Deep learning has transformed protein design, enabling accurate structure prediction, sequence optimization, and de novo protein generation. Advances in single-chain protein structure prediction via AlphaFold2, RoseTTAFold, ESMFold, and others have achieved near-experimental accuracy, inspiring successive work extended to biomolecular complexes via AlphaFold Multimer, RoseTTAFold All-Atom, AlphaFold 3, Chai-1, Boltz-1 and others. Generative models such as ProtGPT2, ProteinMPNN, and RFdiffusion have enabled sequence and backbone design beyond natural evolution-based limitations. More recently, joint sequence-structure co-design models, including ESM3, have integrated both modalities into a unified framework, resulting in improved designability. Despite these advances, challenges still exist pertaining to modeling sequence-structure-function relationships and ensuring robust generalization beyond the regions of protein space spanned by the training data. Future advances will likely focus on joint sequence-structure-function co-design frameworks that are able to model the fitness landscape more effectively than models that treat these modalities independently. Current capabilities, coupled with the dizzying rate of progress, suggest that the field will soon enable rapid, rational design of proteins with tailored structures and functions that transcend the limitations imposed by natural evolution. In this review, we discuss the current capabilities of deep learning methods for protein design, focusing on some of the most revolutionary and capable models with respect to their functionality and the applications that they enable, leading up to the current challenges of the field and the optimal path forward.
Protein-specific large language models (Protein LLMs) are revolutionizing protein science by enabling more efficient protein structure prediction, function annotation, and design. While existing surveys focus on specific aspects or applications, this work provides the first comprehensive overview of Protein LLMs, covering their architectures, training datasets, evaluation metrics, and diverse applications. Through a systematic analysis of over 100 articles, we propose a structured taxonomy of state-of-the-art Protein LLMs, analyze how they leverage large-scale protein sequence data for improved accuracy, and explore their potential in advancing protein engineering and biomedical research. Additionally, we discuss key challenges and future directions, positioning Protein LLMs as essential tools for scientific discovery in protein science. Resources are maintained at https://github.com/Yijia-Xiao/Protein-LLM-Survey.
The motif-scaffolding problem is a central task in computational protein design: Given the coordinates of atoms in a geometry chosen to confer a desired biochemical function (a motif), the task is to identify diverse protein structures (scaffolds) that include the motif and maintain its geometry. Significant recent progress on motif-scaffolding has been made due to computational evaluation with reliable protein structure prediction and fixed-backbone sequence design methods. However, significant variability in evaluation strategies across publications has hindered comparability of results, challenged reproducibility, and impeded robust progress. In response we introduce MotifBench, comprising (1) a precisely specified pipeline and evaluation metrics, (2) a collection of 30 benchmark problems, and (3) an implementation of this benchmark and leaderboard at github.com/blt2114/MotifBench. The MotifBench test cases are more difficult compared to earlier benchmarks, and include protein design problems for which solutions are known but on which, to the best of our knowledge, state-of-the-art methods fail to identify any solution.
Although machine learning has transformed protein structure prediction of folded protein ground states with remarkable accuracy, intrinsically disordered proteins and regions (IDPs/IDRs) are defined by diverse and dynamical structural ensembles that are predicted with low confidence by algorithms such as AlphaFold. We present a new machine learning method, IDPForge (Intrinsically Disordered Protein, FOlded and disordered Region GEnerator), that exploits a transformer protein language diffusion model to create all-atom IDP ensembles and IDR disordered ensembles that maintains the folded domains. IDPForge does not require sequence-specific training, back transformations from coarse-grained representations, nor ensemble reweighting, as in general the created IDP/IDR conformational ensembles show good agreement with solution experimental data, and options for biasing with experimental restraints are provided if desired. We envision that IDPForge with these diverse capabilities will facilitate integrative and structural studies for proteins that contain intrinsic disorder.
