MEDUSA: Prediction of Protein Flexibility from Sequence

Persistent Sheaf Laplacian Analysis of Protein Flexibility

Protein flexibility, measured by the B-factor or Debye-Waller factor, is essential for protein functions such as structural support, enzyme activity, cellular communication, and molecular transport. Theoretical analysis and prediction of protein flexibility are crucial for protein design, engineering, and drug discovery. In this work, we introduce the persistent sheaf Laplacian (PSL), an effective tool in topological data analysis, to model and analyze protein flexibility. By representing the local topology and geometry of protein atoms through the multiscale harmonic and nonharmonic spectra of PSLs, the proposed model effectively captures protein flexibility and provides accurate, robust predictions of protein B-factors. Our PSL model demonstrates an increase in accuracy of 32% compared to the classical Gaussian network model (GNM) in predicting B-factors for a data set of 364 proteins. Additionally, we construct a blind machine learning prediction method utilizing global and local protein features. Extensive computations and comparisons validate the effectiveness of the proposed PSL model for B-factor predictions.

Strategies for the design of biomimetic cell-penetrating peptides using AI-driven in silico tools for drug delivery

Cell-penetrating peptides (CPP) have gained rapid attention over the last 25 years; this is attributed to their versatility, customisation, and 'Trojan horse' delivery that evades the immune system. However, the current CPP rational design process is limited, as it requires several rounds of peptide synthesis, prediction and wet-lab validation, which is expensive, time-consuming and requires extensive knowledge in peptide chemistry. Artificial intelligence (AI) has emerged as a promising alternative which can augment the design process, for example by determining physiochemical characteristics, secondary structure, solvent accessibility, disorder and flexibility, as well as predicting in vivo behaviour such as toxicity and peptidase degradation. Other more recent tools utilise supervised machine learning (ML) to predict the penetrative ability of an amino acid sequence. The use of AI in the CPP design process has the potential to reduce development costs and increase the chances of success with respect to delivery. This review provides a survey of in silico tools and AI platforms which can be utilised in the design process, and the key features that should be taken into consideration when designing next generation CPPs.

Virtual Screening and Molecular Dynamics of Cytokine-Drug Complexes for Atherosclerosis Therapy

Cardiovascular disease remains the leading global cause of mortality, largely driven by atherosclerosis, a chronic inflammatory condition characterized by lipid accumulation and immune-cell infiltration in arterial walls. Macrophages play a central role by forming foam cells and secreting pro-atherogenic cytokines, such as TNF-α, IFN-γ, and IL-1β, which destabilize atherosclerotic plaques, expanding the lipid core and increasing the risk of thrombosis and ischemia. Despite the significant health burden of subclinical atherosclerosis, few targeted therapies exist. Current treatments, including monoclonal antibodies, are limited by high costs and immunosuppressive side effects, underscoring the urgent need for alternative therapeutic strategies. In this study, we employed in silico drug repositioning to identify multitarget inhibitors against TNF-α, IFN-γ, and IL-1β, leveraging a virtual screening of 2750 FDA-approved drugs followed by molecular dynamics simulations to assess the stability of selected cytokine-ligand complexes. This computational approach provides structural insights into potential inhibitors. Additionally, we highlight nutraceutical options, such as fatty acids (oleic, linoleic and eicosapentaenoic acid), which exhibited strong and stable interactions with key cytokine targets. Our study suggests that these bioactive compounds could serve as effective new therapeutic approaches for atherosclerosis.

Multiscale Differential Geometry Learning for Protein Flexibility Analysis

Protein structural fluctuations, measured by Debye-Waller factors or B-factors, are known to be closely associated with protein flexibility and function. Theoretical approaches have also been developed to predict B-factor values, which reflect protein flexibility. Previous models have made significant strides in analyzing B-factors by fitting experimental data. In this study, we propose a novel approach for B-factor prediction using differential geometry theory, based on the assumption that the intrinsic properties of proteins reside on a family of low-dimensional manifolds embedded within the high-dimensional space of protein structures. By analyzing the mean and Gaussian curvatures of a set of low-dimensional manifolds defined by kernel functions, we develop effective and robust multiscale differential geometry (mDG) models. Our mDG model demonstrates a 27% increase in accuracy compared to the classical Gaussian network model (GNM) in predicting B-factors for a dataset of 364 proteins. Additionally, by incorporating both global and local protein features, we construct a highly effective machine-learning model for the blind prediction of B-factors. Extensive least-squares approximations and machine learning-based blind predictions validate the effectiveness of the mDG modeling approach for B-factor predictions.

The 'very moment' when UDG recognizes a flipped-out uracil base in dsDNA

Uracil-DNA glycosylase (UDG) is the first enzyme in the base-excision repair (BER) pathway, acting on uracil bases in DNA. How UDG finds its targets has not been conclusively resolved yet. Based on available structural and other experimental evidence, two possible pathways are under discussion. In one, the action of UDG on the DNA bases is believed to follow a 'pinch-push-pull' model, in which UDG generates the base-flip in an active manner. A second scenario is based on the exploitation of bases flipping out thermally from the DNA. Recent molecular dynamics (MD) studies of DNA in trinucleosome arrays have shown that base-flipping can be readily induced by the action of mechanical forces on DNA alone. This alternative mechanism could possibly enhance the probability for the second scenario of UDG-uracil interaction via the formation of a recognition complex of UDG with flipped-out base. In this work, we describe DNA structures with flipped-out uracil bases generated by MD simulations which we then subject to docking simulations with the UDG enzyme. Our results for the UDG-uracil recognition complex support the view that base-flipping induced by DNA mechanics can be a relevant mechanism of uracil base recognition by the UDG glycosylase in chromatin.

