The paper introduces UAE-3D, a multi-modal VAE for 3D molecule generation that compresses molecules into a unified latent space, maintaining near-zero reconstruction error. This approach simplifies latent diffusion modeling by eliminating the need to handle multi-modality and equivariance separately. Experiments on GEOM-Drugs and QM9 datasets show UAE-3D establishes new benchmarks in de novo and conditional 3D molecule generation, with significant improvements in efficiency and quality.
Pascal Fua from EPFL presented an approach to implementing convolutional neural nets that output complex 3D surface meshes. The method overcomes limitations in converting implicit representations to explicit surface representations. Applications include single view reconstruction, physically-driven shape optimization, and bio-medical image segmentation. Why it matters: This research advances geometric deep learning by enabling end-to-end trainable models for 3D surface mesh generation, with potential impact on various applications in computer vision and biomedical imaging in the region.
KAUST's Peter Wonka discusses the challenges and advancements in creating data-rich, three-dimensional maps for various applications. His team is working with Boeing on 3D modeling tools for aerospace design. KAUST-funded FalconViz uses UAV drones to create 3D maps of disaster areas for first responders. Why it matters: This highlights KAUST's contribution to cutting-edge 3D modeling and its practical applications in industries like aerospace and disaster response in the region.
KAUST researchers developed a crystallization process for organic molecules with potential applications in electronics, pharmaceuticals, and food. They produced "strained organic semiconductors," which can lead to high-performance, low-cost, flexible, and transparent electronic devices. The team combined X-ray beams with high-speed cameras to record the crystallization process, revealing that quick evaporation and nanoscale thinness play a role in producing ideal crystal lattices. Why it matters: This new method offers unprecedented control over crystal formation, potentially revolutionizing the production of plastic electronics and impacting other industries relying on specific crystal structures.
A DeepMind researcher presented work on incorporating symmetries into machine learning models, with applications to lattice-QCD and molecular dynamics. The work includes permutation and translation-invariant normalizing flows for free-energy estimation in molecular dynamics. They also presented U(N) and SU(N) Gauge-equivariant normalizing flows for pure Gauge simulations and its extensions to incorporate fermions in lattice-QCD. Why it matters: Applying symmetry principles to generative models could improve AI's ability to model complex physical systems relevant to materials science and other fields in the region.
KAUST researchers in the Functional Materials Design, Discovery & Development group have discovered a minimal edge transitive net with high connectivity. This net was used as a blueprint for the design and construction of metal-organic frameworks (MOFs). Specifically, a new rare earth nonanuclear carboxylate-based cluster was used as an 18-connected MBB to form gea-MOF-1. Why it matters: This work contributes to the advancement of solid-state materials design, which could have broad implications for energy and environmental sustainability in the region.
Daisuke Kihara from Purdue University presented a seminar at MBZUAI on using deep learning for biomolecular structure modeling. His lab is developing 3D structure modeling methods, especially for cryo-electron microscopy (cryo-EM) data. They are also working on RNA structure prediction and peptide docking using deep neural networks inspired by AlphaFold2. Why it matters: Applying advanced deep learning techniques to biomolecular structure prediction can accelerate drug discovery and our understanding of molecular functions.
KAUST researchers used cryogenic electron microscopy (cryo-EM) to study the 3D structure of protein complexes involved in DNA replication and repair. They investigated the interaction between the Y-family TLS polymerase Pol K and mono-ubiquitylated PCNA. The study revealed that DNA binding is required for Pol K to form a rigid, active complex with PCNA. Why it matters: Understanding these structural interactions may provide insights into cancer development and drug resistance mechanisms.