Biology subjects

Diaz, D. J.

Publications and source records attributed to Diaz, D. J..

13 recordsLinked to original sources

bioRxiv2026

Folding scFv--Antigen Complexes at Scale

AO_SCPLOWBSTRACTC_SCPLOWAccurate modeling of antibody-antigen (Ab-Ag) complexes is central to biologic development, yet the reliability and failures of modern Ab-Ag folding pipelines remain poorly characterized. Single-chain variable fragments (scFvs) are thera-peutically important antibodies, but large-scale evaluations of structure prediction models on scFv-Ag complexes are largely lacking. We introduce a scalable bench-marking pipeline that generates large ensembles of scFv-Ag structure predictions by cofolding a curated subset of 3,800 Ab-Ag complexes from SAbDab using multiple state-of-the-art models under diverse inference-time settings. The resulting dataset, SCALE (scFv-Ag CompLex Ensembles), includes standardized scFv-Ag sequences and around 200,000 predicted complexes spanning different models, sampling strategies, and auxiliary inputs. Using SCALE, we evaluate model performance in recovering correct scFv-Ag interfaces and assess the ability of existing confidence metrics to select the best structure from prediction ensembles. We find that while confidence scores effectively distinguish easy from hard scFv-Ag complexes, they often fail to identify the highest-quality interface for a given target. Further analysis shows that near-correct interfaces typically appear in ensembles but at low frequency, and inference-time choices such as sampling, recycling, and using evolutionary or structural information are crucial for accurate scFv-Ag complex predictions. Dataset and analysis code are available at https://huggingface.co/datasets/ravishah1/SCALE

bioinformatics↗

bioRxiv2026

Promera: a unified model for biomolecular structure prediction, filtering, and design

Generative models have become staple tools for modeling and designing biomolecular structures. However, although these tools have improved in structural prediction accuracy, their ability to filter designed binders--an essential use case--remains insufficient; whereas design methods have focused more on unconstrained binder generation rather than capabilities enabled by controllable design. We introduce Promera, a unified generative model that combines all-atom structure prediction with improved filtering and controllable design. We find that Promeras confidence metrics are more accurate for filtering binders from non-binders for both miniproteins and nanobodies, while its co-folding performance surpasses popular open-source models (OpenFold3-p2, Boltz-2) on therapeutically relevant categories. As a design model, Promera generates binders by predicting masked protein sequences with optional epitope, paratope, and template constraints. Remarkably, our nanobody designs match the in silico success rates from backprop-based techniques (mBER) when evaluated under co-folding confidence filters. We further provide two in silico demonstrations of the the versatile capabilities of our design method: epitope targeting of the Andes hantavirus glycoprotein with VHHs and active state stabilization of the {beta}2 andrenergic GPCR. We conclude by proposing a scaling law for co-folding models, suggesting a path for further performance improvement.

bioinformatics↗

bioRxiv2025

Efficient Search of Ultra-Large Synthesis On-Demand Libraries with Chemical Language Models

Ultra-large building block catalogs provide inexpensive access to billions of synthesis-on-demand molecules, but the combinatorial scale renders conventional virtual screening impractical. We present Vector Virtual Screen (VVS), a score-function-agnostic machine learning framework for efficient navigation of combinatorial libraries and rapid identification of promising molecules for experimental validation. VVS comprises four key innovations: (i) the Embedding Decomposer, which factors molecules into building blocks in latent space; (ii) ChemRank, a correlation-based loss that improves retrieval precision; (iii) BBKNN, an algorithm for nearest-neighbor search directly in building block space; and (iv) a multi-scale hill-climbing algorithm for gradient-based navigation of molecular embedding vector databases. Across diverse scoring functions, VVS consistently outperforms existing methods in retrieving high-scoring molecules while evaluating only a fraction of the library, achieving orders-of-magnitude runtime improvements. By turning ultra-large libraries into tractable search spaces, VVS enables virtual screening to keep pace with the rapid expansion of chemical space and adapt seamlessly to future advances in scoring functions.

