Pythia

Pythia optimizes CRISPR integration, tagging, and editing strategies by leveraging deep-learning models and triplet base-pair repeat repair arms (trimologies).

Key Features:

• Deep-learning prediction of repair outcomes: Predicts repair at the genome-cargo interface based on sequence context specific rules.
• Trimology design: Generates triplet base-pair repeat repair arms that correspond to microhomologies at double-strand breaks.
• Large cargo integration: Facilitates integration of large cargo (>2 kb) while protecting the targeted locus and transgene from excessive damage.
• Small edit optimization: Predicts optimal repair arms for single- or double nucleotide edits using oligonucleotide templates.

Scientific Applications:

• Genomic integration in human cells: Successful integrations occurred in >30 loci in human cells.
• In vivo integration models: Germline transmissible transgene integration in Xenopus and endogenous tagging of tubulin in adult mice brains.
• Non-dividing cell editing: Enables integration during early embryonic cleavage and in non-dividing differentiated cells.
• Therapeutic and experimental genome engineering: Facilitates genomic integration and editing for experimental and therapeutic purposes across diverse target cell types.

Methodology:

Pythia uses deep-learning models trained on repair outcomes at the genome-cargo interface to predict sequence context specific repair patterns and design optimal trimologies.