Large dataset enables prediction of repair after CRISPR-Cas9 editing in primary T cells

Area of Science:

• Molecular Biology
• Genetics
• Bioinformatics

Background:

• Understanding DNA repair mechanisms after CRISPR-Cas9 gene editing is crucial for therapeutic development.
• Limited data exists on Cas9-induced DNA damage repair in primary human cells.

Purpose of the Study:

• To characterize and predict DNA repair outcomes following Cas9 cleavage in primary human T cells.
• To develop a machine learning model for predicting insertion and deletion (indel) sequences at targeted genomic sites.

Main Methods:

• Sequencing of on-target repair outcomes at 1,656 genomic sites in primary human T cells.
• Training a machine learning model, SPROUT, using the generated repair outcome data.

Main Results:

• SPROUT accurately predicts the length and probability of nucleotide insertions and deletions.
• The model also predicts the specific sequences of these insertions and deletions.

Conclusions:

• SPROUT facilitates the design of effective SpCas9 guide RNAs for therapeutic applications in primary human cells.
• This predictive model enhances the precision and safety of CRISPR-Cas9-based gene therapies.