Protein Structure Prediction Research Articles

Diversity and function of soluble heterodisulfide reductases in methane-metabolizing archaea.

Soluble heterodisulfide reductase subunit A (HdrA) is an ancient protein central to energy metabolism, facilitating the recycling of intermediates in methane metabolism and performing flavin-based electron bifurcation for energy conservation. In this study, we investigated the functional diversity and evolutionary dynamics of HdrA in methane-metabolizing archaea. An analysis of 1,152 HdrA sequences from 624 genomes revealed that HdrA diversified through internal domain modifications, resulting in 28 distinct classes and 4 major types (types I, Ia, II, and III). Functional genes in HdrA gene clusters revealed variations in mid-potential electron donors, including NADH, F420H2, H2, and formate. Two major types of HdrA have not previously been studied in detail. Type II HdrA resulted from a fusion of two different classes of type I HdrA. Particularly, a consistent gene cluster containing type II HdrA, molybdopterin oxidoreductase, and F420 dehydrogenase was identified in anaerobic methane-oxidizing archaea and methanogens. Protein sequence and structural predictions suggested that the molybdopterin oxidoreductase protein had lost its catalytic function, and F420H2 served as the mid-potential electron donor or acceptor for the Hdr protein complex. This gene cluster may expand to include additional type I HdrA and HdrD, potentially supporting two electron bifurcation events to lower electron potential for ferredoxin reduction. Type III HdrA, with an inserted GltD domain compared to type I HdrA, appears to have altered the electron transfer route and may use NADH as its mid-potential electron donor or acceptor. The remarkable functional flexibility of HdrA likely helps methane-metabolizing archaea adapt to diverse anaerobic environments.IMPORTANCEAll methanogenic archaea use heterodisulfide of coenzymes M and B as the terminal electron acceptor. In anaerobic methane- and alkane-oxidizing archaea, the reverse reaction occurs. The cycling of heterodisulfide is vital to the energy conservation of these anaerobic microorganisms. Soluble heterodisulfide reductase is an ancient protein fulfilling this function via flavin-based electron bifurcation or confurcation. Despite being present in the vast majority of methane- and alkane-metabolizing archaea, the diversity and evolution of this key protein have not been investigated. This study reveals substantial domain variation and structural changes in the key bifurcating subunit HdrA in methane- and alkane-metabolizing archaea. The resulting flexibility of HdrA enables the protein complex to vary its interacting subunits and electron carriers based on the organisms' primary metabolism. Our findings shed light on how methane- and alkane-metabolizing archaea thrive in various anaerobic environments, contributing to our broader understanding of carbon cycling and energy conservation.

The Frontier Exploration of Algorithm Innovation and Experimental Verification in Intelligent Protein Design

Intelligent protein design is a frontier topic in the cross field of modern biotechnology and AI. Through the combination of algorithm innovation and experimental verification, it breaks through the limitations of traditional protein design. In this paper, the progress of algorithm innovation in intelligent protein design is summarized, especially the application of advanced algorithms such as deep learning, generative model and reinforcement learning in protein structure prediction, function optimization and interaction analysis. Taking DeepThermoNet, a deep learning algorithm, as an example, the effect of protein mutant designed by DeepThermonet in improving the thermal stability of β -glucosidase was verified by experiments. The results showed that the mutant designed by the algorithm group was significantly better than the mutant designed by the traditional method in melting temperature (Tm) and enzyme activity retention rate. The experimental verification not only proves the effectiveness of the algorithm design, but also optimizes the algorithm model through feedback, forming a closed loop of "algorithm design-experimental verification-model optimization". This paper further discusses the interactive relationship between algorithm innovation and experimental verification, looks forward to the future development direction of intelligent protein design, including interdisciplinary integration, new algorithm development and data resource expansion, and points out the limitations of current research and the key direction of future work. Intelligent protein design is expected to provide new theoretical and technical support for drug research and development, biocatalyst development and biomaterial design, and promote innovation and development in related fields.

An informatics-based analysis platform identifies diverse microbial species with plastic-degrading potential.

An informatics-based analysis platform identifies diverse microbial species with plastic-degrading potential.

Equi-MOI ratio for rapid baculovirus-mediated multiprotein co-expression in insect cells integrating selenomethionine for structural studies.

