Decoupling deaminase recruitment and transgene targeting resolves tradeoffs in multiplexed base editing workflows

Designing base editors for therapeutic applications requires more than high editing efficiency. It demands precision, the ability to correct a target nucleotide without introducing unintended bystander edits at adjacent positions. Achieving this at scale, across thousands of pathogenic variant sequences, has historically required extensive empirical screening.

To address this challenge, Revvity developed a scalable discovery workflow using the Pin-point™ modular platform. By combining high-throughput screening with machine learning, the approach enables rapid identification and prediction of precision editors across thousands of therapeutically relevant genomic targets.

The workflow integrates four complementary capabilities:

• Arrayed screening: Combinatorial assembly of Cas and deaminase modules to characterize editing profiles across diverse configurations (CBE and ABE)
• Pooled screening: High-throughput evaluation of ~7,000 pathogenic variant target sequences from ClinVar using lentiviral sensor libraries
• Predictive modeling: Machine learning models (gradient-boosted tree algorithm) trained on pooled screening data to predict editing outcomes, outperforming published models (ρ = 0.66–0.7)
• Cas module comparison: Distinct editing profiles of SpCas9 vs. OpenCRISPR1-based modular editors, highlighting how Cas-deaminase interactions shape precision and efficiency

Key findings:

• Therapeutic candidates identified by pooled screening of ~7,000 pathogenic variant sequences across 1,680 genes
• ML models predict editing outcomes with performance within the range of comparable published models
• Tuning the Cas:deaminase ratio can further optimize the balance between editing precision and efficiency
• OpenCRISPR1 (AI-generated Cas9 ortholog by Profluent Bio) shows distinct editing profiles vs. SpCas9, expanding the design space for precision editors

Download the poster to explore the full screening workflow, predictive model performance, and therapeutic target validation data.

The Pin-point™ base editing platform technology is available for clinical or diagnostic study and commercialization under a commercial license from Revvity.