Uncovering the structural stability of Magnaporthe oryzae effectors: a secretome-wide in silico analysis

Rice blast, caused by the ascomycete fungus Magnaporthe oryzae, is a deadly disease and a major threat to global food security. The pathogen secretes small proteinaceous effectors, virulence factors, inside the host to manipulate and perturb the host immune system, allowing the pathogen to colonize and establish a successful infection. While the molecular functions of several effectors are characterized, very little is known about the structural stability of these effectors. We analyzed a total of 554 small secretory proteins (SSPs) from the M. oryzae secretome to decipher key features of intrinsic disorder (ID) and the structural dynamics of the selected putative effectors through thorough and systematic in silico studies. Our results suggest that out of the total SSPs, 66% were predicted as effector proteins, released either into the apoplast or cytoplasm of the host cell. Of these, 68% were found to be intrinsically disordered effector proteins (IDEPs). Among the six distinct classes of disordered effectors, we observed peculiar relationships between the localization of several effectors in the apoplast or cytoplasm and the degree of disorder. We determined the degree of structural disorder and its impact on protein foldability across all the putative small secretory effector proteins from the blast pathogen, further validated by molecular dynamics simulation studies. This study provides definite clues toward unraveling the mystery behind the importance of structural distortions in effectors and their impact on plant-pathogen interactions. The study of these dynamical segments may help identify new effectors as well.Communicated by Ramaswamy H. Sarma.

Rigidifying Flexible Sites: A Promising Strategy to Improve Thermostability of Lysophospholipase From Pyrococcus abyssi

High thermostability of the enzymes is one of the distinguishing characteristics that increase their industrial utility. In the current research work, rigidifying the flexible amino acid residues of a lysophospholipase (Pa-LPL) from Pyrococcus abyssi was used as a protein engineering approach to improve its thermostability. A truncated variant of Pa-LPL (t-LPL∆12) was constructed by trimming its 12 amino acid residues (50-61) through overlap extension PCR. The truncated enzyme worked optimally at 65°C and pH 6.5 with remarkable thermostability at 65°C-85°C. In comparison to wild-type Pa-LPL, 5.8 and 1.2-fold increase in half-life (t1/2) of t-LPL∆12 was observed at 65 (optimum temperature) and 95°C, respectively. The activity of t-LPL∆12 was stimulated by 1 mM Cu2+ followed by Ca2+, Ni2+, Co2+, and Mg2+. Both substrate docking and experimental results indicated that the truncated enzyme could hydrolyze a variety of p-nitrophenyl esters. K m, V max, and K cat for enzymatic hydrolysis of p-nitrophenyl butyrate were calculated to be 1 ± 0.087 mM, 1456 ± 36.474 U/mg, and 1.397 × 1011 min-1, respectively. In short, broad substrate specificity and thermostability of t-LPL∆12 are some of the distinctive features that make it an ideal candidate for degumming of vegetable oils.

mdCATH: A Large-Scale MD Dataset for Data-Driven Computational Biophysics

Recent advancements in protein structure determination are revolutionizing our understanding of proteins. Still, a significant gap remains in the availability of comprehensive datasets that focus on the dynamics of proteins, which are crucial for understanding protein function, folding, and interactions. To address this critical gap, we introduce mdCATH, a dataset generated through an extensive set of all-atom molecular dynamics simulations of a diverse and representative collection of protein domains. This dataset comprises all-atom systems for 5,398 domains, modeled with a state-of-the-art classical force field, and simulated in five replicates each at five temperatures from 320 K to 450 K. The mdCATH dataset records coordinates and forces every 1 ns, for over 62 ms of accumulated simulation time, effectively capturing the dynamics of the various classes of domains and providing a unique resource for proteome-wide statistical analyses of protein unfolding thermodynamics and kinetics. We outline the dataset structure and showcase its potential through four easily reproducible case studies, highlighting its capabilities in advancing protein science.

Bacterial Diversity, Metabolic Profiling, and Application Potential of Antarctic Soil Metagenomes

Antarctica has attracted increasing interest in understanding its microbial communities, metabolic potential, and as a source of microbial hydrolytic enzymes with industrial applications, for which advances in next-generation sequencing technologies have greatly facilitated the study of unculturable microorganisms. In this work, soils from seven sub-Antarctic islands and Union Glacier were studied using a whole-genome shotgun metagenomic approach. The main findings were that the microbial community at all sites was predominantly composed of the bacterial phyla Actinobacteria and Cyanobacteria, and the families Streptomycetaceae and Pseudonocardiaceae. Regarding the xenobiotic biodegradation and metabolism pathway, genes associated with benzoate, chloroalkane, chloroalkene, and styrene degradation were predominant. In addition, putative genes encoding industrial enzymes with predicted structural properties associated with improved activity at low temperatures were found, with catalases and malto-oligosyltrehalose trehalohydrolase being the most abundant. Overall, our results show similarities between soils from different Antarctic sites with respect to more abundant bacteria and metabolic pathways, especially at higher classification levels, regardless of their geographic location. Furthermore, our results strengthen the potential of Antarctic soils as a source of industrially relevant enzymes.

iACP-DFSRA: Identification of Anticancer Peptides Based on a Dual-channel Fusion Strategy of ResCNN and Attention