bioinformatics↗

bioRxiv2025

Generating functional and multistate proteins with a multimodal diffusion transformer

Generating proteins with the full diversity and complexity of functions found in nature is a grand challenge in protein design. Here, we present ProDiT, a multimodal diffusion model that unifies sequence and structure modeling paradigms to enable the design of functional proteins at scale. Trained on sequences, 3D structures, and annotations for 214M proteins across the evolutionary landscape, ProDiT generates diverse, novel proteins that preserve known active and binding site motifs and can be successfully conditioned on a wide range of molecular functions, spanning 465 Gene Ontology terms. We introduce a diffusion sampling protocol to design proteins with multiple functional states, and demonstrate this protocol by scaffolding enzymatic active sites from carbonic anhydrase and lysozyme to be allosterically deactivated by a calcium effector. Our results showcase ProDiTs unique capacity to satisfy design specifications inaccessible to existing generative models, thereby expanding the protein design toolkit.

bioinformatics↗

bioRxiv2025

Ambient Proteins: Training Diffusion Models on Low Quality Structures

We present Ambient Protein Diffusion, a framework for training protein diffusion models that generates structures with unprecedented diversity and quality. State-of-the-art generative models are trained on computationally derived structures from AlphaFold2 (AF), as experimentally determined structures are relatively scarce. The resulting models are therefore limited by the quality of synthetic datasets. Since the accuracy of AF predictions degrades with increasing protein length and complexity, de novo generation of long, complex proteins remains challenging. Ambient Protein Diffusion overcomes this problem by treating low-confidence AF structures as corrupted data. Rather than simply filtering out low-quality AF structures, our method adjusts the diffusion objective for each structure based on its corruption level, allowing the model to learn from both high and low quality structures. Empirically, Ambient Protein Diffusion yields major improvements: on proteins with 700 residues, diversity increases from 45% to 86% from the previous state-of-the-art, and designability improves from 68% to 86%. We will make all of our code, models and datasets available under the following repository: https://github.com/jozhang97/ambient-proteins.

bioengineering↗

bioRxiv2024

Distilling structural representations into protein sequence models

Protein language models, like the popular ESM2, are widely used tools for extracting evolution-based protein representations and have achieved significant success on downstream biological tasks. Representations based on sequence and structure models, however, show significant performance differences depending on the downstream task. A major open problem is to obtain representations that best capture both the evolutionary and structural properties of proteins in general. Here we introduce Implicit Structure Model (ISM), a sequence-only input model with structurally-enriched representations that outperforms state-of-the-art sequence models on several well-studied benchmarks including mutation stability assessment and structure prediction. Our key innovations are a microenvironment-based autoencoder for generating structure tokens and a self-supervised training objective that distills these tokens into ESM2s pre-trained model. We have made ISMs structure-enriched weights easily available: integrating ISM into any application using ESM2 requires changing only a single line of code. Our code is available at https://github.com/jozhang97/ISM.

bioinformatics↗

bioRxiv2024

A systematic evaluation of the language-of-viral-escape model using multiple machine learning frameworks

Predicting the evolutionary patterns of emerging and endemic viruses is key for mitigating their spread in host populations. In particular, it is critical to rapidly identify mutations with the potential for immune escape or increased disease burden (variants of concern). Knowing which circulating mutations are such variants of concern can inform treatment or mitigation strategies such as alternative vaccines or targeted social distancing. A recent study proposed that variants of concern can be identified using two quantities extracted from protein language models, grammaticality and semantic change. These quantities are defined in analogy to concepts from natural language processing. Grammaticality is intended to be a measure of whether a variant viral protein is viable, and semantic change is intended to be a measure of potential for immune escape. Here, we systematically test this hypothesis, taking advantage of several high-throughput datasets that have become available, and also testing additional machine learning models for calculating the grammaticality metric. We find that grammaticality can be a measure of protein viability, though the more traditional metric {Delta}{Delta}G appears to be more effective. By contrast, we do not find compelling evidence that semantic change is a useful tool for identifying immune escape mutations.

