Focus Area: Drug Discovery
BoltzGen: Toward Universal Binder Design
We introduce BoltzGen, an all-atom generative model for designing proteins and peptides across all modalities to bind a wide range of biomolecular targets. BoltzGen builds strong structural reasoning capabilities about target-binder interactions into its generative design process. This is achieved by unifying design and structure prediction, resulting in a single model that also reaches state-of-the-art folding performance. BoltzGen’s generation process can be controlled with a flexible design specification language over covalent bonds, structure constraints, binding sites, and more. We experimentally validate these capabilities in a total of eight diverse wetlab design campaigns with functional and affinity readouts across 26 targets. The experiments span binder modalities from nanobodies to disulfide-bonded peptides and include targets ranging from disordered proteins to small molecules. For instance, we test 15 nanobody and protein binder designs against each of nine novel targets with low similarity to any protein with a known bound structure. For both binder modalities, this yields nanomolar binders for 66% of targets. We release model weights, data, and both inference and training code at: https://github.com/HannesStark/boltzgen.Co-authors: Hannes Stark, Felix Faltings, MinGyu Choi, Yuxin Xie, Eunsu Hur, Timothy O’Donnell, Anton Bushuiev, Talip Uçar, Saro Passaro, Weian Mao, Mateo Reveiz, Roman Bushuiev, Tomáš Pluska, Josef Sivic, Karsten Kreis, Arash Vahdat, Shamayeeta Ray, Jonathan T. Goldstein, Andrew Savinov, Jacob A. Hambalek, Anshika Gupta, Diego A. Taquiri-Diaz, Yaotian Zhang, A. Katherine Hatstat, Angelika Arada, Nam Hyeong Kim, Ethel Tackie-Yarboi, Dylan Boselli, Lee Schnaider, Chang C. Liu, Gene-Wei Li, Denes Hnisz, David M. Sabatini, William F. DeGrado, Jeremy Wohlwend, Gabriele Corso, Regina Barzilay, Tommi Jaakkola Learn more
Discovery and artificial intelligence-guided mechanistic elucidation of a narrow-spectrum antibiotic
Current clinical antibiotics are largely broad-spectrum agents that can alter the gut microbiome and promote colonization by Enterobacteriaceae, which are often drug resistant. This includes adherent-invasive Escherichia coli (AIEC), particularly in patients with inflammatory bowel disease, in which dysbiosis creates a niche for this pathogen to colonize. There is an urgent and unmet need for novel narrow-spectrum and microbiome-sparing antibiotics. Here we screened 10,747 bioactive small molecules for antibacterial activity against AIEC and discovered enterololin, an antibacterial compound with targeted activity against Enterobacteriaceae species. Enterololin could overcome intrinsic and acquired resistance mechanisms in clinical isolates when combined with a subinhibitory concentration of SPR741, a polymyxin B analogue used here to increase outer membrane permeability in Gram-negative bacteria. Molecular substructure- and deep learning-guided mechanism-of-action investigations revealed that enterololin perturbs lipoprotein trafficking through a mechanism involving the LolCDE complex, laboratory-evolved resistant mutants predominantly mapped to lolC and lolE, with an in vitro frequency of resistance of ~10−8 to 10−7. Enterololin showed low mammalian cytotoxicity (HEK293 half-maximal inhibitory concentration ~100 µg ml−1) and suppressed AIEC infection in mouse models when administered in combination with SPR741, while largely preserving the overall microbiome composition. This study highlights the utility of deep learning methods for predicting molecular interactions and identifies a promising Enterobacteriaceae-specific antibacterial candidate for further development.Co-authors: Denise B. Catacutan, Vian Tran, Autumn Arnold, Jeremie Alexander, Gabriele Corso, Yeganeh Yousefi, Megan M. Tu, Stewart McLellan, Dominique Tertigas, Jakob Magolan, Michael G. Surette, Eric D. Brown, Brian K. Coombes Learn more
Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
