Single-sequence protein structure prediction by integrating protein language models.

Protein structure prediction has been greatly improved by deep learning in the past few years. However, the most successful methods rely on multiple sequence alignment (MSA) of the sequence homologs of the protein under prediction. In nature, a protein folds in the absence of its sequence homologs and thus, a MSA-free structure prediction method is desired. Here, we develop a single-sequence-based protein structure prediction method RaptorX-Single by integrating several protein language models and a structure generation module and then study its advantage over MSA-based methods. Our experimental results indicate that in addition to running much faster than MSA-based methods such as AlphaFold2, RaptorX-Single outperforms AlphaFold2 and other MSA-free methods in predicting the structure of antibodies (after fine-tuning on antibody data), proteins of very few sequence homologs, and single mutation effects. By comparing different protein language models, our results show that not only the scale but also the training data of protein language models will impact the performance. RaptorX-Single also compares favorably to MSA-based AlphaFold2 when the protein under prediction has a large number of sequence homologs.

ABAG-docking benchmark: a non-redundant structure benchmark dataset for antibody-antigen computational docking.

Accurate prediction of antibody-antigen complex structures is pivotal in drug discovery, vaccine design and disease treatment and can facilitate the development of more effective therapies and diagnostics. In this work, we first review the antibody-antigen docking (ABAG-docking) datasets. Then, we present the creation and characterization of a comprehensive benchmark dataset of antibody-antigen complexes. We categorize the dataset based on docking difficulty, interface properties and structural characteristics, to provide a diverse set of cases for rigorous evaluation. Compared with Docking Benchmark 5.5, we have added 112 cases, including 14 single-domain antibody (sdAb) cases and 98 monoclonal antibody (mAb) cases, and also increased the proportion of Difficult cases. Our dataset contains diverse cases, including human/humanized antibodies, sdAbs, rodent antibodies and other types, opening the door to better algorithm development. Furthermore, we provide details on the process of building the benchmark dataset and introduce a pipeline for periodic updates to keep it up to date. We also utilize multiple complex prediction methods including ZDOCK, ClusPro, HDOCK and AlphaFold-Multimer for testing and analyzing this dataset. This benchmark serves as a valuable resource for evaluating and advancing docking computational methods in the analysis of antibody-antigen interaction, enabling researchers to develop more accurate and effective tools for predicting and designing antibody-antigen complexes. The non-redundant ABAG-docking structure benchmark dataset is available at https://github.com/Zhaonan99/Antibody-antigen-complex-structure-benchmark-dataset.

An end-to-end deep learning method for protein side-chain packing and inverse folding.

Protein side-chain packing (PSCP), the task of determining amino acid side-chain conformations given only backbone atom positions, has important applications to protein structure prediction, refinement, and design. Many methods have been proposed to tackle this problem, but their speed or accuracy is still unsatisfactory. To address this, we present AttnPacker, a deep learning (DL) method for directly predicting protein side-chain coordinates. Unlike existing methods, AttnPacker directly incorporates backbone 3D geometry to simultaneously compute all side-chain coordinates without delegating to a discrete rotamer library or performing expensive conformational search and sampling steps. This enables a significant increase in computational efficiency, decreasing inference time by over 100× compared to the DL-based method DLPacker and physics-based RosettaPacker. Tested on the CASP13 and CASP14 native and nonnative protein backbones, AttnPacker computes physically realistic side-chain conformations, reducing steric clashes and improving both rmsd and dihedral accuracy compared to state-of-the-art methods SCWRL4, FASPR, RosettaPacker, and DLPacker. Different from traditional PSCP approaches, AttnPacker can also codesign sequences and side chains, producing designs with subnative Rosetta energy and high in silico consistency.

Improved the heterodimer protein complex prediction with protein language models.

