# BBERT cuts computing time and memory use in bacterial genome tests

_Published Thursday, October 29, 2026 at 11:06 PM EDT · Science, AI · Latest · Tier 2 — Notable_

![BBERT cuts computing time and memory use in bacterial genome tests — Primary](https://scx2.b-cdn.net/gfx/news/hires/2026/new-ai-tool-helps-scie-3.jpg)

Researchers at the Hebrew University of Jerusalem and partner institutions developed BBERT, an AI tool that reduced computing demands when reconstructing bacterial genomes in tests published in Nucleic Acids Research. It recognizes patterns in DNA fragments as short as 100 letters without requiring a matching reference sequence.

Filtering simulated DNA mixtures before reconstruction reduced average processing time from about 66 hours to 25 hours per data set, retaining approximately 96% of average bacterial-genome coverage. In wild legume root samples, filtering halved reconstruction time and reduced peak memory use from 188 to 82 gigabytes. Some bacterial information was lost, and the tool can confuse chloroplast or archaeal DNA with bacterial DNA.

## Sources

- [Phys.org](https://phys.org/news/2026-10-ai-tool-scientists-bacteria-hidden.html)

---
Canonical: https://techandbusiness.org/newswire/__OjFvRlbQSjbRinjKqPVw
Published: 2026-10-30T03:06:54.538Z
Story chronology: 2026-10-30T02:00:02.000Z
Retrieved: 2026-10-30T06:02:02.886Z
Publisher: Tech & Business (techandbusiness.org)