Advances in artificial intelligence (AI) are fueling a new paradigm of discoveries in natural sciences. Today, AI has started to advance natural sciences by improving, accelerating, and enabling our understanding of natural phenomena at a wide range of spatial and temporal scales, giving rise to a new area of research known as AI for science (AI4Science). Being an emerging research paradigm, AI4Science is unique in that it is an enormous and highly interdisciplinary area. Thus, a unified and technical treatment of this field is needed, yet challenging. This work aims to provide a technically thorough account of a subarea of AI4Science; namely, AI for quantum, atomistic, and continuum systems. These areas aim at understanding the physical world from the subatomic (wavefunctions and electron density), atomic (molecules, proteins, materials, and interactions), to macro (fluids, climate, and subsurface) scales and form an important subarea of AI4Science. A unique advantage of focusing on these areas is that they largely share a common set of challenges, thereby allowing a unified and foundational treatment. A key common challenge is how to capture physics first principles, especially symmetries, in natural systems by deep learning methods. We provide an in-depth yet intuitive account of techniques to achieve equivariance to symmetry transformations. We also discuss other common technical challenges, including explainability, out-of-distribution generalization, knowledge transfer with foundation and large language models, and uncertainty quantification. To facilitate learning and education, we provide categorized lists of resources that we found to be useful. We strive to be thorough and unified and hope this initial effort may trigger more community interests and efforts to further advance AI4Science.Co-authors: Xuan Zhang, Limei Wang, Jacob Helwig, Youzhi Luo, Cong Fu, Yaochen Xie, Meng Liu, Yuchao Lin, Zhao Xu, Keqiang Yan, Keir Adams, Maurice Weiler, Xiner Li, Tianfan Fu, Yucheng Wang, Alex Strasser, Haiyang Yu, YuQing Xie, Xiang Fu, Shenglong Xu, Yi Liu, Yuanqi Du, Alexandra Saxton, Hongyi Ling, Hannah Lawrence, Hannes Stärk, Shurui Gui, Carl Edwards, Nicholas Gao, Adriana Ladera, Tailin Wu, Elyssa F. Hofgard, Aria Mansouri Tehrani, Rui Wang, Ameya Daigavane, Montgomery Bohde, Jerry Kurtin, Qian Huang, Tuong Phung, Minkai Xu, Chaitanya K. Joshi, Simon V. Mathis, Kamyar Azizzadenesheli, Ada Fang, Alán Aspuru-Guzik, Erik Bekkers, Michael Bronstein, Marinka Zitnik, Anima Anandkumar, Stefano Ermon, Pietro Liò, Rose Yu, Stephan Günnemann, Jure Leskovec, Heng Ji, Jimeng Sun, Regina Barzilay, Tommi Jaakkola, Connor W. Coley, Xiaoning Qian, Xiaofeng Qian, Tess Smidt and Shuiwang Ji Learn more
LEAPS: A discrete neural sampler via locally equivariant networks
We propose LEAPS, an algorithm to sample from discrete distributions known up to normalization by learning a rate matrix of a continuous-time Markov chain (CTMC). LEAPS can be seen as a continuous-time formulation of annealed importance sampling and sequential Monte Carlo methods, extended so that the variance of the importance weights is offset by the inclusion of the CTMC. To derive these importance weights, we introduce a set of Radon-Nikodym derivatives of CTMCs over their path measures. Because the computation of these weights is intractable with standard neural network parameterizations of rate matrices, we devise a new compact representation for rate matrices via what we call locally equivariant functions. To parameterize them, we introduce a family of locally equivariant multilayer perceptrons, attention layers, and convolutional networks, and provide an approach to make deep networks that preserve the local equivariance. This property allows us to propose a scalable training algorithm for the rate matrix such that the variance of the importance weights associated to the CTMC are minimal. We demonstrate the efficacy of LEAPS on problems in statistical physics. We provide code in https: //github.com/malbergo/leaps/.Co-authors: Peter Holderrieth, Michael Albergo Learn more
Boltz-2: Towards accurate and efficient binding prediction
Accurately modeling biomolecular interactions is a central challenge in modern biology. While recent advances, such as AlphaFold3 and Boltz-1, have substantially improved our ability to predict biomolecular complex structures, these models still fall short in predicting binding affinity, a critical property underlying molecular function and therapeutic efficacy. Here, we present Boltz-2, a new structural biology foundation model that exhibits strong performance for both structure and affinity prediction. Boltz-2 introduces controllability features including experimental method conditioning, distance constraints, and multi-chain template integration for structure prediction, and is, to our knowledge, the first