Effective molecular representation learning is of great importance to facilitate molecular property prediction. Recent advances for molecular representation learning have shown great promise in applying graph neural networks to model molecules. Moreover, a few recent studies design self-supervised learning methods for molecular representation to address insufficient labelled molecules; however, these self-supervised frameworks treat the molecules as topological graphs without fully utilizing the molecular geometry information. The molecular geometry, also known as the three-dimensional spatial structure of a molecule, is critical for determining molecular properties. To this end, we propose a novel geometry-enhanced molecular representation learning method (GEM). The proposed GEM has a specially designed geometry-based graph neural network architecture as well as several dedicated geometry-level self-supervised learning strategies to learn the molecular geometry knowledge. We compare GEM with various state-of-the-art baselines on different benchmarks and show that it can considerably outperform them all, demonstrating the superiority of the proposed method.

No abstract

The centrosome is the main microtubule-organizing center in animal cells. It comprises of two centrioles and the surrounding pericentriolar material. Protein organization at the outer layer of the centriole and outward has been studied extensively; however, an overall picture of the protein architecture at the centriole core has been missing. Here we report a direct view of Drosophila centriolar proteins at ∼50-nm resolution. This reveals a Sas6 ring at the C-terminus, where it overlaps with the C-terminus of Cep135. The ninefold symmetrical pattern of Cep135 is further conveyed through Ana1–Asterless axes that extend past the microtubule wall from between the blades. Ana3 and Rcd4, whose termini are close to Cep135, are arranged in ninefold symmetry that does not match the above axes. During centriole biogenesis, Ana3 and Rcd4 are sequentially loaded on the newly formed centriole and are required for centriole-to-centrosome conversion through recruiting the Cep135–Ana1–Asterless complex. Together, our results provide a spatiotemporal map of the centriole core and implications of how the structure might be built.

No abstract