Automated high-throughput genome editing platform with an AI learning in situ prediction model

Area of Science:

• Genetics
• Bioengineering
• Computational Biology

Background:

• Generating cell disease models with pathogenic single nucleotide variants (SNVs) is crucial for developing genome editing therapeutics and basic research.
• Traditional methods for creating these models are manual, time-consuming, costly, and prone to errors.

Purpose of the Study:

• To develop an automated high-throughput platform for efficient generation of genome-edited cells.
• To create a predictive model for cytosine base editor (CBE) performance using genome editing data.

Main Methods:

• An automated high-throughput platform was designed to perform large-scale in situ genome editing.
• A Chromatin Accessibility Enabled Learning Model (CAELM) was developed, integrating chromatin accessibility and sequence context data.
• The model was trained on data generated by the automated platform to predict CBE editing outcomes.

Main Results:

• The automated platform successfully generated thousands of edited cell samples with high efficiency within a week.
• The CAELM accurately predicted the performance of in situ base editing.
• The model effectively utilized both chromatin accessibility and sequence context for prediction.

Conclusions:

• The automated platform significantly enhances the efficiency and scalability of generating genome-edited cell models.
• The CAELM provides an accurate method for predicting base editor performance, reducing experimental trial and error.
• This integrated approach is expected to accelerate the development and application of base editor-based genetic therapies.