Anticancer peptides (ACPs) have been widely applied in the treatment of cancer owing to good safety, rational side effects, and high selectivity. However, the number of ACPs that have been experimentally validated is limited as identification of ACPs is extremely expensive. Hence, accurate and cost-effective identification methods for ACPs are urgently needed. In this work, we proposed a deep learning-based model, named iACP-DFSRA, for ACPs identification. Specifically, we adopted two kinds of sequence embedding technologies, ProtBert_BFD pre-training language model and handcrafted features to encode protein sequences. Then, the LightGBM was used for feature selection, and the selected features were input into ResCNN and Attention mechanism, respectively, to extract local and global features. Finally, the concatenate features were deeply fused by using the Attention mechanism to allow key features to be paid more attention to by the model and make predictions by fully connected layer. The results of 10-fold cross-validation demonstrated that the iACP-DFSRA model delivered improved results in most metrics with Sp of 94.15%, Sn of 95.32%, Acc of 94.74% and MCC of 89.48% compared to the latest AACFlow model. Indeed, the iACP-DFSRA model is the only model with Acc > 90% and MCC > 80% on this independent test dataset. Furthermore, we have further demonstrated the superiority of our model on additional datasets. In addition, t-SNE and SHAP interpretation analysis demonstrated that it is crucial to use two channels for feature extraction and use the Attention mechanism for deep fusion, which helps the iACP-DFSRA to predict ACPs more effectively.

Assessment of Four Theoretical Approaches to Predict Protein Flexibility in the Crystal Phase and Solution

In this paper, we evaluated the ability of four coarse-grained methods to predict protein flexible regions with potential biological importance, UNRES-flex, UNRES-DSSP-flex (based on the united residue model of polypeptide chains without and with secondary structure restraints, respectively), CABS-flex (based on the C-α, C-β, and side chain model), and nonlinear rigid block normal mode analysis (NOLB) with a set of 100 protein structures determined by NMR spectroscopy or X-ray crystallography, with all secondary structure types. End regions with high fluctuations were excluded from analysis. The Pearson and Spearman correlation coefficients were used to quantify the conformity between the calculated and experimental fluctuation profiles, the latter determined from NMR ensembles and X-ray B-factors, respectively. For X-ray structures (corresponding to proteins in a crowded environment), NOLB resulted in the best agreement between the predicted and experimental fluctuation profiles, while for NMR structures (corresponding to proteins in solution), the ranking of performance is CABS-flex > UNRES-DSSP-flex > UNRES-flex > NOLB; however, CABS-flex sometimes exaggerated the extent of small fluctuations, as opposed to UNRES-DSSP-flex.

In Silico Functional and Structural Analysis of Non-synonymous Single Nucleotide Polymorphisms (nsSNPs) in Human Paired Box 4 Gene

In human genome, members of Paired box (PAX) transcription factor family are highly sequence-specific DNA-binding proteins. Among PAX gene family members, PAX4 gene has significant role in growth, proliferation, differentiation, and insulin secretion of pancreatic β-cells. Single nucleotide polymorphisms (SNPs) in PAX4 gene progress in the pathogenesis of various human diseases. Hence, the molecular mechanism of how these SNPs in PAX4 gene significantly progress diseases pathogenesis needs to be elucidated. For the reason, a series of bioinformatic analyzes were done to identify the SNPs of PAX4 gene that contribute in diseases pathogenesis. From the analyzes, 4145 SNPs (rsIDs) in PAX4 gene were obtained, where, 362 missense (8.73%), 169 synonymous (4.08%), and 2323 intron variants (56.04%). The rest SNPs were unspecified. Among the 362 missense variants, 118 nsSNPs were found as deleterious in SIFT analysis. Among those, 25 nsSNPs were most probably damaging and 23 were deleterious as observed in PolyPhen-2 and PROVEAN analyzes, respectively. Following all analyzes, 14 nsSNPs (rs149708455, rs115887120, rs147279315, rs35155575, rs370095957, rs373939873, rs145468905, rs121917718, rs2233580, rs3824004, rs372751660, rs369459316, rs375472849, rs372497946) were common and observed as deleterious, probably damaging, affective and diseases associated. Following structural analyzes, 11 nsSNPs guided proteins were found as most unstable and highly conserved. Among these, R20W, R39Q, R45Q, R60H, G65D, and A223D mutated proteins were highly harmful. Hence, the results from above-mentioned integrated comprehensive bioinformatic analyzes guide how different nsSNPs in PAX4 gene alter structural and functional characteristics of the protein that might progress diseases pathogenesis in human including type 2 diabetes.

In silico functional, structural and pathogenicity analysis of missense single nucleotide polymorphisms in human MCM6 gene