Equi-MOI ratio for rapid baculovirus-mediated multiprotein co-expression in insect cells integrating selenomethionine for structural studies.

Artificial Intelligence: A New Tool for Structure-Based G Protein-Coupled Receptor Drug Discovery

Understanding protein structures can facilitate the development of therapeutic drugs. Traditionally, protein structures have been determined through experimental approaches such as X-ray crystallography, NMR spectroscopy, and cryo-electron microscopy. While these methods are effective and are considered the gold standard, they are very resource-intensive and time-consuming, ultimately limiting their scalability. However, with recent developments in computational biology and artificial intelligence (AI), the field of protein prediction has been revolutionized. Innovations like AlphaFold and RoseTTAFold enable protein structure predictions to be made directly from amino acid sequences with remarkable speed and accuracy. Despite the enormous enthusiasm associated with these newly developed AI-approaches, their true potential in structure-based drug discovery remains uncertain. In fact, although these algorithms generally predict overall protein structures well, essential details for computational ligand docking, such as the exact location of amino acid side chains within the binding pocket, are not predicted with the necessary accuracy. Additionally, docking methodologies are considered more as a hypothesis generator rather than a precise predictor of ligand–target interactions, and thus, usually identify many false-positive hits among only a few correctly predicted interactions. In this paper, we are reviewing the latest development in this cutting-edge field with emphasis on the GPCR target class to assess the potential role of AI approaches in structure-based drug discovery.

Replicating PET Hydrolytic Activity by Positioning Active Sites with Smaller Synthetic Protein Scaffolds.

Evolutionary constraints significantly limit the diversity of naturally occurring enzymes, thereby reducing the sequence repertoire available for enzyme discovery and engineering. Recent breakthroughs in protein structure prediction and de novo design, powered by artificial intelligence, now enable to create enzymes with desired functions without solely relying on traditional genome mining. Here, a computational strategy is demonstrated for creating new-to-nature polyethylene terephthalate hydrolases (PET hydrolases) by leveraging the known catalytic mechanisms and implementing multiple deep learning algorithms and molecular computations. This strategy includes the extraction of functional motifs from a template enzyme (here leaf-branch compost cutinase, LCC,is used), regeneration of new protein sequences, computational screening, experimental validation, and sequence refinement. PET hydrolytic activity is successfully replicated with designer enzymes that are at least 30% shorter in sequence length than LCC. Among them, RsPETase1 stands out due to its robust expressibility. It exhibits comparable catalytic efficiency (kcat/Km) to LCC and considerable thermostability with a melting temperature of 56°C, despite sharing only 34% sequence similarity with LCC. This work suggests that enzyme diversity can be expanded by recapitulating functional motifs with computationally built protein scaffolds, thus generating opportunities to acquire highly active and robust enzymes that do not exist in nature.

DAWN OF SYNTHETIC BIOLOGY ENGINEERING LIFE AT THE MICROSCOPIC SCALE: A REVIEW

Synthetic biology, converging biology, engineering and computer science, allows the design of new biological systems, promising revolutions in healthcare, agriculture and environmental sustainability. Its core principles—modularity, abstraction hierarchies, orthogonality, predictability, and standardization—enable systematic biological engineering. Modularity breaks complex systems into manageable parts, while abstraction hierarchies organize these parts by complexity. Orthogonality ensures independent function of synthetic components and predictability is achieved through modeling and computation. Standardization promotes reproducibility and collaboration. Mechanistically, synthetic biology manipulates DNA, designs genetic circuits and metabolic pathways and applies physical and computational principles. Techniques like PCR and DNA sequencing construct recombinant DNA. Genetic circuits control gene expression and metabolic engineering optimizes pathways. Integrating Artificial Intelligence (AI) and Machine Learning (ML) accelerates innovation by analyzing data, predicting protein structures and automating experiments, improving drug and therapy development.Synthetic biology can address global challenges like infectious diseases, climate change and food security, in addition to the potential applications in the medical and pharmaceutical sectors. By understanding its principles and using advanced technologies, researchers can realize the field's potential for a better future. Peer Review History: Received 12 December 2024; Reviewed 6 January 2025; Accepted 17 February; Available online 15 March 2025 Academic Editor: Dr. Jennifer Audu-Peter, University of Jos, Nigeria, drambia44@gmail.com

Mining Highly Active Oleate Hydratases by Structure Clustering, Sequence Clustering, and Ancestral Sequence Reconstruction.

