Published Paper
Download Model
GitHub Repository (legacy)
The links marked legacy point to the original published source code. For the model as this web server runs it, use the Download Model link above.
Python Package
DeepPrime is a deep learning-based model predicting prime editing efficiencies for all possible pegRNAs of a given target sequence, published in Yu et al. Cell (2023).
The DeepPrime web server can accommodate most applications using default parameters, and you can use the most appropriate prime editing (PE) model for your experimental conditions. It can evaluate all possible prime editing guide RNAs (pegRNAs) for a given target according to the predicted prime editing efficiency by the DeepPrime score.
For the best results, use a 60 bp flank on each side of the edit, giving a 121 bp input sequence. That is the shortest input in which every candidate pegRNA's 74 nt scoring window fits, so it offers the fullest set of designs. A shorter pair is designed from as well, and offers fewer candidates: a candidate whose scoring window runs past either end of the sequence is skipped. The edit itself may sit anywhere in the pair.
Input Unedited Sequence(s) (121 bp recommended):
CTCACGTGAGCTCTTTGAGCTTGCCTGTCTCTGTGGGCTGAAGGCTGTTCCCTGTTTCCTTCAGCTCTACGTCTCCTCCGAGAGCCGCTTCAACACCCTGGCCGAGTTGGTTCATCATCAT
Input Edited Sequence(s) (121 bp recommended):
CTCACGTGAGCTCTTTGAGCTTGCCTGTCTCTGTGGGCTGAAGGCTGTTCCCTGTTTCCTACAGCTCTACGTCTCCTCCGAGAGCCGCTTCAACACCCTGGCCGAGTTGGTTCATCATCAT
CTCACGTGAGCTCTTTGAGCTTGCCTGTCTCTGTGGGCTGAAGGCTGTTCCCTGTTTCCTCTTTCAGCTCTACGTCTCCTCCGAGAGCCGCTTCAACACCCTGGCCGAGTTGGTTCATCAT
CTCACGTGAGCTCTTTGAGCTTGCCTGTCTCTGTGGGCTGAAGGCTGTTCCCTGTTTCCTAGCTCTACGTCTCCTCCGAGAGCCGCTTCAACACCCTGGCCGAGTTGGTTCATCATCATTC
Inputs that specify an edit. An unedited/edited sequence pair, ClinVar ID and COSMIC ID name the edit to install, and the pegRNA is designed to write it. No separate homology-directed repair (HDR) donor is offered on this workflow: in prime editing the pegRNA's own reverse-transcription template is the repair template, so a second single-stranded donor would be a different experiment rather than an extra column.
© Laboratory of Genome Editing (Yonsei University)
Note: SynDesign is also available with FASTA or gene input (1 bp substitution), enabling guide RNA design for saturation genome editing.