bioinformatics↗

bioRxiv2023

Stability Oracle: A Structure-Based Graph-Transformer for Identifying Stabilizing Mutations

Stabilizing proteins is a fundamental challenge in protein engineering and is almost always a prerequisite for the development of industrial and pharmaceutical biotechnologies. Here we present Stability Oracle: a structure-based graph-transformer framework that achieves state-of-the-art performance on predicting the effect of a point mutation on a proteins thermodynamic stability ({Delta}{Delta}G). A strength of our model is its ability to identify stabilizing mutations, which often make up a small fraction of a proteins mutational landscape. Our framework introduces several data and machine learning innovations to overcome well-known challenges in data scarcity and bias, generalization, and computation time. Stability Oracle is first pretrained on over 2M masked microenvironments and then fine-tuned using a novel data augmentation technique, Thermodynamic Permutations (TP), applied to a [~]120K curated subset of the mega-scale cDNA display proteolysis dataset. This technique increases the original 120K mutations to over 2M thermodynamically valid {Delta}{Delta}G measurements to generate the first structure training set that samples and balances all 380 mutation types. By using the masked microenvironment paradigm, Stability Oracle does not require a second mutant structure and instead uses amino acid structural embeddings to represent a mutation. This architectural design accelerates training and inference times: we can both train on 2M instances with just 119 structures and generate deep mutational scan (DMS) predictions from only the wildtype structure. We benchmark Stability Oracle with both experimental and AlphaFold structures of all proteins on T2837, a test set that aggregates the common test sets (SSym, S669, p53, and Myoglobin) with all additional experimental data from proteins with over a 30% sequence similarity overlap. We used TP augmented T2837 to evaluate performance for engineering protein stability: Stability Oracle correctly identifies 48% of stabilizing mutations ({Delta}{Delta}G < -0.5 kcal/mol) and 74% of its stabilizing predictions are indeed stabilizing (18% and 8% of predictions were neutral and destabilizing, respectively). For a fair comparison between sequence and structure-based fine-tuned deep learning models, we build on the Prostata framework and fine-tune the sequence embeddings of ESM2 on our training set (Prostata-IFML). A head-to-head comparison demonstrates that Stability Oracle outperforms Prostata-IFML on regression and classification even though the model is 548 times smaller and is pretrained with 4000 times fewer proteins, highlighting the advantages of learning from structures.

biochemistry↗

bioRxiv2023

Synthetic microbial sensing and biosynthesis of amaryllidaceae alkaloids

A major challenge to achieving industry-scale biomanufacturing of therapeutic alkaloids is the slow process of biocatalyst engineering. Amaryllidaceae alkaloids, such as the Alzheimers medication galantamine, are complex plant secondary metabolites with recognized therapeutic value. Due to their difficult synthesis they are regularly sourced by extraction and purification from low-yielding plants, including the wild daffodil Narcissus pseudonarcissus. Engineered biocatalytic methods have the potential to stabilize the supply chain of amaryllidaceae alkaloids. Here, we propose a highly efficient biosensor-AI technology stack for biocatalyst development, which we apply to engineer amaryllidaceae alkaloid production in Escherichia coli. Directed evolution is used to develop a highly sensitive (EC50= 20 uM) and specific biosensor for the key amaryllidaceae alkaloid branchpoint 4-OMethylnorbelladine. A machine learning model (MutComputeX) was subsequently developed and used to generate activity-enriched variants of a plant methyltransferase, which were rapidly screened with the biosensor. Functional enzyme variants were identified that yielded a 60% improvement in product titer, 17-fold reduced remnant substrate, and 3-fold lower off-product regioisomer formation.

synthetic biology↗

bioRxiv2023

Sequence and structure based deep learning models represent different aspects of protein biochemistry