AI model to approach the performance of free-energy perturbation (FEP) methods in estimating small molecule–protein binding affinity. Crucially, it achieves strong correlation with experimental readouts on many benchmarks, while being at least 1000× more computationally efficient than FEP. By coupling Boltz-2 with a generative model for small molecules, we demonstrate an effective workflow to find diverse, synthesizable, high-affinity binders, as estimated by absolute FEP simulations on the TYK2 target. To foster broad adoption and further innovation at the intersection of machine learning and biology, we are releasing Boltz-2 weights, inference, and training code 1 under a permissive open license, providing a robust and extensible foundation for both academic and industrial research.Co-authors: Saro Passaro, Gabriele Corso, Jeremy Wohlwend, Mateo Reveiz, Stephan Thaler, Vignesh Ram Somnath, Noah Getz, Tally Portnoi, Julien Roy, Hannes Stark, David Kwabi-Addo, Dominique Beaini, Tommi Jaakkola, Regina Barzilay Learn more
ASKCOS: Open-Source, Data-Driven Synthesis Planning
The advancement of machine learning and the availability of large-scale reaction datasets have accelerated the development of data-driven models for computer-aided synthesis planning (CASP) in the past decade. In this Account, we describe the range of data-driven methods and models that have been incorporated into the newest version of ASKCOS, an open-source software suite for synthesis planning that we have been developing since 2016. This ongoing effort has been driven by the importance of bridging the gap between research and development, making research advances available through a freely available practical tool. ASKCOS integrates modules for retrosynthetic planning, modules for complementary capabilities of condition prediction and reaction product prediction, and several supplementary modules and utilities with various roles in synthesis planning. For retrosynthetic planning, we have developed an Interactive Path Planner (IPP) for user-guided search as well as a Tree Builder for automatic planning with two well-known tree search algorithms, Monte Carlo Tree Search (MCTS) and Retro*. Four one-step retrosynthesis models covering template-based and template-free strategies form the basis of retrosynthetic predictions and can be used simultaneously to combine their advantages and propose diverse suggestions. Strategies for assessing the feasibility of proposed reaction steps and evaluating the full pathways are built on top of several pioneering efforts that we have made in the subtasks of reaction condition recommendation, pathway scoring and clustering, and the prediction of reaction outcomes including the major product, impurities, site selectivity, and regioselectivity. In addition, we have also developed auxiliary capabilities in ASKCOS based on our past and ongoing work for solubility prediction and quantum mechanical descriptor prediction, which can provide more insight into the suitability of proposed reaction solvents or the hypothetical selectivity of desired transformations. For each of these capabilities, we highlight its relevance in the context of synthesis planning and present a comprehensive overview of how it is built on top of not only our work but also of other recent advancements in the field. We also describe in detail how chemists can easily interact with these capabilities via user-friendly interfaces. ASKCOS has assisted hundreds of medicinal, synthetic, and process chemists in their day-to-day tasks by complementing expert decision making and route ideation. It is our belief that CASP tools are an important part of modern chemistry research and offer ever-increasing utility and accessibility.Co-authors: Zhengkai Tu, Sourabh J. Choure, Mun Hong Fong, Jihye Roh, Itai Levin, Kevin Yu, Joonyoung F. Joung, Nathan Morgan, Shih-Cheng Li, Xiaoqi Sun, Huiqian Lin, Mark Murnin, Jordan P. Liles, Thomas J. Struble, Michael E. Fortunato, Mengjie Liu, William H. Green, Klavs F. Jensen Learn more
ShEPhERD: Diffusing shape, electrostatics, and pharmacophores for bioisosteric drug design
