Random adversarial threshold search enables automated DNA screening

Abstract Custom DNA synthesis underpins modern biology, but hazardous genes in the wrong hands could threaten many lives and public trust in science. In 1992, a virology-trained mass murderer tried and failed to obtain physical samples of Ebola; today, viruses can be assembled from synthetic DNA fragments. Screening orders for hazards is unreliable and expensive because similarity search algorithms yield false alarms requiring expert human review. Here we develop “random adversarial threshold” (RAT) search, which looks for exact matches to short nucleic acid and peptide subsequence windows from hazards and predicted functional variants that aren’t found in any known innocuous genes. To experimentally assess sensitivity, we used RAT search to protect nine windows from the M13 bacteriophage virus, then invited a “red team” to launch up to 21,000 attacks at each window and measure the fitness of their designed mutants. We identified defensible windows from regulated pathogens, built a curated test database that our M13 experiments indicate will block 99.999% of functional attacks, and verified its sensitivity against orders designed to evade detection. RAT search offers a way to safeguard biotechnology by securely automating DNA synthesis screening. Summary Searching for exact matches to pre-computed functional variants unique to hazardous genes enables sensitive, secure, and automated DNA synthesis screening.