Single nucleotide polymorphisms (SNPs) are one of the most common determinants and potential biomarkers of human disease pathogenesis. SNPs could alter amino acid residues, leading to the loss of structural and functional integrity of the encoded protein. In humans, members of the minichromosome maintenance (MCM) family play a vital role in cell proliferation and have a significant impact on tumorigenesis. Among the MCM members, the molecular mechanism of how missense SNPs of minichromosome maintenance complex component 6 (MCM6) contribute to DNA replication and tumor pathogenesis is underexplored and needs to be elucidated. Hence, a series of sequence and structure-based computational tools were utilized to determine how mutations affect the corresponding MCM6 protein. From the dbSNP database, among 15,009 SNPs in the MCM6 gene, 642 missense SNPs (4.28%), 291 synonymous SNPs (1.94%), and 12,500 intron SNPs (83.28%) were observed. Out of the 642 missense SNPs, 33 were found to be deleterious during the SIFT analysis. Among these, 11 missense SNPs (I123S, R207C, R222C, L449F, V456M, D463G, H556Y, R602H, R633W, R658C, and P815T) were found as deleterious, probably damaging, affective and disease-associated. Then, I123S, R207C, R222C, V456M, D463G, R602H, R633W, and R658C missense SNPs were found to be highly harmful. Six missense SNPs (I123S, R207C, V456M, D463G, R602H, and R633W) had the potential to destabilize the corresponding protein as predicted by DynaMut2. Interestingly, five high-risk mutations (I123S, V456M, D463G, R602H, and R633W) were distributed in two domains (PF00493 and PF14551). During molecular dynamics simulations analysis, consistent fluctuation in RMSD and RMSF values, high Rg and hydrogen bonds in mutant proteins compared to wild-type revealed that these mutations might alter the protein structure and stability of the corresponding protein. Hence, the results from the analyses guide the exploration of the mechanism by which these missense SNPs of the MCM6 gene alter the structural integrity and functional properties of the protein, which could guide the identification of ways to minimize the harmful effects of these mutations in humans.

Application of SUMO fusion technology for the enhancement of stability and activity of lysophospholipase from Pyrococcus abyssi

Heterologous production of proteins in Escherichia coli has raised several challenges including soluble production of target proteins, high levels of expression and purification. Fusion tags can serve as the important tools to overcome these challenges. SUMO (small ubiquitin-related modifier) is one of these tags whose fusion to native protein sequence can enhance its solubility and stability. In current research, a simple, efficient and cost-effective method is being discussed for the construction of pET28a-SUMO vector. In order to improve the stability and activity of lysophospholipase from Pyrococcus abyssi (Pa-LPL), a 6xHis-SUMO tag was fused to N-terminal of Pa-LPL by using pET28a-SUMO vector. Recombinant SUMO-fused enzyme (6 H-S-PaLPL) works optimally at 35 °C and pH 6.5 with remarkable thermostability at 35-95 °C. Thermo-inactivation kinetics of 6 H-S-PaLPL were also studied at 35-95 °C with first order rate constant (kIN) of 5.58 × 10- 2 h-1 and half-life of 12 ± 0 h at 95 °C. Km and Vmax for the hydrolysis of 4-nitrophenyl butyrate were calculated to be 2 ± 0.015 mM and 3882 ± 22.368 U/mg, respectively. 2.4-fold increase in Vmax of Pa-LPL was observed after fusion of 6xHis-SUMO tag to its N-terminal. It is the first report on the utilization of SUMO fusion tag to enhance the overall stability and activity of Pa-LPL. Fusion of 6xHis-SUMO tag not only aided in the purification process but also played a crucial role in increasing the thermostability and activity of the enzyme. SUMO-fused enzyme, thus generated, can serve as an important candidate for degumming of vegetable oils at industrial scale.

Residue coevolution and mutational landscape for OmpR and NarL response regulator subfamilies

DNA-binding response regulators (DBRRs) are a broad class of proteins that operate in tandem with their partner kinase proteins to form two-component signal transduction systems in bacteria. Typical DBRRs are composed of two domains where the conserved N-terminal domain accepts transduced signals and the evolutionarily diverse C-terminal domain binds to DNA. These domains are assumed to be functionally independent, and hence recombination of the two domains should yield novel DBRRs of arbitrary input/output response, which can be used as biosensors. This idea has been proved to be successful in some cases; yet, the error rate is not trivial. Improvement of the success rate of this technique requires a deeper understanding of the linker-domain and inter-domain residue interactions, which have not yet been thoroughly examined. Here, we studied residue coevolution of DBRRs of the two main subfamilies (OmpR and NarL) using large collections of bacterial amino acid sequences to extensively investigate the evolutionary signatures of linker-domain and inter-domain residue interactions. Coevolutionary analysis uncovered evolutionarily selected linker-domain and inter-domain residue interactions of known experimental structures, as well as previously unknown inter-domain residue interactions. We examined the possibility of these inter-domain residue interactions as contacts that stabilize an inactive conformation of the DBRR where DNA binding is inhibited for both subfamilies. The newly gained insights on linker-domain/inter-domain residue interactions and shared inactivation mechanisms improve the understanding of the functional mechanism of DBRRs, providing clues to efficiently create functional DBRR-based biosensors. Additionally, we show the feasibility of applying coevolutionary landscape models to predict the functionality of domain-swapped DBRR proteins. The presented result demonstrates that sequence information can be used to filter out bioengineered DBRR proteins that are predicted to be nonfunctional due to a high negative predictive value.

