DNA sequencing errors.
  • The improvement achieved a 30% reduction in variant detection errors.
  • The work helps scientists at PacBio analyze genetic data more accurately.
  • The solution reduces costs for genetic analysis.
  • Aaron Wenger is Senior Director at PacBio.
  • Higher accuracy may enable discovery of hidden disease-causing mutations.
  • DeepConsensus was developed by Google Research.

Entities

Institutions

  • Google DeepMind
  • Google Research
  • PacBio

Sources