PEGG: Prime Editing Guide Generator¶
This is PEGG 3.1. For the original PEGG 2.0 documentation, click here¶
Click here to read the Nature Biotechnology article¶
Warning
Version 3.1 fixes a silent bug affecting silent bystanders on indels.
Who is affected: users of version 3.0 or 3.0.1 who designed pegRNAs with silent_bystander=True for
insertions or deletions (INS, DEL, INDEL). Substitutions (SNP, DNP, ONP) were never affected, and
neither were in-frame indels whose length change is a multiple of 3.
What went wrong: the reading-frame anchor was walked back by the length of the RTT rather than the number of
reference bases it spans. On an indel read on the opposite strand these differ, shifting the frame by
(ref_len - alt_len) % 3. Bystanders that were supposed to be synonymous could be non-synonymous, and
no error was raised – the shifted frame was self-consistent, so the internal translation check passed.
What to do: if you generated a library under those conditions, regenerate the affected designs with version
3.1. Ordinary pegRNA design (silent_bystander=False) was never affected.
PEGG is a python package that designs prime editing guide RNAs (pegRNAs) and base editing guide RNAs (gRNAs) for use in precision genome editing. Unlike the existing, web-based programs for pegRNA design, PEGG is suitable for designing thousands of pegRNAs at once, giving users the ability to design entire libraries of pegRNAs and gRNAs. Uniquely, PEGG can design paired pegRNA or gRNA-sensor cassettes that include a synthetic version of the target locus, allowing for the calibration of guide editing activity in pooled screens (see above bioRxiv preprint for more information).
PEGG’s main functions are:
Generating pegRNAs or gRNAs based on a list of input mutations. The input format is extremely flexible, allowing for users to input a list of genome coordinates or sequences with desired edits.
Ranking and filtering these pegRNAs based on their properties, including On-Target (Azimuth) Scores.
Automated oligo generation (with the option to include a synthetic “sensor” region).
Automated pegRNA/gRNA library design with included safe-targeting mutations, non-targeting guides, and silent substitution controls.
Visualization Tools for pegRNA and gRNA design.
Optional silent bystander mutations: synonymous edits placed alongside the intended edit to evade mismatch repair and, where they fall in the PAM, reduce re-nicking. Both the ordinary and bystander-carrying designs are returned, so they can be compared directly.
PEGG was updated to version 2.0 with new features including (1) increased input mutation format flexibility, (2) Dynamically computed Azimuth on-target scores, (3) a new base editing module, (4) improved library design functionality, as well as some minor bug fixes with INS/DEL design.
Version 3.0 adds silent bystander mutations. pegRNAs can carry synonymous edits alongside the intended one, and both
designs are returned in the same table – the ordinary pegRNA and up to silent_per_mut bystander-carrying variants
of it – distinguished by the has_silent_bystander column, so either set can be filtered out afterwards. Sensors,
oligos, and polyT / restriction-site filtration all reflect the bystander mutations, and all variant types are
supported (SNP, ONP, INS, DEL, INDEL). The feature is off by default: with silent_bystander=False the output is
identical to version 2.0.
The synonymous mutation logic is informed by the silent bystander design approach in PRIDICT2.0 (Mathis et al., Nature Protocols 2025).
Version 3.1 adds optional pegRNA scoring with OptiPrime
(Hsu et al., Nature Biotechnology 2026), a model trained on large-scale measured prime editing outcomes that
explicitly accounts for mismatch repair. We recommend it for library design: PEGG’s built-in PEGG2_Score
is hand-weighted rather than fitted to data. OptiPrime installs separately and the only cost is runtime; see
🎯 OptiPrime Scoring. Version 3.1 also fixes the indel reading-frame bug described above.
Installation¶
PEGG is available through the python package index. To install, use pip:
pip install pegg
Note¶
PEGG has been tested with python versions 3.9 and 3.10. Python versions higher than 3.10 are not compatible with the scikit-learn package version needed to compute protospacer on-target scores. To get it to install, you may need to use a virtual environment :
conda create -n myenv python=3.9
Additionally, some users (particularly on Windows) have reported installation issues stemming from the cyvcf2 package used for translating ClinVar IDs to a format compatiable with PEGG. A version without this package and its functionality is available for local pip installation in the following dropbox link (pegg-2.0.92-py3-none-any.whl) .
Contents:
PEGG is an open source python package. If you use PEGG, please cite it using the following citation:
Gould, S.I., Wuest, A.N., Dong, K. et al. High-throughput evaluation of genetic variants with prime editing sensor libraries. Nat Biotechnol (2024). https://doi.org/10.1038/s41587-024-02172-9