A Perspective on the Prospective Use of AI in Protein Structure Prediction

AlphaFold2 (AF2) and RoseTTaFold (RF) have revolutionized structural biology, serving as highly reliable and effective methods for predicting protein structures. This article explores their impact and limitations, focusing on their integration into experimental pipelines and their application in diverse protein classes, including membrane proteins, intrinsically disordered proteins (IDPs), and oligomers. In experimental pipelines, AF2 models help X-ray crystallography in resolving the phase problem, while complementarity with mass spectrometry and NMR data enhances structure determination and protein flexibility prediction. Predicting the structure of membrane proteins remains challenging for both AF2 and RF due to difficulties in capturing conformational ensembles and interactions with the membrane. Improvements in incorporating membrane-specific features and predicting the structural effect of mutations are crucial. For intrinsically disordered proteins, AF2's confidence score (pLDDT) serves as a competitive disorder predictor, but integrative approaches including molecular dynamics (MD) simulations or hydrophobic cluster analyses are advocated for accurate dynamics representation. AF2 and RF show promising results for oligomeric models, outperforming traditional docking methods, with AlphaFold-Multimer showing improved performance. However, some caveats remain in particular for membrane proteins. Real-life examples demonstrate AF2's predictive capabilities in unknown protein structures, but models should be evaluated for their agreement with experimental data. Furthermore, AF2 models can be used complementarily with MD simulations. In this Perspective, we propose a "wish list" for improving deep-learning-based protein folding prediction models, including using experimental data as constraints and modifying models with binding partners or post-translational modifications. Additionally, a meta-tool for ranking and suggesting composite models is suggested, driving future advancements in this rapidly evolving field.

ATLAS: protein flexibility description from atomistic molecular dynamics simulations

Dynamical behaviour is one of the most crucial protein characteristics. Despite the advances in the field of protein structure resolution and prediction, analysis and prediction of protein dynamic properties remains a major challenge, mostly due to the low accessibility of data and its diversity and heterogeneity. To address this issue, we present ATLAS, a database of standardised all-atom molecular dynamics simulations, accompanied by their analysis in the form of interactive diagrams and trajectory visualisation. ATLAS offers a large-scale view and valuable insights on protein dynamics for a large and representative set of proteins, by combining data obtained through molecular dynamics simulations with information extracted from experimental structures. Users can easily analyse dynamic properties of functional protein regions, such as domain limits (hinge positions) and residues involved in interaction with other biological molecules. Additionally, the database enables exploration of proteins with uncommon dynamic properties conditioned by their environment such as chameleon subsequences and Dual Personality Fragments. The ATLAS database is freely available at https://www.dsimb.inserm.fr/ATLAS.

Unsupervised Analysis of Optical Imaging Data for the Discovery of Reactivity Patterns in Metal Alloy

Operando wide-field optical microscopy imaging yields a wealth of information about the reactivity of metal interfaces, yet the data are often unstructured and challenging to process. In this study, the power of unsupervised machine learning (ML) algorithms is harnessed to analyze chemical reactivity images obtained dynamically by reflectivity microscopy in combination with ex situ scanning electron microscopy to identify and cluster the chemical reactivity of particles in Al alloy. The ML analysis uncovers three distinct clusters of reactivity from unlabeled datasets. A detailed examination of representative reactivity patterns confirms the chemical communication of generated OH- fluxes within particles, as supported by statistical analysis of size distribution and finite element modelling (FEM). The ML procedures also reveal statistically significant patterns of reactivity under dynamic conditions, such as pH acidification. The results align well with a numerical model of chemical communication, underscoring the synergy between data-driven ML and physics-driven FEM approaches.

LibGENiE - A bioinformatic pipeline for the design of information-enriched enzyme libraries

Enzymes are potent catalysts with high specificity and selectivity. To leverage nature's synthetic potential for industrial applications, various protein engineering techniques have emerged which allow to tailor the catalytic, biophysical, and molecular recognition properties of enzymes. However, the many possible ways a protein can be altered forces researchers to carefully balance between the exhaustiveness of an enzyme screening campaign and the required resources. Consequently, the optimal engineering strategy is often defined on a case-by-case basis. Strikingly, while predicting mutations that lead to an improved target function is challenging, here we show that the prediction and exclusion of deleterious mutations is a much more straightforward task as analyzed for an engineered carbonic acid anhydrase, a transaminase, a squalene-hopene cyclase and a Kemp eliminase. Combining such a pre-selection of allowed residues with advanced gene synthesis methods opens a path toward an efficient and generalizable library construction approach for protein engineering. To give researchers easy access to this methodology, we provide the website LibGENiE containing the bioinformatic tools for the library design workflow.

Protein Fitness Prediction Is Impacted by the Interplay of Language Models, Ensemble Learning, and Sampling Methods

Advances in machine learning (ML) and the availability of protein sequences via high-throughput sequencing techniques have transformed the ability to design novel diagnostic and therapeutic proteins. ML allows protein engineers to capture complex trends hidden within protein sequences that would otherwise be difficult to identify in the context of the immense and rugged protein fitness landscape. Despite this potential, there persists a need for guidance during the training and evaluation of ML methods over sequencing data. Two key challenges for training discriminative models and evaluating their performance include handling severely imbalanced datasets (e.g., few high-fitness proteins among an abundance of non-functional proteins) and selecting appropriate protein sequence representations (numerical encodings). Here, we present a framework for applying ML over assay-labeled datasets to elucidate the capacity of sampling techniques and protein encoding methods to improve binding affinity and thermal stability prediction tasks. For protein sequence representations, we incorporate two widely used methods (One-Hot encoding and physiochemical encoding) and two language-based methods (next-token prediction, UniRep; masked-token prediction, ESM). Elaboration on performance is provided over protein fitness, protein size, and sampling techniques. In addition, an ensemble of protein representation methods is generated to discover the contribution of distinct representations and improve the final prediction score. We then implement multiple criteria decision analysis (MCDA; TOPSIS with entropy weighting), using multiple metrics well-suited for imbalanced data, to ensure statistical rigor in ranking our methods. Within the context of these datasets, the synthetic minority oversampling technique (SMOTE) outperformed undersampling while encoding sequences with One-Hot, UniRep, and ESM representations. Moreover, ensemble learning increased the predictive performance of the affinity-based dataset by 4% compared to the best single-encoding candidate (F1-score = 97%), while ESM alone was rigorous enough in stability prediction (F1-score = 92%).

