Therapeutic antibody candidates often require extensive engineering to improve key functional and developability properties before clinical development. This can be achieved through iterative design, where starting molecules are optimized over several rounds of in vitro experiments. While protein structure can provide a strong inductive bias, it is rarely used in iterative design due to the lack of structural data for continually evolving lead molecules over the course of optimization. In this work, we propose a strategy for iterative antibody optimization that leverages both sequence and structure as well as accumulating lab measurements of binding and developability. Building on prior work, we first train a sequence-structure diffusion generative model that operates on antibody-antigen complexes. We then outline an approach to use this model, together with carefully predicted antibody-antigen complexes, to optimize lead candidates throughout the iterative design process. Further, we describe a guided sampling approach that biases generation toward desirable properties by integrating models trained on experimental data from iterative design. We evaluate our approach in multiple in silico and in vitro experiments, demonstrating that it produces high-affinity binders at multiple stages of an active antibody optimization campaign. Therapeutic antibodies are a flexible and rapidly-growing class of drugs that have already successfully been used to treat a wide range of diseases (Carter & Lazar, 2018).

Molecular docking is a cornerstone of drug discovery, relying on high-resolution ligand-bound structures to achieve accurate predictions. However, obtaining these structures is often costly and time-intensive, limiting their availability. In contrast, ligand-free structures are more accessible but suffer from reduced docking performance due to pocket geometries being less suited for ligand accommodation in apo structures. Traditional methods for artificially inducing these conformations, such as molecular dynamics simulations, are computationally expensive. In this work, we introduce Sesame, a generative model designed to predict this conformational change efficiently. By generating geometries better suited for ligand accommodation at a fraction of the computational cost, Sesame aims to provide a scalable solution for improving virtual screening workflows.

Exhaustive virtual screening is highly informative but often intractable against the expensive objective functions involved in modern drug discovery. This problem is exacerbated in combinatorial contexts such as multi-vector expansion, where molecular spaces can quickly become ultra-large. Here, we introduce Scalable Active Learning via Synthon Acquisition (SALSA): a simple algorithm applicable to multi-vector expansion which extends pool-based active learning to non-enumerable spaces by factoring modeling and acquisition over synthon or fragment choices. Through experiments on ligand-and structure-based objectives, we highlight SALSA's sample efficiency, and its ability to scale to spaces of trillions of compounds. Further, we demonstrate application toward multi-parameter objective design tasks on three protein targets - finding SALSA-generated molecules have comparable chemical property profiles to known bioactives, and exhibit greater diversity and higher scores over an industry-leading generative approach. Given the strong association between a molecule's core scaffold and its chemical properties, a common workflow is to iteratively design, make, and test changes at targeted R-groups in order to advance therapeutics through the discovery pipeline (Schneider, 2017). Exhaustive virtual screening of R-group changes aids designers and medicinal chemists in the search for promising, synthesizable molecular structures, but quickly becomes intractable against computationally expensive scores as the number of possible attachments increases.

The ability to make zero-shot predictions about the fitness consequences of protein sequence changes with pre-trained machine learning models enables many practical applications. Such models can be applied for downstream tasks like genetic variant interpretation and protein engineering without additional labeled data. The advent of capable protein structure prediction tools has led to the availability of orders of magnitude more precomputed predicted structures, giving rise to powerful structure-based fitness prediction models. Through our experiments, we assess several modeling choices for structure-based models and their effects on downstream fitness prediction. Zero-shot fitness prediction models can struggle to assess the fitness landscape within disordered regions of proteins, those that lack a fixed 3D structure. We confirm the importance of matching protein structures to fitness assays and find that predicted structures for disordered regions can be misleading and affect predictive performance. Lastly, we evaluate an additional structure-based model on the ProteinGym substitution benchmark and show that simple multi-modal ensembles are strong baselines.

Designing antibody sequences to better resemble those observed in natural human repertoires is a key challenge in biologics development. We introduce IgCraft: a multi-purpose model for paired human antibody sequence generation, built on Bayesian Flow Networks. IgCraft presents one of the first unified generative modeling frameworks capable of addressing multiple antibody sequence design tasks with a single model, including unconditional sampling, sequence inpainting, inverse folding, and CDR motif scaffolding. Our approach achieves competitive results across the full spectrum of these tasks while constraining generation to the space of human antibody sequences, exhibiting particular strengths in CDR motif scaffolding (grafting) where we achieve state-of-the-art performance in terms of humanness and preservation of structural properties. By integrating previously separate tasks into a single scalable generative model, IgCraft provides a versatile platform for sampling human antibody sequences under a variety of contexts relevant to antibody discovery and engineering. Monoclonal antibodies are an important class of therapies that comprise an increasingly large share of the global pharmaceutical market (Ecker et al., 2015). Key to the success of these molecules as therapeutics lies not only in their ability to selectively bind their target with high affinity, but also in their favorable developability, a property that broadly describes the suitability of a functional compound to become a viable drug, often a function of immunogenicity, solubility, and a number of other factors. Conventional antibody discovery typically relies on either animal immunization (Lee et al., 2014) or high-throughput screening of large sequence libraries (Bradbury et al., 2011) to isolate potential candidates. While in vitro screening methods are faster, cheaper, and have ethical advantages compared to immunization, naturally-derived antibodies tend to exhibit better developa-bility properties, including favorable pharmacokinetics, high specificity, and low immunogenicity (Jain et al., 2017).

