C-SVR Crispr: Prediction of CRISPR/Cas12 guideRNA activity using deep learning models
Problem
CRISPR gene editing is limited by low guide RNA activity, and predicting gRNA activity (indel frequency) in advance is difficult.
Why it matters
Poor gRNA activity wastes time and resources in gene editing research and potential therapies.
What the AI did
Compared Multiple Linear Regression, CNN, and a hybrid CNN-SVR regressor, optimizing convolution depth and kernel size, to predict gRNA activity.
Results
Hybrid CNN-SVR outperformed state-of-the-art by up to 40% increase in Spearman correlation for gRNA activity prediction.
Impact
Authors report improved gRNA activity prediction; no independent verification stated in source.
Record
| Venue | Alexandria Engineering Journal |
|---|---|
| Published | 2021 |
| Work type | article |
| Authors | Zubaida Said Ameen, Mehmet Özsöz, Auwalu Saleh Mubarak, Fadi M. Al-Turjman, Sertan Serte |
| Citations | 26 (OpenAlex, at import time) |
| References | 37 |
| Model or method | CNN-SVR hybrid (CNN and MLR compared) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1016/j.aej.2021.02.007 |
| Open access | gold |
| Keywords | CRISPR, Cas12, guide RNA, CNN, SVR, deep learning |
Importance rating
significant — thousands to millions affected
Abstract calls CRISPR the most important gene editing tool and low gRNA activity a key limitation; better prediction saves time across editing research and CRISPR-based detection.