Baylor College of Medicine researchers developed an advanced AI model, DeepMVP, that links protein modifications to genetic mutations influencing diseases. This powerful tool significantly outperforms existing models, enhancing therapeutic development. Proteins, vital for bodily functions, undergo post-translational modifications (PTMs) that can affect their behavior and roles in cellular processes. When PTMs are disrupted by mutations, they may contribute to health issues like cancer and heart disease. The researchers created PTMAtlas, a comprehensive database tracking nearly 400,000 PTM sites from over 241 datasets, enabling DeepMVP to predict PTM locations with high accuracy. In tests involving known mutation-PTM pairs, DeepMVP achieved 81% accuracy in identifying PTM sites and 97% in predicting the impact of mutations. This tool promises to expedite research in genetics, cancer biology, and drug development, and is freely accessible at https://deepmvp.ptmax.org/.
Source link
Share
Read more