NatureApr 22, 2025

Customizing CRISPR-Cas9 DNA targeting using engineering and machine learning

CRISPR Gene Editing Weekly Brief ↗PubMed ↗DOI ↗

Updated Feb 3, 2026

Abstract

Nearly 1,000 engineered SpCas9 enzymes were characterized for their PAM requirements using machine learning.

A machine learning algorithm, PAMmla, was developed to predict PAM specificity based on the amino acid sequence of SpCas9 enzymes.
The engineered SpCas9 variants identified through this method showed improved performance as nucleases and base editors in human cells.
These bespoke enzymes may reduce the risk of off-target editing compared to traditional generalist CRISPR-Cas systems.
An in silico-directed evolution approach allows for targeted design of Cas9 enzymes for specific applications, including allele-selective targeting.
The study suggests a shift towards personalized Cas9 variants, enhancing safety and efficiency in genome editing.

Simplified

Engineering and characterizing proteins can be time-consuming and cumbersome, motivating the development of generalist CRISPR-Cas enzymesto enable diverse genome-editing applications. However, such enzymes have caveats such as an increased risk of off-target editing. Here, to enable scalable reprogramming of Cas9 enzymes, we combined high-throughput protein engineering with machine learning to derive bespoke editors that are more uniquely suited to specific targets. Through structure-function-informed saturation mutagenesis and bacterial selections, we obtained nearly 1,000 engineered SpCas9 enzymes and characterized their protospacer-adjacent motif (PAM)requirements to train a neural network that relates amino acid sequence to PAM specificity. By utilizing the resulting PAM machine learning algorithm (PAMmla) to predict the PAMs of 64 million SpCas9 enzymes, we identified efficacious and specific enzymes that outperform evolution-based and engineered SpCas9 enzymes as nucleases and base editors in human cells while reducing off-targets. An in silico-directed evolution method enables user-directed Cas9 enzyme design, including for allele-selective targeting of the RHOallele in human cells and mice. Together, PAMmla integrates machine learning and protein engineering to curate a catalogue of SpCas9 enzymes with distinct PAM requirements, motivating a shift away from generalist enzymes towards safe and efficient bespoke Cas9 variants. 1-4 3,5,6 7P23H

Full Text

Full text is available at the source.

View Full Text ↗

Funding

Competing interests

Competing interests: R.A.S. and B.P.K. are inventors on a patent application filed by Mass General Brigham (MGB) that describes the development of PAMmla. B.P.K. and R.T.W. are inventors on additional patents or patent applications filed by MGB that describe genome engineering technologies related to the current study. S.Q.T. is an inventor on a patent application for GUIDE-seq, and is a member of the scientific advisory boards of Ensoma and Prime Medicine. L.P. has financial interests in Edilytics and SeQure Dx. Q.L. is a consultant for Entrada Therapeutics. B.P.K. is a consultant for EcoR1 capital, Novartis Venture Fund and Jumble Therapeutics, and is on the scientific advisory boards of Acrigen Biosciences, Life Edit Therapeutics and Prime Medicine. B.P.K. has a financial interest in Prime Medicine, Inc., a company developing therapeutic CRISPR–Cas technologies for gene editing. The interests of L.P. and B.P.K. were reviewed and are managed by MGH and MGB in accordance with their conflict-of-interest policies. The other authors declare no competing interests.

PubMed

Customizing CRISPR-Cas9 DNA targeting using engineering and machine learning

Abstract

Full Text

Funding

Competing interests

You found one interesting study. We’ll send the next 7.

What Lands in Your Inbox Each Week:

Recent issues from the crispr gene editing brief

Abstract

Full Text

Funding

Competing interests

Related papers

You found one interesting study. We’ll send the next 7.

What Lands in Your Inbox Each Week:

Recent issues from the crispr gene editing brief