Predicting DNA-binding protein and coronavirus protein flexibility using protein dihedral angle and sequence feature

The flexibility of protein structure is related to various biological processes, such as molecular recognition, allosteric regulation, catalytic activity, and protein stability. At the molecular level, protein dynamics and flexibility are important factors to understand protein function. DNA-binding proteins and Coronavirus proteins are of great concern and relatively unique proteins. However, exploring the flexibility of DNA-binding proteins and Coronavirus proteins through experiments or calculations is a difficult process. Since protein dihedral rotational motion can be used to predict protein structural changes, it provides key information about protein local conformation. Therefore, this paper introduces a method to improve the accuracy of protein flexibility prediction, DihProFle (Prediction of DNA-binding proteins and Coronavirus proteins flexibility introduces the calculated dihedral Angle information). Based on protein dihedral Angle information, protein evolution information, and amino acid physical and chemical properties, DihProFle realizes the prediction of protein flexibility in two cases on DNA-binding proteins and Coronavirus proteins, and assigns flexibility class to each protein sequence position. In this study, compared with the flexible prediction using sequence evolution information, and physicochemical properties of amino acids, the flexible prediction accuracy based on protein dihedral Angle information, sequence evolution information and physicochemical properties of amino acids improved by 2.2% and 3.1% in the nonstrict and strict conditions, respectively. And DihProFle achieves better performance than previous methods for protein flexibility analysis. In addition, we further analyzed the correlation of amino acid properties and protein dihedral angles with residues flexibility. The results show that the charged hydrophilic residues have higher proportion in the flexible region, and the rigid region tends to be in the angular range of the protein dihedral angle (such as the ψ angle of amino acid residues is more flexible than rigid in the range of 91°-120°). Therefore, the results indicate that hydrophilic residues and protein dihedral angle information play an important role in protein flexibility.

An agnostic analysis of the human AlphaFold2 proteome using local protein conformations

Knowledge of the 3D structure of proteins is a valuable asset for understanding their precise biological mechanisms. However, the cost of production of 3D structures and experimental difficulties limit their obtaining. The proposal of 3D structural models is consequently an appealing alternative. The release of the AlphaFold Deep Learning approach has revolutionized the field. The recent near-complete human proteome proposal makes it possible to analyse large amounts of data and evaluate the results of the approach in greater depth. The 3D human proteome was thus analysed in light of the classic secondary structures, and many less-used protein local conformations (PolyProline II helices, type of γ-turns, of β-turns and of β-bulges, curvature of the helices, and a structural alphabet). Without questioning the global quality of the approach, this analysis highlights certain local conformations, which maybe poorly predicted and they could therefore be better addressed.

Molecular and structural characterization of a novel high-prevalence antigen of the Augustine blood group system

BACKGROUND

An antibody directed against a high-prevalence red blood cell (RBC) antigen was detected in a 67-year-old female patient of North African ancestry with a history of a single pregnancy and blood transfusion. So far, the specificity of the proband's alloantibody remained unknown in our immunohematology reference laboratory.

STUDY DESIGN AND METHODS

Whole-exome sequencing (WES) was performed on the proband's DNA. The reactivity to the SLC29A1-encoded ENT1 adenosine transporter was investigated by flow cytometry analyses of ENT1-expressing HEK293 cells, and RBCs from Augustine-typed individuals. Erythrocyte protein expression level, nucleoside-binding capacity, and molecular structure of the proband's ENT1 variant were further explored by western blot, flow cytometry, and molecular dynamics calculations, respectively.

RESULTS

A missense variant was identified in the SLC29A1 gene, which encodes the Augustine blood group system. It arises from homozygosity for a rare c.242A > G missense mutation that results in a nonsynonymous p.Asn81Ser substitution within the large extracellular loop of ENT1. Flow cytometry analyses demonstrated that the proband's antibody was reactive against HEK-293 cells transfected with control but not proband's SLC29A1 cDNA. Consistent with this finding, proband's antibody was found to be reactive with At(a-) (AUG:-2), but not AUG:-1 (null phenotype) RBCs. Data from structural analysis further supported that the proband's p.Asn81Ser variation does not alter ENT1 binding of its specific inhibitor NBMPR.

CONCLUSION

Our study provides evidence for a novel high-prevalence antigen, AUG4 (also called ATAM after the proband's name) in the Augustine blood group system, encoded by the rare SLC29A1 variant allele AUG*04 (c.242A > G, p.Asn81Ser).