Oleate hydratases (Ohys) catalyze the conversion of oleic acid (OA) to 10-(R)-hydroxystearic acid (10-HSA), a compound widely used in the chemical industry. However, the limited activity of Ohys has hindered their broader applications. To address this limitation, we propose a novel strategy for mining highly active Ohys through structure clustering, sequence clustering, and ancestral sequence reconstruction (SSA strategy). Structure clustering via AI-driven protein structure prediction followed by classification enhanced the ability to mine target Ohys. Ancestral enzyme reconstruction was carried out based on mining results from both structure and sequence clustering. This strategy significantly reduces the time and cost of the discovery process. Among the 1304 Ohys screened via SSA, 13 candidates were selected. Seven candidates demonstrated high activity. Ohy 64, identified through structure clustering, exhibited the highest activity. Ancestral enzymes that were reconstructed from structure clustering targets were 3 times more likely to exhibit high catalytic activity than those identified through sequence clustering. Four critical, hydrophobic residues were identified through structure and sequence comparisons between StOhy and targets mined by SSA. Site-directed mutagenesis revealed that these hydrophobic residues conferred varying levels of enzyme activity, confirming that increased hydrophobicity at these positions enhances cofactor FAD binding, thus improving enzyme catalytic efficiency.

Genetic and protein structure prediction analyses identify a rare pathogenic PKD1 variant causing autosomal dominant polycystic kidney disease.

Autosomal dominant polycystic kidney disease (ADPKD) is one of the most common monogenic kidney disorders. The diagnosis of ADPKD requires imaging findings showing multiple kidney cysts or genetic testing, in cases where a family history is unknown. We report a patient diagnosed with ADPKD based on the identification of a rare PKD1 variant. The patient was a 41-year-old female in whom cysts and calcification in the kidneys were incidentally detected. Whole-exome sequencing identified a rare PKD1 variant (NM_001009944.3: c.11557G > A/p.E3853K). Protein structure prediction of the PKD1-PKD2 complex showed that the variant may be pathogenic, leading to the diagnosis of autosomal dominant polycystic kidney disease. A detailed family history revealed that her relatives also had ADPKD, further supporting the diagnosis of ADPKD. Comprehensive genetic analysis and protein structure prediction led to the diagnosis of ADPKD and the identification of rare causative genes. These methods are useful for diagnosing hereditary kidney diseases of unknown etiologies.

Conformational ensembles for protein structure prediction

Acquisition of conformational ensembles for a protein is a challenging task, which is actually involving to the solution for protein folding problem and the study of intrinsically disordered protein. Despite AlphaFold with artificial intelligence acquired unprecedented accuracy to predict structures, its result is limited to a single state of conformation and it cannot provide multiple conformations to display protein intrinsic disorder. To overcome the barrier, a FiveFold approach was developed with a single sequence method. It applied the protein folding shape code (PFSC) uniformly to expose local folds of five amino acid residues, formed the protein folding variation matrix (PFVM) to reveal local folding variations along sequence, obtained a massive number of folding conformations in PFSC strings, and then an ensemble of multiple conformational protein structures is constructed. The P53_HUMAN as a well-known protein and LEF1_HUMAN and Q8GT36_SPIOL as typical disordered proteins are token as the benchmark to evaluate the predicted outcomes. The results demonstrated an effective algorithm and biological meaningful process well to predict protein multiple conformation structures.

Heterologous expression and characterization of xylose-tolerant GH 43 family β-xylosidase/α-L-arabinofuranosidase from Limosilactobacillus fermentum and its application in xylan degradation.