PyMOLfold is a flexible and open-source plugin designed to seamlessly integrate AI-based protein structure prediction and visualization within the widely used PyMOL molecular graphics system. By leveraging state-of-the-art protein folding models such as ESM3, Boltz-1, and Chai-1, PyMOLfold allows researchers to directly predict protein tertiary structures from amino acid sequences without requiring external tools or complex workflows. Furthermore, with certain models, users can provide a SMILES string of a ligand and have the small molecule placed in the protein structure. This unique capability bridges the gap between computational folding and structural visualization, enabling users to input a primary sequence, perform a folding prediction, and immediately explore the resulting 3D structure within the same intuitive platform.
Since the completion of the human genome sequencing project in 2001, significant progress has been made in areas such as gene regulation editing and protein structure prediction. However, given the vast amount of genomic data, the segments that can be fully annotated and understood remain relatively limited. If we consider the genome as a book, constructing its equivalents of words, sentences, and paragraphs has been a long-standing and popular research direction. Recently, studies on transfer learning in large language models have provided a novel approach to this challenge.Multilingual transfer ability, which assesses how well models fine-tuned on a source language can be applied to other languages, has been extensively studied in multilingual pre-trained models. Similarly, the transfer of natural language capabilities to "DNA language" has also been validated. Building upon these findings, we first trained a foundational model capable of transferring linguistic capabilities from English to DNA sequences. Using this model, we constructed a vocabulary of DNA words and mapped DNA words to their English equivalents.Subsequently, we fine-tuned this model using English datasets for paragraphing and sentence segmentation to develop models capable of segmenting DNA sequences into sentences and paragraphs. Leveraging these models, we processed the GRCh38.p14 human genome by segmenting, tokenizing, and organizing it into a "book" comprised of genomic "words," "sentences," and "paragraphs." Additionally, based on the DNA-to-English vocabulary mapping, we created an "English version" of the genomic book. This study offers a novel perspective for understanding the genome and provides exciting possibilities for developing innovative tools for DNA search, generation, and analysis.
Recent breakthroughs in large-scale generative modeling have demonstrated the potential of foundation models in domains such as natural language, computer vision, and protein structure prediction. However, their application in the energy and smart grid sector remains limited due to the scarcity and heterogeneity of high-quality data. In this work, we propose a method for creating high-fidelity electricity consumption time series data for rare and unseen context variables (e.g. location, building type, photovoltaics). Our approach, Context Encoding and Normalizing Time Series Generation, or CENTS, includes three key innovations: (i) A context normalization approach that enables inverse transformation for time series context variables unseen during training, (ii) a novel context encoder to condition any state-of-the-art time-series generator on arbitrary numbers and combinations of context variables, (iii) a framework for training this context encoder jointly with a time-series generator using an auxiliary context classification loss designed to increase expressivity of context embeddings and improve model performance. We further provide a comprehensive overview of different evaluation metrics for generative time series models. Our results highlight the efficacy of the proposed method in generating realistic household-level electricity consumption data, paving the way for training larger foundation models in the energy domain on synthetic as well as real-world data.
Considering the significance of proteins, computational protein science has always been a critical scientific field, dedicated to revealing knowledge and developing applications within the protein sequence-structure-function paradigm. In the last few decades, Artificial Intelligence (AI) has made significant impacts in computational protein science, leading to notable successes in specific protein modeling tasks. However, those previous AI models still meet limitations, such as the difficulty in comprehending the semantics of protein sequences, and the inability to generalize across a wide range of protein modeling tasks. Recently, LLMs have emerged as a milestone in AI due to their unprecedented language processing & generalization capability. They can promote comprehensive progress in fields rather than solving individual tasks. As a result, researchers have actively introduced LLM techniques in computational protein science, developing protein Language Models (pLMs) that skillfully grasp the foundational knowledge of proteins and can be effectively generalized to solve a diversity of sequence-structure-function reasoning problems. While witnessing prosperous developments, it's necessary to present a systematic overview of computational protein science empowered by LLM techniques. First, we summarize existing pLMs into categories based on their mastered protein knowledge, i.e., underlying sequence patterns, explicit structural and functional information, and external scientific languages. Second, we introduce the utilization and adaptation of pLMs, highlighting their remarkable achievements in promoting protein structure prediction, protein function prediction, and protein design studies. Then, we describe the practical application of pLMs in antibody design, enzyme design, and drug discovery. Finally, we specifically discuss the promising future directions in this fast-growing field.