AlphaFold-Multimer has greatly improved the protein complex structure prediction, but its accuracy also depends on the quality of the multiple sequence alignment (MSA) formed by the interacting homologs (i.e. interologs) of the complex under prediction. Here we propose a novel method, ESMPair, that can identify interologs of a complex using protein language models. We show that ESMPair can generate better interologs than the default MSA generation method in AlphaFold-Multimer. Our method results in better complex structure prediction than AlphaFold-Multimer by a large margin (+10.7% in terms of the Top-5 best DockQ), especially when the predicted complex structures have low confidence. We further show that by combining several MSA generation methods, we may yield even better complex structure prediction accuracy than Alphafold-Multimer (+22% in terms of the Top-5 best DockQ). By systematically analyzing the impact factors of our algorithm we find that the diversity of MSA of interologs significantly affects the prediction accuracy. Moreover, we show that ESMPair performs particularly well on complexes in eucaryotes.

Accurate protein function prediction via graph attention networks with predicted structure information.

Experimental protein function annotation does not scale with the fast-growing sequence databases. Only a tiny fraction (<0.1%) of protein sequences has experimentally determined functional annotations. Computational methods may predict protein function very quickly, but their accuracy is not very satisfactory. Based upon recent breakthroughs in protein structure prediction and protein language models, we develop GAT-GO, a graph attention network (GAT) method that may substantially improve protein function prediction by leveraging predicted structure information and protein sequence embedding. Our experimental results show that GAT-GO greatly outperforms the latest sequence- and structure-based deep learning methods. On the PDB-mmseqs testset where the train and test proteins share <15% sequence identity, our GAT-GO yields Fmax (maximum F-score) 0.508, 0.416, 0.501, and area under the precision-recall curve (AUPRC) 0.427, 0.253, 0.411 for the MFO, BPO, CCO ontology domains, respectively, much better than the homology-based method BLAST (Fmax 0.117, 0.121, 0.207 and AUPRC 0.120, 0.120, 0.163) that does not use any structure information. On the PDB-cdhit testset where the training and test proteins are more similar, although using predicted structure information, our GAT-GO obtains Fmax 0.637, 0.501, 0.542 for the MFO, BPO, CCO ontology domains, respectively, and AUPRC 0.662, 0.384, 0.481, significantly exceeding the just-published method DeepFRI that uses experimental structures, which has Fmax 0.542, 0.425, 0.424 and AUPRC only 0.313, 0.159, 0.193.

A tale of solving two computational challenges in protein science: neoantigen prediction and protein structure prediction.

In this article, we review two challenging computational questions in protein science: neoantigen prediction and protein structure prediction. Both topics have seen significant leaps forward by deep learning within the past five years, which immediately unlocked new developments of drugs and immunotherapies. We show that deep learning models offer unique advantages, such as representation learning and multi-layer architecture, which make them an ideal choice to leverage a huge amount of protein sequence and structure data to address those two problems. We also discuss the impact and future possibilities enabled by those two applications, especially how the data-driven approach by deep learning shall accelerate the progress towards personalized biomedicine.

Improving protein structure prediction using templates and sequence embedding.

MOTIVATION: Protein structure prediction has been greatly improved by deep learning, but the contribution of different information is yet to be fully understood. This article studies the impacts of two kinds of information for structure prediction: template and multiple sequence alignment (MSA) embedding. Templates have been used by some methods before, such as AlphaFold2, RoseTTAFold and RaptorX. AlphaFold2 and RosetTTAFold only used templates detected by HHsearch, which may not perform very well on some targets. In addition, sequence embedding generated by pre-trained protein language models has not been fully explored for structure prediction. In this article, we study the impact of templates (including the number of templates, the template quality and how the templates are generated) on protein structure prediction accuracy, especially when the templates are detected by methods other than HHsearch. We also study the impact of sequence embedding (generated by MSATransformer and ESM-1b) on structure prediction. RESULTS: We have implemented a deep learning method for protein structure prediction that may take templates and MSA embedding as extra inputs. We study the contribution of templates and MSA embedding to structure prediction accuracy. Our experimental results show that templates can improve structure prediction on 71 of 110 CASP13 (13th Critical Assessment of Structure Prediction) targets and 47 of 91 CASP14 targets, and templates are particularly useful for targets with similar templates. MSA embedding can improve structure prediction on 63 of 91 CASP14 (14th Critical Assessment of Structure Prediction) targets and 87 of 183 CAMEO targets and is particularly useful for proteins with shallow MSAs. When both templates and MSA embedding are used, our method can predict correct folds (TMscore > 0.5) for 16 of 23 CASP14 FM targets and 14 of 18 Continuous Automated Model Evaluation (CAMEO) targets, outperforming RoseTTAFold by 5% and 7%, respectively. AVAILABILITY AND IMPLEMENTATION: Available at https://github.com/xluo233/RaptorXFold. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Dual interface trapezium liquid prism with beam steering function.

