Our paper DeepMVP: deep learning models trained on high-quality data accurately predict PTM sites and...

Our paper DeepMVP: deep learning models trained on high-quality data accurately predict PTM sites and variant-induced alterations has been published in Nature Methods. Congratulations to Bo, Chenwei, Kai, and all co-authors! This study introduces PTMAtlas, a curated compendium of 397,524 PTM sites generated through systematic reprocessing of 241 public mass-spectrometry datasets, and DeepMVP, a deep learning framework trained on PTMAtlas to predict PTM sites for phosphorylation, acetylation, methylation, sumoylation, ubiquitination and N-glycosylation. Together, they provide a robust platform for PTM research and a scalable framework for assessing the functional consequences of coding variants through the lens of PTMs. PTMAtlas, DeepMVP and a Python package for seamless integration of DeepMVP into genomics pipelines are available at http://deepmvp.ptmax.org. Among the various reports covering this study, two that appear to be AI-generated are particularly effective in making the work easily accessible and engaging to a broad audience: a youtube video and a Chinese article.