Editing Efficiency Across Crop Families: A Systematic Review and Meta-Analysis of CRISPR/SpCas9 Knockout Outcomes in Cucurbitaceae, Brassicaceae, Solanaceae and Poaceae
This meta-analysis reveals that apparent differences in CRISPR/SpCas9 editing efficiency across major crop families are primarily driven by publication bias and within-study clustering rather than intrinsic taxonomic biology, with a bias-adjusted pooled efficiency of approximately 45%.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Imagine you're trying to bake the perfect batch of cookies, but instead of flour and sugar, you're using a high-tech molecular scissors called CRISPR to edit the DNA of plants. Scientists have been trying to figure out a simple question: How often does this editing actually work?
For a long time, the answer seemed to depend entirely on the "family" of the plant you were editing. Some researchers claimed that if you edited plants in the Brassicaceae family (like broccoli and cabbage), you'd get a huge success rate, while others said editing Poaceae (grasses like wheat and rice) was a total gamble. It was like saying, "Chocolate chip cookies always turn out great, but oatmeal raisin is a disaster."
But a new study by Olagunju and Oladunjoye decided to stop guessing and start doing the math. They gathered data from 22 different scientific studies, looking at 172 specific attempts to edit plants from four major families: cucurbits (squash), brassicas (cabbage), solanaceae (tomatoes), and poaceae (grasses).
The Big Surprise: It's Not About the Plant Family
When they first looked at the raw numbers, the "cookie family" theory seemed true. The Brassicaceae family looked like the golden child with a 73.8% success rate, while the Poaceae family looked like the underdog at 47.8%. That's a huge gap!
However, the authors realized they were falling for a trick. Imagine you have a classroom of students. If you only ask the top three students from one specific school how they did on a test, and then compare them to the top three students from a different school, you might think one school is just naturally smarter. But what if the first school just happened to have a really easy test that day, or the students studied together in a secret club?
In this study, the "secret club" was the specific laboratory doing the work. The researchers found that 64.4% of the differences in success rates came from the study itself (the lab, the specific tools, the team), not the plant family. When they used a special statistical tool to account for these "lab clusters" (called a cluster-robust test), the magic difference between plant families vanished. The math showed that family is not an independent predictor of success. The apparent gap between broccoli and wheat was just an illusion created by how the data was grouped.
The "Hidden Failures" Problem
There's another twist. The study also looked at a phenomenon called "publication bias." Think of it like a talent show where only the contestants who win the gold medal get to tell their story. If a scientist tries to edit a plant and gets 0% success, they often just throw the data in the trash and never publish it. They only publish the stories where they got a high success rate.
Because of this, the "average" success rate reported in the literature looked like 61.8%. But when the authors used a method called "trim-and-fill" to guess how many of those hidden failures were missing from the story, the real average dropped significantly. They estimated that the true success rate is closer to 45.2%.
What Does This Mean for the Future?
The authors are very careful not to say they have "solved" the problem of plant editing. Instead, they suggest that the wide range of success rates (from 5% to 98%) is the real story.
- The "Unadjusted" View: If you just look at the published papers, you might expect a 61.8% success rate.
- The "Realistic" View: If you account for the hidden failures and the fact that labs vary wildly, you should probably expect something closer to 45%.
- The "Wild Card" View: Because the results vary so much from lab to lab, the authors warn that for any specific new experiment, the success rate could be anywhere between 5% and 98%.
The Takeaway
The main lesson here isn't that some plants are "harder" to edit than others. It's that the method and the lab matter way more than the family tree of the plant. If you are a scientist planning to edit a new crop, don't just look at how well it worked for broccoli; look at how well it worked for the specific tools and techniques you plan to use.
The study concludes that we need to stop treating these numbers as a crystal ball that predicts the future. Instead, we should treat them as a budgeting tool: "If I want to get 10 edited plants, I should probably try to make 20 or 30, because the odds are a coin flip, not a guarantee." The apparent differences between plant families are just noise; the real signal is that science is messy, and every new experiment is a fresh roll of the dice.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.