We build a “dataset of datasets” to help explore the possibilities of three-dimensional molecular learning.
RNA molecules fold into complex three-dimensional shapes that are difficult to determine experimentally or predict computationally. We introduce a deep-learning method that significantly improves prediction of 3D RNA structures. Understanding these structures may aid in the discovery of drugs for currently untreatable diseases.
We combine graph and geometric aspects of biomolecular structure within a single unified framework.
We investigate the role of GPCR phosphorylation in modulating arrestin binding and conformation, revealing the structural basis for the long-standing “barcode” hypothesis.
We use a novel class of neural network architectures to accurately predict the structures of 3D protein complexes.
We present a 100x-sized protein interface prediction dataset and achieve state-of-the-art results through using 3DCNNs.
We investigate the structural mechanism of GPCR-mediated arrestin activation, providing a foundation for the design of functionally selective (‘biased’) GPCR-targeted ligands with desired effects on arrestin signalling.
A user-assisted geolocation system than can use natural skylines to often pinpoint a photo’s location to within 100m2.