Diffusion models represent a class of generative models that produce data by denoising a sample corrupted by white noise. Despite the success of diffusion models in computer vision, audio synthesis, and point cloud generation, so far they overlook inherent multiscale structures in data and have a slow generation process due to many iteration steps. In physics, the renormalization group offers a fundamental framework for linking different scales and giving an accurate coarse-grained model. Here we introduce a renormalization group-based diffusion model that leverages multiscale nature of data distributions for realizing a high-quality data generation. In the spirit of renormalization group procedures, we define a flow equation that progressively erases data information from fine-scale details to coarse-grained structures. Through reversing the renormalization group flows, our model is able to generate high-quality samples in a coarse-to-fine manner. We validate the versatility of the model through applications to protein structure prediction and image generation. Our model consistently outperforms conventional diffusion models across standard evaluation metrics, enhancing sample quality and/or accelerating sampling speed by an order of magnitude. The proposed method alleviates the need for data-dependent tuning of hyperparameters in the generative diffusion models, showing promise for systematically increasing sample efficiency based on the concept of the renormalization group.
Proteins are sequences of amino acids that serve as the basic building blocks of living organisms. Despite rapidly growing databases documenting structural and functional information for various protein sequences, our understanding of proteins remains limited because of the large possible sequence space and the complex inter- and intra-molecular forces. Deep learning, which is characterized by its ability to learn relevant features directly from large datasets, has demonstrated remarkable performance in fields such as computer vision and natural language processing. It has also been increasingly applied in recent years to the data-rich domain of protein sequences with great success, most notably with Alphafold2's breakout performance in the protein structure prediction. The performance improvements achieved by deep learning unlocks new possibilities in the field of protein bioinformatics, including protein design, one of the most difficult but useful tasks. In this paper, we broadly categorize problems in protein bioinformatics into three main categories: 1) structural prediction, 2) functional prediction, and 3) protein design, and review the progress achieved from using deep learning methodologies in each of them. We expand on the main challenges of the protein design problem and highlight how advances in structural and functional prediction have directly contributed to design tasks. Finally, we conclude by identifying important topics and future research directions.
We address protein structure prediction in the 3D Hydrophobic-Polar lattice model through two novel deep learning architectures. For proteins under 36 residues, our hybrid reservoir-based model combines fixed random projections with trainable deep layers, achieving optimal conformations with 25% fewer training episodes. For longer sequences, we employ a long short-term memory network with multi-headed attention, matching best-known energy values. Both architectures leverage a stabilized Deep Q-Learning framework with experience replay and target networks, demonstrating consistent achievement of optimal conformations while significantly improving training efficiency compared to existing methods.
AlphaFold, a groundbreaking protein prediction model, has revolutionized protein structure prediction, populating the AlphaFold Protein Database (AFDB) with millions of predicted structures. However, AlphaFold's accuracy in predicting proteins with intricate topologies, such as knots, remains a concern. This study investigates AlphaFold's performance in predicting knotted proteins and explores potential solutions to enhance the AFDB's reliability. Forty-five experimentally verified knotted protein structures from the KnotProt database were compared to their AlphaFold-generated counterparts. Knot analysis was performed using PyKnot, a PyMOL plugin, employing both Gauss codes and Alexander-Briggs knot notations. Results showed 95.6% accuracy in predicting the general shape of knots using Alexander-Briggs notation. However, Gauss code analysis revealed a 55.6% discrepancy, indicating AlphaFold's limitations in accurately representing the intricate orientation and directionality of knots. This Applications of Knot Theory for the improvement of the AlphaFold Protein Database suggests potential inaccuracies in a significant portion of the AFDB's knotted protein structures. The study underscores the need for improved knot representation in AlphaFold and proposes potential solutions, including transitioning to a single-module design or removing incorrectly predicted structures from the AFDB. These findings highlight the importance of continuous refinement for AI-based protein structure prediction tools to ensure the accuracy and reliability of protein databases for research and drug development.