In this paper, a dual interface trapezium liquid prism with beam steering function is implemented and analyzed. The electrowetting-on-dielectric method is used to perform the desired beam steering function without mechanical moving parts. This work examines deflection angles at different applied voltages to determine the beam steering range. The deflection angle can be experimentally measured from 0° to 3.43°. The proposed liquid prism can be applied in the field of optical manipulation, solar collecting system and so on.

Deep graph learning of inter-protein contacts.

MOTIVATION: Inter-protein (interfacial) contact prediction is very useful for in silico structural characterization of protein-protein interactions. Although deep learning has been applied to this problem, its accuracy is not as good as intra-protein contact prediction. RESULTS: We propose a new deep learning method GLINTER (Graph Learning of INTER-protein contacts) for interfacial contact prediction of dimers, leveraging a rotational invariant representation of protein tertiary structures and a pretrained language model of multiple sequence alignments. Tested on the 13th and 14th CASP-CAPRI datasets, the average top L/10 precision achieved by GLINTER is 54% on the homodimers and 52% on all the dimers, much higher than 30% obtained by the latest deep learning method DeepHomo on the homodimers and 15% obtained by BIPSPI on all the dimers. Our experiments show that GLINTER-predicted contacts help improve selection of docking decoys. AVAILABILITY AND IMPLEMENTATION: The software is available at https://github.com/zw2x/glinter. The datasets are available at https://github.com/zw2x/glinter/data. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Three-phase electrowetting liquid lens with deformable liquid iris.

Inspired by the arrangement of iris and crystalline lens in human eyes, we propose a three-phase electrowetting liquid lens with a deformable liquid iris (TELL-DLI). The proposed electrowetting liquid lens has three-phase fluid: air, conductive liquid, and dyed insulating liquid. The insulating liquid is distributed on the inner wall of the chamber in a ring shape. By applying voltage, the contact angle is changed, so that the dyed insulating liquid contracts towards the center, which is similar to the contraction of iris and the function of crystalline lens muscle in human eyes. The variation range of focal length is from -451.9 mm to -107.9 mm. The variation range of the aperture is from 4.89 mm to 0.6 mm. Under the step voltage of 200 V, the TELL-DLI can be switched between the maximum aperture state and the zero aperture state, and the switching time is ∼150/200 ms. Because of the discrete electrodes, TELL-DLI can regionally control the shape and position of the iris, and switch between circle, ellipse, sector, and strip. The TELL-DLI has a wide application prospect in imaging systems, such as microscopic imaging system, and has the potential to be applied in the field of complex beam navigation.

Annotating functional effects of non-coding variants in neuropsychiatric cell types by deep transfer learning.