The degradation of hemicellulose, including xylan, is an important industrial process as it provides cheap and sustainable source of economically valuable monosaccharides. β-xylosidases are key enzymes required for complete degradation of xylan and are used in the production of monosaccharides, such as xylose. In this study, we characterized a novel, xylose-tolerant β-xylosidase isolated from Limosilactobacillus fermentum SK152. Sequence analysis and protein structure prediction revealed that the putative β-xylosidase belongs to the glycoside hydrolase (GH) family 43 subfamily 11 and exhibits high homology with other characterised GH43 β-xylosidases from fungal and bacterial sources. The putative β-xylosidase was named LfXyl43. The catalytic residues of LfXyl43, which are highly conserved among GH 43 β-xylosidases, were predicted. To fully characterise LfXyl43, the gene encoding it was heterologously expressed in Escherichia coli. Biochemical characterisation revealed that the recombinant LfXyl43 (rLfXyl43) was active against artificial and natural substrates containing β-1,4-xylanopyranosyl residues, such as p-nitrophenyl-β-D-xylopyranoside (pNPX) and oNPX. Moreover, it demonstrated weak α-L-arabinofuranosidase activity. The optimal activity of rLfXyl43 was obtained at pH 7.0at 35°C. rLfXyl43 could degrade xylo-oligosaccharides, such as xylobiose, xylotriose, and xylotetraose, and showed hydrolysing activity towards beechwood xylan. Moreover, rLfXyl43 demonstrated synergy with a commercial xylanase in degrading rye and wheat arabinoxylan. The activity of rLfXyl43 was not affected by the addition of metal ions, chemical reagents, or high concentrations of NaCl. Notably, rLfXyl43 exhibited tolerance to high xylose concentrations, with a K i value of 100.1, comparable to that of other xylose-tolerant GH 43 β-xylosidases. To our knowledge, this is the first β-xylosidase identified from a lactic acid bacterium with high tolerance to salt and xylose. Overall, rLfXyl43 exhibits great potential as a novel β-xylosidase for use in the degradation of lignocellulosic material, especially xylan hemicellulose. Its high activity against xylo-oligosaccharides, mild catalytic conditions, and tolerance to high xylose concentrations makes it a suitable enzyme for industrial applications.

Protein-Peptide Docking with ESMFold Language Model.

Designing peptide therapeutics requires precise peptide docking, which remains a challenge. We assessed the ESMFold language model, originally designed for protein structure prediction, for its effectiveness in protein-peptide docking. Various docking strategies, including polyglycine linkers and sampling-enhancing modifications, were explored. The number of acceptable-quality models among top-ranking results is comparable to traditional methods and generally lower than AlphaFold-Multimer or Alphafold 3, though ESMFold surpasses it in some cases. The combination of result quality and computational efficiency underscores ESMFold's potential value as a component in a consensus approach for high-throughput peptide design.

Molecular Interactions of Mycobacterial Transporter for Novel Antimicrobial Strategies

Efflux mechanisms for extruding antimicrobials, mediated by multidrug transporters, are key contributors to multidrug resistance in mycobacteria. The current study focused on molecular interaction analysis of Mycobacterium tuberculosis multidrug transporter implicated in multidrug and antimicrobial resistance. We screened a library of efflux transporter inhibitors against the protein structure to identify a lead compound that can potentially inhibit the transporter significantly. The efflux transporter sequence was modeled based on crystallized templates using protein structure prediction and molecular docking. The analysis deduced molecular interactions and critical binding residues that can be targeted as novel biotherapeutics strategies against multidrug transporters of mycobacteria. This study paves the way for targeting multidrug and antimicrobial resistance in the mycobacteria, offering hope for developing effective treatments.

Revisiting the Plasmodium falciparum druggable genome using predicted structures and data mining

Identification of novel drug targets is a key component of modern drug discovery. While antimalarial targets are often identified through the mechanism of action studies on phenotypically derived inhibitors, this method tends to be time- and resource-consuming. The discoverable target space is also constrained by existing compound libraries and phenotypic assay conditions. Leveraging recent advances in protein structure prediction, we systematically assessed the Plasmodium falciparum genome and identified 867 candidate protein targets with evidence of small-molecule binding and blood-stage essentiality. Of these, 540 proteins showed strong essentiality evidence and lack inhibitors that have progressed to clinical trials. Expert review and rubric-based scoring of this subset based on additional criteria such as selectivity, structural information, and assay developability yielded 27 high-priority antimalarial target candidates. This study also provides a genome-wide data resource for P. falciparum and implements a generalizable framework for systematically evaluating and prioritizing novel pathogenic disease targets.

Regularly updated benchmark sets for statistically correct evaluations of AlphaFold applications.

AlphaFold2 changed structural biology by providing high-quality structure predictions for all possible proteins. Since its inception, a plethora of applications were built on AlphaFold2, expediting discoveries in virtually all areas related to protein science. In many cases, however, optimism seems to have made scientists forget about data leakage, a serious issue that needs to be addressed when evaluating machine learning methods. Here we provide a rigorous benchmark set that can be used in a broad range of applications built around AlphaFold2/3.