Data scarcity and distribution shifts often hinder the ability of machine learning models to generalize when applied to proteins and other biological data. Self-supervised pre-training on large datasets is a common method to enhance generalization. However, striving to perform well on all possible proteins can limit model's capacity to excel on any specific one, even though practitioners are often most interested in accurate predictions for the individual protein they study. To address this limitation, we propose an orthogonal approach to achieve generalization. Building on the prevalence of self-supervised pre-training, we introduce a method for self-supervised fine-tuning at test time, allowing models to adapt to the test protein of interest on the fly and without requiring any additional data. We study our test-time training (TTT) method through the lens of perplexity minimization and show that it consistently enhances generalization across different models, their scales, and datasets. Notably, our method leads to new state-of-the-art results on the standard benchmark for protein fitness prediction, improves protein structure prediction for challenging targets, and enhances function prediction accuracy.
Artificial intelligence based on machine learning and deep learning has made significant advances in various fields such as protein structure prediction and climate modeling. However, a central challenge remains: the "black box" nature of AI, where precise quantitative relationships between inputs and outputs are often lacking. Here, by analyzing 323 AI models trained to predict human essential proteins, we uncovered a ratio law showing that model performance and the ratio of minority to majority samples can be closely linked by two concise equations. Moreover, we mathematically proved that an AI model achieves its optimal performance on a balanced dataset. More importantly, we next explore whether this finding can further guide us to enhance AI models' performance. Therefore, we divided the imbalanced dataset into several balanced subsets to train base classifiers, and then applied a bagging-based ensemble learning strategy to combine these base models. As a result, the equation-guided strategy substantially improved model performance, with increases of 4.06% and 5.28%, respectively, outperforming traditional dataset balancing techniques. Finally, we confirmed the broad applicability and generalization of these equations using different types of classifiers and 10 additional, diverse binary classification tasks. In summary, this study reveals two equations precisely linking AI's input and output, which could be helpful for unboxing the mysterious "black box" of AI.
Constructing atomic models from cryo-electron microscopy (cryo-EM) maps is a crucial yet intricate task in structural biology. While advancements in deep learning, such as convolutional neural networks (CNNs) and graph neural networks (GNNs), have spurred the development of sophisticated map-to-model tools like DeepTracer and ModelAngelo, their efficacy notably diminishes with low-resolution maps beyond 4 {\AA}. To address this shortfall, our research introduces DeepTracer-LowResEnhance, an innovative framework that synergizes a deep learning-enhanced map refinement technique with the power of AlphaFold. This methodology is designed to markedly improve the construction of models from low-resolution cryo-EM maps. DeepTracer-LowResEnhance was rigorously tested on a set of 37 protein cryo-EM maps, with resolutions ranging between 2.5 to 8.4 {\AA}, including 22 maps with resolutions lower than 4 {\AA}. The outcomes were compelling, demonstrating that 95.5\% of the low-resolution maps exhibited a significant uptick in the count of total predicted residues. This denotes a pronounced improvement in atomic model building for low-resolution maps. Additionally, a comparative analysis alongside Phenix's auto-sharpening functionality delineates DeepTracer-LowResEnhance's superior capability in rendering more detailed and precise atomic models, thereby pushing the boundaries of current computational structural biology methodologies.
Recent advancements in protein structure prediction, particularly AlphaFold2, have revolutionized structural biology by achieving near-experimental accuracy ($\text{average RMSD} < 1.5\text{\AA}$). However, the computational demands of these models (approximately 30 minutes per protein on an RTX 4090) significantly limit their application in high-throughput protein screening. While large language models like ESM (Evolutionary Scale Modeling) have shown promise in extracting structural information directly from protein sequences, rapid assessment of protein structure quality for large-scale analyses remains a major challenge. We introduce pLDDT-Predictor, a high-speed protein screening tool that achieves a $250,000\times$ speedup compared to AlphaFold2 by leveraging pre-trained ESM2 protein embeddings and a Transformer architecture. Our model predicts AlphaFold2's pLDDT (predicted Local Distance Difference Test) scores with a Pearson correlation of 0.7891 and processes proteins in just 0.007 seconds on average. Using a comprehensive dataset of 1.5 million diverse protein sequences (ranging from 50 to 2048 amino acids), we demonstrate that pLDDT-Predictor accurately classifies high-confidence structures (pLDDT $>$ 70) with 91.2\% accuracy and achieves an MSE of 84.8142 compared to AlphaFold2's predictions. The source code and pre-trained models are freely available at https://github.com/jw-chae/pLDDT_Predictor, enabling the research community to perform rapid, large-scale protein structure quality assessments.