Genomewide association studies (GWAS) have identified a large number of loci associated with neuropsychiatric traits, however, understanding the molecular mechanisms underlying these loci remains difficult. To help prioritize causal variants and interpret their functions, computational methods have been developed to predict regulatory effects of non-coding variants. An emerging approach to variant annotation is deep learning models that predict regulatory functions from DNA sequences alone. While such models have been trained on large publicly available dataset such as ENCODE, neuropsychiatric trait-related cell types are under-represented in these datasets, thus there is an urgent need of better tools and resources to annotate variant functions in such cellular contexts. To fill this gap, we collected a large collection of neurodevelopment-related cell/tissue types, and trained deep Convolutional Neural Networks (ResNet) using such data. Furthermore, our model, called MetaChrom, borrows information from public epigenomic consortium to improve the accuracy via transfer learning. We show that MetaChrom is substantially better in predicting experimentally determined chromatin accessibility variants than popular variant annotation tools such as CADD and delta-SVM. By combining GWAS data with MetaChrom predictions, we prioritized 31 SNPs for Schizophrenia, suggesting potential risk genes and the biological contexts where they act. In summary, MetaChrom provides functional annotations of any DNA variants in the neuro-development context and the general method of MetaChrom can also be extended to other disease-related cell or tissue types.

The epidemiology of varicella and effectiveness of varicella vaccine in Ganyu, China: a long-term community surveillance study.

BACKGROUND: The real-world data of long-term protection under moderate vaccination coverage is limited. This study aimed to evaluate varicella epidemiology and the long-term effectiveness under moderate coverage levels in Ganyu District, Lianyungang City, Jiangsu Province. METHODS: This was a population-based, retrospective birth cohort study based on the immunization information system (IIS) and the National Notifiable Disease Surveillance System (NNDSS) in Ganyu District. Varicella cases reported from 2009 to 2020 were included to describe the epidemiology of varicella, and eleven-year consecutive birth cohorts (2008-2018) were included to estimate the vaccine effectiveness (VE) of varicella by Cox regression analysis. RESULTS: A total of 155,232 native children and 3,251 varicella cases were included. The vaccination coverage was moderate with 37.1%, correspondingly, the annual incidence of varicella infection increased 4.4-fold from 2009 to 2020. A shift of the varicella cases to older age groups was observed, with the peak proportion of cases shifting from 5-6 year-old to 7-8 year-old. The adjusted effectiveness of one dose of vaccine waned over time, and the adjusted VE decreased from 72.9% to 41.8% in the one-dose group. CONCLUSIONS: The insufficient vaccination coverage (37.1%) may have contributed in part to the rising annual incidence of varicella infection, and a shift of varicella cases to older age groups occurred. The effectiveness of one dose of varicella vaccine was moderate and waned over time. It is urgent to increase varicella vaccine coverage to 80% to reduce the incidence of varicella and prevent any potential shift in the age at infection in China.

Improved protein model quality assessment by integrating sequential and pairwise features using deep learning.

MOTIVATION: Accurately estimating protein model quality in the absence of experimental structure is not only important for model evaluation and selection but also useful for model refinement. Progress has been steadily made by introducing new features and algorithms (especially deep neural networks), but the accuracy of quality assessment (QA) is still not very satisfactory, especially local QA on hard protein targets. RESULTS: We propose a new single-model-based QA method ResNetQA for both local and global quality assessment. Our method predicts model quality by integrating sequential and pairwise features using a deep neural network composed of both 1D and 2D convolutional residual neural networks (ResNet). The 2D ResNet module extracts useful information from pairwise features such as model-derived distance maps, co-evolution information, and predicted distance potential from sequences. The 1D ResNet is used to predict local (global) model quality from sequential features and pooled pairwise information generated by 2D ResNet. Tested on the CASP12 and CASP13 datasets, our experimental results show that our method greatly outperforms existing state-of-the-art methods. Our ablation studies indicate that the 2D ResNet module and pairwise features play an important role in improving model quality assessment. AVAILABILITY AND IMPLEMENTATION: https://github.com/AndersJing/ResNetQA. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Study of real-valued distance prediction for protein structure prediction with deep learning.

