Spurred on by recent advances in neural modeling and wet-lab methods, structural biology, the study of the three-dimensional (3D) atomic structure of proteins and other macromolecules, has emerged as an area of great promise for machine learning. The shape of macromolecules is intrinsically linked to their biological function (e.g., much like the shape of a bike is critical to its transportation purposes), and thus machine learning algorithms that can better predict and reason about these shapes promise to unlock new scientific discoveries in human health as well as increase our ability to design novel medicines.
Moreover, fundamental challenges in structural biology motivate the development of new learning systems that can more effectively capture physical inductive biases, respect natural symmetries, and generalize across atomic systems of varying sizes and granularities. Through the Machine Learning in Structural Biology workshop, we aim to include a diverse range of participants and spark a conversation on the required representations and learning algorithms for atomic systems, as well as dive deeply into how to integrate these with novel wet-lab capabilities.
| Opening Remarks (Talk) | |
| Keynote -- Michael Levitt (Talk) | |
| Invited Talk - Charlotte Deane: Predicting the conformational ensembles of proteins (Talk) | |
| Invited Talk - Frank Noe: Deep Markov State Models versus Covid-19 (Talk) | |
| Invited Talk - Andrea Thorn: Finding Secondary Structure in Cryo-EM maps: HARUSPEX (Talk) | |
| Break | |
| Keynote - David Baker: Rosetta design of COVID antivirals and diagnostics (Talk) | |
| Morning Poster Session (Poster Session) | |
| Contributed Talk - Predicting Chemical Shifts with Graph Neural Networks (Talk) | |
| Contributed Talk - Cryo-ZSSR: multiple-image super-resolution based on deep internal learning (Talk) | |
| Contributed Talk - Wasserstein K-Means for Clustering Tomographic Projections (Talk) | |
| Lunch + Panel Discussion on Future of ML for Structural Biology (Starts at 1pm) (Lunch) | |
| Invited Talk - Possu Huang (Talk) | |
| Contributed talks intro (Intro) | |
| Contributed Talk - ProGen: Language Modeling for Protein Generation (Talk) | |
| Contributed Talk - Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences (Talk) | |
| Contributed Talk - SidechainNet: An All-Atom Protein Structure Dataset for Machine Learning (Talk) | |
| Contributed Talk - Generating 3D Molecular Structures Conditional on a Receptor Binding Site with Deep Generative Models (Talk) | |
| Contributed Talk - Learning from Protein Structure with Geometric Vector Perceptrons (Talk) | |
| Afternoon Poster Session (Poster Session) | |
| Invited Talk - Mohammed AlQuraishi: (Nearly) end-to-end differentiable learning of protein structure (Talk) | |
| Invited Talk - Chaok Seok: Ab initio protein structure prediction by global optimization of neural network energy: Can AI learn physics? (Talk) | |
| Concluding Remarks (Talk) | |
| Happy Hour | |
| Exploring generative atomic models in cryo-EM reconstruction (Poster Session) | |
| Design-Bench: Benchmarks for Data-Driven Offline Model-Based Optimization (Poster Session) | |
| Conservative Objective Models: A Simple Approach to Effective Model-Based Optimization (Poster Session) | |
| GEFA: Early Fusion Approach in Drug-Target Affinity Prediction (Poster Session) | |
| Protein model quality assessment using rotation-equivariant, hierarchical neural networks (Poster Session) | |
| Sequence and stucture based deep learning models for the identification of peptide binding sites (Poster Session) | |
| Fast and adaptive protein structure representations for machine learning (Poster Session) | |
| Combining variational autoencoder representations with structural descriptors improves prediction of docking scores (Poster Session) | |
| Designing a Prospective COVID-19 Therapeutic with Reinforcement Learning (Poster Session) | |
| MXMNet: A Molecular Mechanics-Driven Neural Network Based on Multiplex Graph for Molecules (Poster Session) | |
| Is Transfer Learning Necessary for Protein Landscape Prediction? (Poster Session) | |
| DHS-Crystallize: Deep-Hybrid-Sequence based method for predicting protein Crystallization (Poster Session) | |
| Cross-Modality Protein Embedding for Compound-Protein Affinity and Contact Prediction (Poster Session) | |
| Profile Prediction: An Alignment-Based Pre-Training Task for Protein Sequence Models (Poster Session) | |
| Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models (Poster Session) | |
| Pre-training Protein Language Models with Label-Agnostic Binding Pairs Enhances Performance in Downstream Tasks (Poster Session) | |
| ESM-1b: Optimizing Evolutionary Scale Modeling (Poster Session) | |
| Learning Super-Resolution Electron Density Map of Proteins using 3D U-Net (Poster Session) | |
| The structure-fitness landscape of pairwise relations in generative sequence models (Poster Session) | |