Protein structure is key to understanding protein function and is essential for progress in bioengineering, drug discovery, and molecular biology. Recently, with the incorporation of generative AI, the power and accuracy of computational protein structure prediction/design have been improved significantly. However, ethical concerns such as copyright protection and harmful content generation (biosecurity) pose challenges to the wide implementation of protein generative models. Here, we investigate whether it is possible to embed watermarks into protein generative models and their outputs for copyright authentication and the tracking of generated structures. As a proof of concept, we propose a two-stage method FoldMark as a generalized watermarking strategy for protein generative models. FoldMark first pretrain watermark encoder and decoder, which can minorly adjust protein structures to embed user-specific information and faithfully recover the information from the encoded structure. In the second step, protein generative models are fine-tuned with watermark-conditioned Low-Rank Adaptation (LoRA) modules to preserve generation quality while learning to generate watermarked structures with high recovery rates. Extensive experiments are conducted on open-source protein structure prediction models (e.g., ESMFold and MultiFlow) and de novo structure design models (e.g., FrameDiff and FoldFlow) and we demonstrate that our method is effective across all these generative models. Meanwhile, our watermarking framework only exerts a negligible impact on the original protein structure quality and is robust under potential post-processing and adaptive attacks.
In recent years, advances in artificial intelligence (AI) have transformed structural biology, particularly protein structure prediction. Though AI-based methods, such as AlphaFold (AF), often predict single conformations of proteins with high accuracy and confidence, predictions of alternative folds are often inaccurate, low-confidence, or simply not predicted at all. Here, we review three blind spots that alternative conformations reveal about AF-based protein structure prediction. First, proteins that assume conformations distinct from their training-set homologs can be mispredicted. Second, AF overrelies on its training set to predict alternative conformations. Third, degeneracies in pairwise representations can lead to high-confidence predictions inconsistent with experiment. These weaknesses suggest approaches to predict alternative folds more reliably.
Protein structures are important for understanding their functions and interactions. Currently, many protein structure prediction methods are enriching the structure database. Discriminating the origin of structures is crucial for distinguishing between experimentally resolved and computationally predicted structures, evaluating the reliability of prediction methods, and guiding downstream biological studies. Building on works in structure prediction, We developed a structure-sensitive supervised deep learning model, Crystal vs Predicted Evaluator for Protein Structure (CPE-Pro), to represent and discriminate the origin of protein structures. CPE-Pro learns the structural information of proteins and captures inter-structural differences to achieve accurate traceability on four data classes, and is expected to be extended to more. Simultaneously, we utilized Foldseek to encode protein structures into "structure-sequences" and trained a protein Structural Sequence Language Model, SSLM. Preliminary experiments demonstrated that, compared to large-scale protein language models pre-trained on vast amounts of amino acid sequences, the "structure-sequence" enables the language model to learn more informative protein features, enhancing and optimizing structural representations. We have provided the code, model weights, and all related materials on https://github.com/GouWenrui/CPE-Pro-main.git.
Deep learning has deeply influenced protein science, enabling breakthroughs in predicting protein properties, higher-order structures, and molecular interactions. This paper introduces DeepProtein, a comprehensive and user-friendly deep learning library tailored for protein-related tasks. It enables researchers to seamlessly address protein data with cutting-edge deep learning models. To assess model performance, we establish a benchmark evaluating different deep learning architectures across multiple protein-related tasks, including protein function prediction, subcellular localization prediction, protein-protein interaction prediction, and protein structure prediction. Furthermore, we introduce DeepProt-T5, a series of fine-tuned Prot-T5-based models that achieve state-of-the-art performance on four benchmark tasks, while demonstrating competitive results on six of others. Comprehensive documentation and tutorials are available which could ensure accessibility and support reproducibility. Built upon the widely used drug discovery library DeepPurpose, DeepProtein is publicly available at https://github.com/jiaqingxie/DeepProtein.