MOTIVATION: Inter-residue distance prediction by convolutional residual neural network (deep ResNet) has greatly advanced protein structure prediction. Currently, the most successful structure prediction methods predict distance by discretizing it into dozens of bins. Here, we study how well real-valued distance can be predicted and how useful it is for 3D structure modeling by comparing it with discrete-valued prediction based upon the same deep ResNet. RESULTS: Different from the recent methods that predict only a single real value for the distance of an atom pair, we predict both the mean and standard deviation of a distance and then fold a protein by the predicted mean and deviation. Our findings include: (i) tested on the CASP13 FM (free-modeling) targets, our real-valued distance prediction obtains 81% precision on top L/5 long-range contact prediction, much better than the best CASP13 results (70%); (ii) our real-valued prediction can predict correct folds for the same number of CASP13 FM targets as the best CASP13 group, despite generating only 20 decoys for each target; (iii) our method greatly outperforms a very new real-valued prediction method DeepDist in both contact prediction and 3D structure modeling and (iv) when the same deep ResNet is used, our real-valued distance prediction has 1-6% higher contact and distance accuracy than our own discrete-valued prediction, but less accurate 3D structure models. AVAILABILITY AND IMPLEMENTATION: https://github.com/j3xugit/RaptorX-3DModeling. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

ORFLine: a bioinformatic pipeline to prioritize small open reading frames identifies candidate secreted small proteins from lymphocytes.

MOTIVATION: The annotation of small open reading frames (smORFs) of <100 codons (<300 nucleotides) is challenging due to the large number of such sequences in the genome. RESULTS: In this study, we developed a computational pipeline, which we have named ORFLine, that stringently identifies smORFs and classifies them according to their position within transcripts. We identified a total of 5744 unique smORFs in datasets from mouse B and T lymphocytes and systematically characterized them using ORFLine. We further searched smORFs for the presence of a signal peptide, which predicted known secreted chemokines as well as novel micropeptides. Four novel micropeptides show evidence of secretion and are therefore candidate mediators of immunoregulatory functions. AVAILABILITY AND IMPLEMENTATION: Freely available on the web at https://github.com/boboppie/ORFLine. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Deep template-based protein structure prediction.

MOTIVATION: Protein structure prediction has been greatly improved by deep learning, but most efforts are devoted to template-free modeling. But very few deep learning methods are developed for TBM (template-based modeling), a popular technique for protein structure prediction. TBM has been studied extensively in the past, but its accuracy is not satisfactory when highly similar templates are not available. RESULTS: This paper presents a new method NDThreader (New Deep-learning Threader) to address the challenges of TBM. NDThreader first employs DRNF (deep convolutional residual neural fields), which is an integration of deep ResNet (convolutional residue neural networks) and CRF (conditional random fields), to align a query protein to templates without using any distance information. Then NDThreader uses ADMM (alternating direction method of multipliers) and DRNF to further improve sequence-template alignments by making use of predicted distance potential. Finally, NDThreader builds 3D models from a sequence-template alignment by feeding it and sequence coevolution information into a deep ResNet to predict inter-atom distance distribution, which is then fed into PyRosetta for 3D model construction. Our experimental results show that NDThreader greatly outperforms existing methods such as CNFpred, HHpred, DeepThreader and CEthreader. NDThreader was blindly tested in CASP14 as a part of RaptorX server, which obtained the best average GDT score among all CASP14 servers on the 58 TBM targets.

Distance-based protein folding powered by deep learning.

Direct coupling analysis (DCA) for protein folding has made very good progress, but it is not effective for proteins that lack many sequence homologs, even coupled with time-consuming conformation sampling with fragments. We show that we can accurately predict interresidue distance distribution of a protein by deep learning, even for proteins with ∼60 sequence homologs. Using only the geometric constraints given by the resulting distance matrix we may construct 3D models without involving extensive conformation sampling. Our method successfully folded 21 of the 37 CASP12 hard targets with a median family size of 58 effective sequence homologs within 4 h on a Linux computer of 20 central processing units. In contrast, DCA-predicted contacts cannot be used to fold any of these hard targets in the absence of extensive conformation sampling, and the best CASP12 group folded only 11 of them by integrating DCA-predicted contacts into fragment-based conformation sampling. Rigorous experimental validation in CASP13 shows that our distance-based folding server successfully folded 17 of 32 hard targets (with a median family size of 36 sequence homologs) and obtained 70% precision on the top L/5 long-range predicted contacts. The latest experimental validation in CAMEO shows that our server predicted correct folds for 2 membrane proteins while all of the other servers failed. These results demonstrate that it is now feasible to predict correct fold for many more proteins lack of similar structures in the Protein Data Bank even on a personal computer.