Leveraging Sequence Purification for Accurate Prediction of Multiple Conformational States with AlphaFold2.

AlphaFold2 (AF2) has transformed protein structure prediction by harnessing co-evolutionary constraints embedded in multiple sequence alignments (MSAs). MSAs not only encode static structural information, but also hold critical details about protein dynamics, which underpin biological functions. However, these subtle co-evolutionary signatures, which dictate conformational state preferences, are often obscured by noise within MSA data and thus remain challenging to decipher. Here, we introduce AF-ClaSeq, a systematic framework that isolates these co-evolutionary signals through sequence purification and iterative enrichment. By extracting sequence subsets that preferentially encode distinct structural states, AF-ClaSeq enables high-confidence predictions of alternative conformations. Our findings reveal that the successful sampling of alternative states depends not on MSA depth but on sequence purity. Intriguingly, purified sequences encoding specific structural states are distributed across phylogenetic clades and superfamilies, rather than confined to specific lineages. Expanding upon AF2's transformative capabilities, AF-ClaSeq provides a powerful approach for uncovering hidden structural plasticity, advancing allosteric protein and drug design, and facilitating dynamics-based protein function annotation.

Characterization of an α-ketoglutarate-dependent oxygenase involved in converting 2-(2-phenylethyl)chromones into 2-styrylchromones in agarwood.

2-Phenylethylchromones (PECs) and 2-styrylchromones (SCs) are the primary components responsible for the delightful fragrance and bioactivity of agarwood, a highly valuable aromatic resinous heartwood. PECs are derived from a common precursor with a diarylpentanoid skeleton (C6-C5-C6). However, the biosynthesis of SCs remains unclear. In this study, based on the successful conversion of the PEC skeleton, rather than a dehydrogenated diarylpentanoid, into SCs by Aquilaria sinensis suspension cells, we demonstrated that double bond formation of the styryl group in SCs occurs after the creation of the PEC skeleton, not before this step from a dehydrogenated diarylpentanoid precursor. Through transcriptomic data mining, transient expression in Nicotiana benthamiana and A. sinensis suspension cells, we identified a new 2-oxoglutarate-dependent oxygenase (As2OG1) that plays a crucial role in the conversion of PECs into SCs. Further protein structure prediction and mutagenesis studies, combined with probing of the catalytic potential of As2OG1 using chemically synthesized hydroxylated intermediates, suggested that As2OG1 possibly uses diradical or carbocation intermediates, rather than hydroxylated intermediates, to install double bonds in SCs. The results not only provide insights into the molecular mechanism of agarwood formation but also facilitate the overproduction of pharmaceutically important SCs using metabolic engineering approaches.

Proteome-wide assessment of differential missense variant clustering in neurodevelopmental disorders and cancer.

Proteome-wide assessment of differential missense variant clustering in neurodevelopmental disorders and cancer.

NucleoSeeker-precision filtering of RNA databases to curate high-quality datasets.

The structural prediction of biomolecules via computational methods complements the often involved wet-lab experiments. Unlike protein structure prediction, RNA structure prediction remains a significant challenge in bioinformatics, primarily due to the scarcity of annotated RNA structure data and its varying quality. Many methods have used this limited data to train deep learning models but redundancy, data leakage and bad data quality hampers their performance. In this work, we present NucleoSeeker, a tool designed to curate high-quality, tailored datasets from the Protein Data Bank (PDB) database. It is a unified framework that combines multiple tools and streamlines an otherwise complicated process of data curation. It offers multiple filters at structure, sequence, and annotation levels, giving researchers full control over data curation. Further, we present several use cases. In particular, we demonstrate how NucleoSeeker allows the creation of a nonredundant RNA structure dataset to assess AlphaFold3's performance for RNA structure prediction. This demonstrates NucleoSeeker's effectiveness in curating valuable nonredundant tailored datasets to both train novel and judge existing methods. NucleoSeeker is very easy to use, highly flexible, and can significantly increase the quality of RNA structure datasets.

Systems biology of Haemonchus contortus - Advancing biotechnology for parasitic nematode control.

Systems biology of Haemonchus contortus - Advancing biotechnology for parasitic nematode control.