Engineering molecules to exhibit precise 3D intermolecular interactions with their environment forms the basis of chemical design. In ligand-based drug design, bioisosteric analogues of known bioactive hits are often identified by virtually screening chemical libraries with shape, electrostatic, and pharmacophore similarity scoring functions. We instead hypothesize that a generative model which learns the joint distribution over 3D molecular structures and their interaction profiles may facilitate 3D interaction-aware chemical design. We specifically design ShEPhERD1, an SE(3)-equivariant diffusion model which jointly diffuses/denoises 3D molecular graphs and representations of their shapes, electrostatic potential surfaces, and (directional) pharmacophores to/from Gaussian noise. Inspired by traditional ligand discovery, we compose 3D similarity scoring functions to assess ShEPhERD’s ability to conditionally generate novel molecules with desired interaction profiles. We demonstrate ShEPhERD’s potential for impact via exemplary drug design tasks including natural product ligand hopping, protein-blind bioactive hit diversification, and bioisosteric fragment merging.Co-authors: Keir Adams, Kento Abeywardane, Jenna Fromer Learn more
Protein codes promote selective subcellular compartmentalization
Cells have evolved mechanisms to distribute ~10 billion protein molecules to subcellular compartments where diverse proteins involved in shared functions must assemble. Here, we demonstrate that proteins with shared functions share amino acid sequence codes that guide them to compartment destinations. A protein language model, ProtGPS, was developed that predicts with high performance the compartment localization of human proteins excluded from the training set. ProtGPS successfully guided generation of novel protein sequences that selectively assemble in the nucleolus. ProtGPS identified pathological mutations that change this code and lead to altered subcellular localization of proteins. Our results indicate that protein sequences contain not only a folding code, but also a previously unrecognized code governing their distribution to diverse subcellular compartments.Contributors: Henry R. Kilgore, Itamar Chinn, Peter G. Mikhael, Ilan Mitnikov, Catherine Van Dongen, Guy Zylberberg, Lena Afeyan, Salman F. Banani, Susana Wilson-Hawken, Tong Ihn Lee, and Richard A. Young Learn more
Deep learning enhances the prediction of HLA class I-presented CD8+ T cell epitopes in foreign pathogens
Accurate in silico determination of CD8+ T cell epitopes would greatly enhance T cell-based vaccine development, but current prediction models are not reliably successful. Here, motivated by recent successes applying machine learning to complex biology, we curated a dataset of 651,237 unique human leukocyte antigen class I (HLA-I) ligands and developed MUNIS, a deep learning model that identifies peptides presented by HLA-I alleles. MUNIS shows improved performance compared with existing models in predicting peptide presentation and CD8+ T cell epitope immunodominance hierarchies. Moreover, application of MUNIS to proteins from Epstein–Barr virus led to successful identification of both established and novel HLA-I epitopes which were experimentally validated by in vitro HLA-I-peptide stability and T cell immunogenicity assays. MUNIS performs comparably to an experimental stability assay in terms of immunogenicity prediction, suggesting that deep learning can reduce experimental burden and accelerate identification of CD8+ T cell epitopes for rapid T cell vaccine development.Contributors:: Anusha Nathan, Nitan Shalon, Charles R. Crain, Rhoda Tano-Menka, Benjamin Goldberg, Emma Richards, Gaurav D. Gaiha Learn more
Boltz-1 Democratizing Biomolecular Interaction Modeling
Understanding biomolecular interactions is fundamental to advancing fields like drug discovery and protein design. In this paper, we introduce Boltz-1, an open-source deep learning model incorporating innovations in model architecture, speed optimization, and data processing achieving AlphaFold3-level accuracy in predicting the 3D structures of biomolecular complexes. Boltz-1 demonstrates a performance on-par with state-of-the-art commercial models on a range of diverse benchmarks, setting a new benchmark for commercially accessible tools in structural biology. By releasing the training and inference code, model weights, datasets, and benchmarks under the MIT open license, we aim to foster global collaboration, accelerate discoveries, and provide a robust platform for advancing biomolecular modeling.Contributors: Jeremy Wohlwend, Gabriele Corso, Saro Passaro, Mateo Reveiz, Ken Leidal, Wojtek Swiderski, Tally Portnoi, Itamar Chinn, Jacob Silterra Learn more
