摘要
Identifying beneficial or functional mutations for a specific species is a task always relies on the labor-intensive construction of a library containing thousands of single-gene knockout or interference strains. Here, we systematically demonstrated that the task could be done by constructing a genome-scale library, designed by AutoGSL, containing a limited number of large fragment deletion strains. The loss-of- and gain-of-function phenotypes could be efficiently screened out by our chromosome-segment scanning, and the specific gene corresponding to an observed phenotype could be quickly identified and visualized by our computer-aided gene-annotation web tool. Additionally, a Clusters of Orthologous Gene Transformer learned representations were transferred to predict growth phenotypes of the genome-scale library under varying conditions. We further utilized our chromosome segment scanning for gain- or loss-of-function screening (CHASING) strategy to obtain acetoin- and lycopene-overproducing hosts. Our work highlighted the significance of CHASING in functional genomics investigation, robust chassis engineering, and chemical overproduction.
源语言 | 英语 |
---|---|
文章编号 | 112484 |
期刊 | iScience |
卷 | 28 |
期 | 5 |
DOI | |
出版状态 | 已出版 - 16 5月 2025 |