The Conformation of the Intrinsically Disordered N-Terminal Region of Barrier-to-Autointegration Factor (BAF) is Regulated by pH and Phosphorylation

Barrier-to-Autointegration Factor (BAF) is a highly conserved DNA binding protein important for genome integrity. Its localization and function are regulated through phosphorylation. Previously reported structures of BAF suggested that it is fully ordered, but our recent NMR analysis revealed that its N-terminal region is flexible in solution and that S4/T3 di-phosphorylation by VRK1 reduces this flexibility. Here, molecular dynamics (MD) simulation was used to unveil the conformational ensembles accessible to the N-terminal region of BAF either unphosphorylated, mono-phosphorylated on S4 or di-phosphorylated on S4/T3 (pBAF) and to reveal the interactions that contribute to define these ensembles. We show that the intrinsic flexibility observed in the N-terminal region of BAF is reduced by S4 phosphorylation and to a larger extent by S4/T3 di-phosphorylation. Thanks to the atomic description offered by MD supported by the NMR study of several BAF mutants, we identified the dynamic network of salt bridge interactions responsible for the conformational restriction involving pS4 and pT3 with residues located in helix α1 and α6. Using MD, we showed that the flexibility in the N-terminal region of BAF depends on the ionic strength and on the pH. We show that the presence of two negative charges of the phosphoryl groups is required for a substantial decrease in flexibility in pBAF. Using MD supported by NMR, we also showed that H7 deprotonation reduces the flexibility in the N-terminal region of BAF. Thus, the conformation of the intrinsically disordered N-terminal region of BAF is highly tunable, likely related to its diverse functions.

Predicting the effects of rare genetic variants on oncogenic signaling pathways: A computational analysis of HRAS protein function

The HRAS gene plays a crucial role in regulating essential cellular processes for life, and this gene's misregulation is linked to the development of various types of cancers. Nonsynonymous single nucleotide polymorphisms (nsSNPs) within the coding region of HRAS can cause detrimental mutations that disrupt wild-type protein function. In the current investigation, we have employed in-silico methodologies to anticipate the consequences of infrequent genetic variations on the functional properties of the HRAS protein. We have discovered a total of 50 nsSNPs, of which 23 were located in the exon region of the HRAS gene and denoting that they were expected to cause harm or be deleterious. Out of these 23, 10 nsSNPs ([G60V], [G60D], [R123P], [D38H], [I46T], [G115R], [R123G], [P11OL], [A59L], and [G13R]) were identified as having the most delterious effect based on results of SIFT analysis and PolyPhen2 scores ranging from 0.53 to 69. The DDG values -3.21 kcal/mol to 0.87 kcal/mol represent the free energy change associated with protein stability upon mutation. Interestingly, we identified that the three mutations (Y4C, T58I, and Y12E) were found to improve the structural stability of the protein. We performed molecular dynamics (MD) simulations to investigate the structural and dynamic effects of HRAS mutations. Our results showed that the stable model of HRAS had a significantly lower energy value of -18756 kj/mol compared to the initial model of -108915 kj/mol. The RMSD value for the wild-type complex was 4.40 Å, and the binding energies for the G60V, G60D, and D38H mutants were -107.09 kcal/mol, -109.42 kcal/mol, and -107.18 kcal/mol, respectively as compared to wild-type HRAS protein had -105.85 kcal/mol. The result of our investigation presents convincing corroboration for the potential functional significance of nsSNPs in augmenting HRAS expression and adding to the activation of malignant oncogenic signalling pathways.

A genomic approach to analyze the cold adaptation of yeasts isolated from Italian Alps

Microorganisms including yeasts are responsible for mineralization of organic matter in cold regions, and their characterization is critical to elucidate the ecology of such environments on Earth. Strategies developed by yeasts to survive in cold environments have been increasingly studied in the last years and applied to different biotechnological applications, but their knowledge is still limited. Microbial adaptations to cold include the synthesis of cryoprotective compounds, as well as the presence of a high number of genes encoding the synthesis of proteins/enzymes characterized by a reduced proline content and highly flexible and large catalytic active sites. This study is a comparative genomic study on the adaptations of yeasts isolated from the Italian Alps, considering their growth kinetics. The optimal temperature for growth (OTG), growth rate (Gr), and draft genome sizes considerably varied (OTG, 10°C–20°C; Gr, 0.071–0.0726; genomes, 20.7–21.5 Mpb; %GC, 50.9–61.5). A direct relationship was observed between calculated protein flexibilities and OTG, but not for Gr. Putative genes encoding for cold stress response were found, as well as high numbers of genes encoding for general, oxidative, and osmotic stresses. The cold response genes found in the studied yeasts play roles in cell membrane adaptation, compatible solute accumulation, RNA structure changes, and protein folding, i.e., dihydrolipoamide dehydrogenase, glycogen synthase, omega-6 fatty acid, stearoyl-CoA desaturase, ATP-dependent RNA helicase, and elongation of very-long-chain fatty acids. A redundancy for several putative genes was found, higher for P-loop containing nucleoside triphosphate hydrolase, alpha/beta hydrolase, armadillo repeat-containing proteins, and the major facilitator superfamily protein. Hundreds of thousands of small open reading frames (SmORFs) were found in all studied yeasts, especially in Phenoliferia glacialis. Gene clusters encoding for the synthesis of secondary metabolites such as terpene, non-ribosomal peptide, and type III polyketide were predicted in four, three, and two studied yeasts, respectively.

Solution structure of the type I polyketide synthase Pks13 from Mycobacterium tuberculosis

BACKGROUND

Type I polyketide synthases (PKSs) are multifunctional enzymes responsible for the biosynthesis of a group of diverse natural compounds with biotechnological and pharmaceutical interest called polyketides. The diversity of polyketides is impressive despite the limited set of catalytic domains used by PKSs for biosynthesis, leading to considerable interest in deciphering their structure-function relationships, which is challenging due to high intrinsic flexibility. Among nineteen polyketide synthases encoded by the genome of Mycobacterium tuberculosis, Pks13 is the condensase required for the final condensation step of two long acyl chains in the biosynthetic pathway of mycolic acids, essential components of the cell envelope of Corynebacterineae species. It has been validated as a promising druggable target and knowledge of its structure is essential to speed up drug discovery to fight against tuberculosis.