Deep learning models are seeing increased use as methods to predict mutational effects or allowed mutations in proteins. The models commonly used for these purposes include large language models (LLMs) and 3D Convolutional Neural Networks (CNNs). These two model types have very different architectures and are commonly trained on different representations of proteins. LLMs make use of the transformer architecture and are trained purely on protein sequences whereas 3D CNNs are trained on voxelized representations of local protein structure. While comparable overall prediction accuracies have been reported for both types of models, it is not known to what extent these models make comparable specific predictions and/or generalize protein biochemistry in similar ways. Here, we perform a systematic comparison of two LLMs and two structure-based models (CNNs) and show that the different model types have distinct strengths and weaknesses. The overall prediction accuracies are largely uncorrelated between the sequence- and structure-based models. Overall, the two structure-based models are better at predicting buried aliphatic and hydrophobic residues whereas the two LLMs are better at predicting solvent-exposed polar and charged amino acids. Finally, we find that a combined model that takes the individual model predictions as input can leverage these individual model strengths and results in significantly improved overall prediction accuracy.

biochemistry↗

bioRxiv2021

Deep learning redesign of PETase for practical PET degrading applications

Plastic waste poses an ecological challenge1. While current plastic waste management largely relies on unsustainable, energy-intensive, or even hazardous physicochemical and mechanical processes, enzymatic degradation offers a green and sustainable route for plastic waste recycling2. Poly(ethylene terephthalate) (PET) has been extensively used in packaging and for the manufacture of fabrics and single-used containers, accounting for 12% of global solid waste3. The practical application of PET hydrolases has been hampered by their lack of robustness and the requirement for high processing temperatures. Here, we use a structure-based, deep learning algorithm to engineer an extremely robust and highly active PET hydrolase. Our best resulting mutant (FAST-PETase: Functional, Active, Stable, and Tolerant PETase) exhibits superior PET-hydrolytic activity relative to both wild-type and engineered alternatives, (including a leaf-branch compost cutinase and its mutant4) and possesses enhanced thermostability and pH tolerance. We demonstrate that whole, untreated, post-consumer PET from 51 different plastic products can all be completely degraded by FAST-PETase within one week, and in as little as 24 hours at 50 {degrees}C. Finally, we demonstrate two paths for closed-loop PET recycling and valorization. First, we re-synthesize virgin PET from the monomers recovered after enzymatic depolymerization. Second, we enable in situ microbially-enabled valorization using a Pseudomonas strain together with FAST-PETase to degrade PET and utilize the evolved monomers as a carbon source for growth and polyhydroxyalkanoate production. Collectively, our results demonstrate the substantial improvements enabled by deep learning and a viable route for enzymatic plastic recycling at the industrial scale.

bioengineering↗

bioRxiv2021

Learning the local landscape of protein structures with convolutional neural networks

The fundamental problem of protein biochemistry is to predict protein structure from amino acid sequence. The inverse problem, predicting either entire sequences or individual mutations that are consistent with a given protein structure, has received much less attention even though it has important applications in both protein engineering and evolutionary biology. Here, we ask whether 3D convolutional neural networks (3D CNNs) can learn the local fitness landscape of protein structure to reliably predict either the wild-type amino acid or the consensus in a multiple sequence alignment from the local structural context surrounding a site of interest. We find that the network can predict wild type with good accuracy, and that network confidence is a reliable measure of whether a given prediction is likely going to be correct or not. Predictions of consensus are less accurate, and are primarily driven by whether or not the consensus matches the wild type. Our work suggests that high-confidence mis-predictions of the wild type may identify sites that are primed for mutation and likely targets for protein engineering.

bioinformatics↗

bioRxiv2019

A structure-based deep learning framework for protein engineering

While deep learning methods exist to guide protein optimization, examples of novel proteins generated with these techniques require a priori mutational data. Here we report a 3D convolutional neural network that associates amino acids with neighboring chemical microenvironments at state-of-the-art accuracy. This algorithm enables identification of novel gain-of-function mutations, and subsequent experiments confirm substantive phenotypic improvements in stability-associated phenotypes in vivo across three diverse proteins.

synthetic biology↗