Pan-genomic analysis reveals that the evolution of Dietzia species depends on their living habitats.

The bacterial genus Dietzia is widely distributed in various environments. The genomes of 26 diverse strains of Dietzia, including almost all the type strains, were analysed in this study. This analysis revealed a lipid metabolism gene richness, which could explain the ability of Dietzia to live in oil related environments. The pan-genome consists of 83,976 genes assigned into 10,327 gene families, 792 of which are shared by all the genomes of Dietzia. Mathematical extrapolation of the data suggests that the Dietzia pan-genome is open. Both gene duplication and gene loss contributed to the open pan-genome, while horizontal gene transfer was limited. Dietzia strains primarily gained their diverse metabolic capacity through more ancient gene duplications. Phylogenetic analysis of Dietzia isolated from aquatic and terrestrial environments showed two distinct clades from the same ancestor. The genome sizes of Dietzia strains from aquatic environments were significantly larger than those from terrestrial environments, which was mainly due to the occurrence of more gene loss events during the evolutionary progress of the strains from terrestrial environments. The evolutionary history of Dietzia was tightly coupled to environmental conditions, and iron concentrations should be one of the key factors shaping the genomes of the Dietzia lineages.

High stability liquid lens with optical path modulation function.

In this paper, a high stability liquid lens with optical path modulation function is designed and fabricated. The liquid lens has an outer chamber and an inner chamber, and the inner chamber has a structure with three annular anchoring layers. This structure can limit the sliding of the three-phase contact line under electrowetting effect and anchor the position of contact angle with a limited distance. The feasibility of this structure is verified by simulation and practice. The zoom imaging, contact angle, focal length and response time of the liquid lens are analyzed. The structure with three annular anchoring layers provides six anchored precision optical path modulation gears, and the optical path difference can be changed by mechanical hydraulic control, up to 1.17 mm. Widespread applications of the proposed liquid lens are foreseeable such as microscopic imaging and a telescope system, etc.

PredMP: a web server for de novo prediction and visualization of membrane proteins.

MOTIVATION: PredMP is the first web service, to our knowledge, that aims at de novo prediction of the membrane protein (MP) 3D structure followed by the embedding of the MP into the lipid bilayer for visualization. Our approach is based on a high-throughput Deep Transfer Learning (DTL) method that first predicts MP contacts by learning from non-MPs and then predicts the 3D model of the MP using the predicted contacts as distance restraints. This algorithm is derived from our previous Deep Learning (DL) method originally developed for soluble protein contact prediction, which has been officially ranked No. 1 in CASP12. The DTL framework in our approach overcomes the challenge that there are only a limited number of solved MP structures for training the deep learning model. There are three modules in the PredMP server: (i) The DTL framework followed by the contact-assisted folding protocol has already been implemented in RaptorX-Contact, which serves as the key module for 3D model generation; (ii) The 1D annotation module, implemented in RaptorX-Property, is used to predict the secondary structure and disordered regions; and (iii) the visualization module to display the predicted MPs embedded in the lipid bilayer guided by the predicted transmembrane topology. RESULTS: Tested on 510 non-redundant MPs, our server predicts correct folds for ∼290 MPs, which significantly outperforms existing methods. Tested on a blind and live benchmark CAMEO from September 2016 to January 2018, PredMP can successfully model all 10 MPs belonging to the hard category. AVAILABILITY AND IMPLEMENTATION: PredMP is freely accessed on the web at http://www.predmp.com. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.