BioCentury
ARTICLE | Preclinical News

Machine learning tool predicts template-free CRISPR editing

November 8, 2018 4:22 AM UTC

CRISPR veteran David Liu of the Broad Institute of MIT and Harvard and colleagues have developed a machine learning algorithm that predicts the genotypic result of CRISPR-mediated DNA breaks, suggesting CRISPR-based editing could be achieved using the cell's own repair mechanisms without delivering a DNA template along with the guide RNA.

Cas9-induced homology-directed repair (HDR) using a template can lead to precise edits but is limited by low efficiency, according to author Richard Sherwood, assistant professor of medicine at Brigham and Women's Hospital. Moreover, HDR does not occur in non-dividing cells...