Protein databases articles from across Nature Portfolio

A protein foundation model represents protein sequence, structure and function.

A standardized, realistic phantom dataset consisting of ground-truth annotations for six diverse molecular species is provided as a community resource for cryo-electron-tomography algorithm benchmarking.

This work introduces DeepTernary, a deep learning method for rapid and accurate prediction of PROTAC and molecular glue-induced ternary complex structures, achieving state-of-the-art results by learning from a curated dataset, TernaryDB.

In the coming years, crystallographic fragment-screening campaigns will deliver massive amounts of data and challenge existing practices and resources. In the article, options are explored for how best to preserve these data for the community to support further research and developments.

RNA-binding proteins are present in mitotic structures, but their role in cell division is unclear. Here, the authors provide an atlas of RNA-dependent mitotic proteins. They identify AURKA, KIFC1, and TPX2 as RNA-binding proteins and show that RNA regulates their interactions and functions.

A new Global Neurodegeneration Proteomics Consortium provides a wealth of data to boost mechanistic understanding, biomarker research and drug development during aging and in neurodegenerative diseases — and dementia in particular.