| Title: | Functional assessment of all ATM SNVs using prime editing and deep learning |
| Journal: | Cell |
| Published: | 27 Jun 2025 |
| Pubmed: | https://pubmed.ncbi.nlm.nih.gov/40580951/ |
| DOI: | https://doi.org/10.1016/j.cell.2025.05.046 |
Publication 16473
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
Ataxia telangiectasia mutated (ATM), a large gene with 63 exons, plays a critical role in the DNA damage response, and its loss of function increases cancer risk and affects the prognosis of cancer patients. However, interpreting the functional impact of ATM variants remains challenging because most are variants of uncertain significance (VUSs). Here, we assessed the functions of all 27,513 possible single-nucleotide variants (SNVs) in ATM. By using prime editing, we experimentally evaluated the effects of 23,092 SNVs on cell fitness in the presence of olaparib, thereby identifying critical residues. Using cancer genetics data and UK Biobank data, we found that our results are useful for estimating both cancer risk and prognosis. We also developed a deep learning model, DeepATM, which predicted such functional effects of the remaining 4,421 SNVs with unprecedentedly high accuracy. This complete evaluation of ATM variants supports precision medicine and provides a framework for addressing VUSs in other genes.</p>
9 Keywords
- Ataxia Telangiectasia Mutated Proteins
- Deep Learning
- Gene Editing
- Humans
- Neoplasms
- Phthalazines
- Piperazines
- Polymorphism, Single Nucleotide
- Prognosis
10 Authors
- Kwang Seob Lee
- Joon-Goo Min
- Yumin Cheong
- Hyeong-Cheol Oh
- Seung-Youn Jung
- Jeong-In Park
- Myungjae Song
- Jung Hwa Seo
- Sung-Rae Cho
- Hyongbum Henry Kim
1 Application
| Application ID | Title |
|---|---|
| 457621 | Prevalence and incidence of ATM-related cancers in general population. |
Enabling scientific discoveries that improve human health