Accurately predicting drug-drug interactions (DDIs) is crucial for pharmaceutical research and clinical safety. Recent deep learning models often suffer from high computational costs and limited generalization across datasets. In this study, we investigate a simpler yet effective approach using molecular representations such as Morgan fingerprints (MFPS), graph-based embeddings from graph convolutional networks (GCNs), and transformer-derived embeddings from MoLFormer integrated into a straightforward neural network. We benchmark our implementation on DrugBank DDI splits and a drug-drug affinity (DDA) dataset from the Food and Drug Administration. MFPS along with MoLFormer and GCN representations achieve competitive performance across tasks, even in the more challenging leak-proof split, highlighting the sufficiency of simple molecular representations. Moreover, we are able to identify key molecular motifs and structural patterns relevant to drug interactions via gradient-based analyses using the representations under study. Despite these results, dataset limitations such as insufficient chemical diversity, limited dataset size, and inconsistent labeling impact robust evaluation and challenge the need for more complex approaches. Our work provides a meaningful baseline and emphasizes the need for better dataset curation and progressive complexity scaling.

Protein language models have revolutionized structure prediction, but their nonlinear nature obscures how sequence representations inform structure prediction. While sparse autoencoders (SAEs) offer a path to interpretability here by learning linear representations in high-dimensional space, their application has been limited to smaller protein language models unable to perform structure prediction. In this work, we make two key advances: (1) we scale SAEs to ESM2-3B, the base model for ESMFold, enabling mechanistic interpretability of protein structure prediction for the first time, and (2) we adapt Matryoshka SAEs for protein language models, which learn hierarchically organized features by forcing nested groups of latents to reconstruct inputs independently. We demonstrate that our Matryoshka SAEs achieve comparable or better performance than standard architectures. Through comprehensive evaluations, we show that SAEs trained on ESM2-3B significantly outperform those trained on smaller models for both biological concept discovery and contact map prediction. Finally, we present an initial case study demonstrating how our approach enables targeted steering of ESMFold predictions, increasing structure solvent accessibility while fixing the input sequence. To facilitate further investigation by the broader community, we open-source our code, dataset, pretrained models, and visualizer.

Antibody therapies have been employed to address some of today's most challenging diseases, but must meet many criteria during drug development before reaching a patient. Humanization is a sequence optimization strategy that addresses one critical risk called immunogenicity -- a patient's immune response to the drug -- by making an antibody more'human-like' in the absence of a predictive lab-based test for immunogenicity. However, existing humanization strategies generally yield very few humanized candidates, which may have degraded biophysical properties or decreased drug efficacy. Here, we re-frame humanization as a conditional generative modeling task, where humanizing mutations are sampled from a language model trained on human antibody data. We describe a sampling process that incorporates models of therapeutic attributes, such as antigen binding affinity, to obtain candidate sequences that have both reduced immunogenicity risk and maintained or improved therapeutic properties, allowing this algorithm to be readily embedded into an iterative antibody optimization campaign. We demonstrate in silico and in lab validation that in real therapeutic programs our generative humanization method produces diverse sets of antibodies that are both (1) highly-human and (2) have favorable therapeutic properties, such as improved binding to target antigens. Antibodies are the fastest growing drug class, with approved molecules treating a breadth of disorders ranging from cancer to autoimmune disease to infectious disease (Carter & Lazar, 2018). Many candidate therapeutic antibodies are derived from non-human e.g., murine or camelid sources, and modern antibody formats such as multi-specifics or antibody-drug conjugates can require heavy sequence engineering after discovery. This increases the risk of immunogenicity, where Anti-Drug Antibodies (ADAs) result in either fast clearance of the drug or adverse events (Hwang & Foote, 2005). While antibody sequence humanness is only roughly correlated with immunogenicity, humanization is widely employed to decrease immunogenicity risk (Prihoda et al., 2022).

Cryo-EM is a vital technique for determining 3D structure of biological molecules such as proteins and viruses. The cryo-EM reconstruction problem is challenging due to the high noise levels, the missing poses of particles, and the computational demands of processing large datasets. A promising solution to these challenges lies in the use of amortized inference methods, which have shown particular efficacy in pose estimation for large datasets. However, these methods also encounter convergence issues, often necessitating sophisticated initialization strategies or engineered solutions for effective convergence. Building upon the existing cryoAI pipeline, which employs a symmetric loss function to address convergence problems, this work explores the emergence and persistence of these issues within the pipeline. Additionally, we explore the impact of equivariant amortized inference on enhancing convergence. Our investigations reveal that, when applied to simulated data, a pipeline incorporating an equivariant encoder not only converges faster and more frequently than the standard approach but also demonstrates superior performance in terms of pose estimation accuracy and the resolution of the reconstructed volume. Cryo-electron microscopy (cryo-EM) has emerged as an crucial technique in molecular biology and chemistry, enabling the determination of macro-molecular structures such as proteins. In cryo-EM, particle samples are frozen in a thin layer of vitreous ice and exposed to an electron beam. The interaction between electrons and the sample's electrostatic potential scatters electrons in patterns that reflect the molecular structure.

Protein-protein interactions (PPIs) play a crucial role in numerous biological processes. Developing methods that predict binding affinity changes under substitution mutations is fundamental for modelling and re-engineering biological systems. Deep learning is increasingly recognized as a powerful tool capable of bridging the gap between in-silico predictions and in-vitro observations. With this contribution, we propose eGRAL, a novel SE(3) equivariant graph neural network (eGNN) architecture designed for predicting binding affinity changes from multiple amino acid substitutions in protein complexes. eGRAL leverages residue, atomic and evolutionary scales, thanks to features extracted from protein large language models. To address the limited availability of large-scale affinity assays with structural information, we generate a simulated dataset comprising approximately 500,000 data points. Our model is pre-trained on this dataset, then fine-tuned and tested on experimental data.