RESULTS

We report here a quasi-atomic model of Pks13 obtained using small-angle X-ray scattering of the entire protein and various molecular subspecies combined with known high-resolution structures of Pks13 domains or structural homologues. As a comparison, the low-resolution structures of two other mycobacterial polyketide synthases, Mas and PpsA from Mycobacterium bovis BCG, are also presented. This study highlights a monomeric and elongated state of the enzyme with the apo- and holo-forms being identical at the resolution probed. Catalytic domains are segregated into two parts, which correspond to the condensation reaction per se and to the release of the product, a pivot for the enzyme flexibility being at the interface. The two acyl carrier protein domains are found at opposite sides of the ketosynthase domain and display distinct characteristics in terms of flexibility.

CONCLUSIONS

The Pks13 model reported here provides the first structural information on the molecular mechanism of this complex enzyme and opens up new perspectives to develop inhibitors that target the interactions with its enzymatic partners or between catalytic domains within Pks13 itself.

Turning Failures into Applications: The Problem of Protein ΔΔG Prediction

After nearly two decades of research in the field of computational methods based on machine learning and knowledge-based potentials for ΔG and ΔΔG prediction upon variations, we now realize that all the approaches are poorly performing when tested on specific cases and that there is large space for improvement. Why this is so? Is it wrong the underlying assumption that experimental protein thermodynamics in solution reflects the thermodynamics of a single protein? Both machine learning and knowledge-based computational methods are rigorous and we know the solid theory behind. We are now in a critical situation, which suggests that predictions of protein instability upon variation should be considered with care. In the following, we will show how to cope with the problem of understanding which protein positions may be of interest for biotechnological and biomedical purposes. By applying a consensus procedure, we indicate possible strategies for the result interpretation.

VHH Structural Modelling Approaches: A Critical Review

VHH, i.e., VH domains of camelid single-chain antibodies, are very promising therapeutic agents due to their significant physicochemical advantages compared to classical mammalian antibodies. The number of experimentally solved VHH structures has significantly improved recently, which is of great help, because it offers the ability to directly work on 3D structures to humanise or improve them. Unfortunately, most VHHs do not have 3D structures. Thus, it is essential to find alternative ways to get structural information. The methods of structure prediction from the primary amino acid sequence appear essential to bypass this limitation. This review presents the most extensive overview of structure prediction methods applied for the 3D modelling of a given VHH sequence (a total of 21). Besides the historical overview, it aims at showing how model software programs have been shaping the structural predictions of VHHs. A brief explanation of each methodology is supplied, and pertinent examples of their usage are provided. Finally, we present a structure prediction case study of a recently solved VHH structure. According to some recent studies and the present analysis, AlphaFold 2 and NanoNet appear to be the best tools to predict a structural model of VHH from its sequence.

Response to Cold: A Comparative Transcriptomic Analysis in Eight Cold-Adapted Yeasts

Microorganisms have evolved to colonize all biospheres, including extremely cold environments, facing several stressor conditions, mainly low/freezing temperatures. In general, terms, the strategies developed by cold-adapted microorganisms include the synthesis of cryoprotectant and stress-protectant molecules, cold-active proteins, especially enzymes, and membrane fluidity regulation. The strategy could differ among microorganisms and concerns the characteristics of the cold environment of the microorganism, such as seasonal temperature changes. Microorganisms can develop strategies to grow efficiently at low temperatures or tolerate them and grow under favorable conditions. These differences can be found among the same kind of microorganisms and from the same cold habitat. In this work, eight cold-adapted yeasts isolated from King George Island, subAntarctic region, which differ in their growth properties, were studied about their response to low temperatures at the transcriptomic level. Sixteen ORFeomes were assembled and used for gene prediction and functional annotation, determination of gene expression changes, protein flexibilities of translated genes, and codon usage bias. Putative genes related to the response to all main kinds of stress were found. The total number of differentially expressed genes was related to the temperature variation that each yeast faced. The findings from multiple comparative analyses among yeasts based on gene expression changes and protein flexibility by cellular functions and codon usage bias raise significant differences in response to cold among the studied Antarctic yeasts. The way a yeast responds to temperature change appears to be more related to its optimal temperature for growth (OTG) than growth velocity. Yeasts with higher OTG prepare to downregulate their metabolism to enter the dormancy stage. In comparison, yeasts with lower OTG perform minor adjustments to make their metabolism adequate and maintain their growth at lower temperatures.

An in-frame deletion mutation in the degron tail of auxin coreceptor IAA2 confers resistance to the herbicide 2,4-D in Sisymbrium orientale

Synthetic auxin herbicides intersect basic plant developmental biology and applied weed management. We investigated resistance to 2,4-D in the Australian weed Sisymbrium orientale (Indian hedge mustard). We identified a mechanism involving an in-frame 27-bp deletion in the degron tail of auxin coreceptor IAA2, one member of the gene family of Aux/IAA auxin co-receptors. We show that this deletion in IAA2 is a gain-of-function mutation that confers synthetic auxin resistance. This field-evolved mechanism of resistance to synthetic auxin herbicides confirms previous biochemical studies showing the role of the Aux/IAA degron tail in regulating Aux/IAA protein degradation upon auxin perception. The deletion mutation could be generated in crops using gene-editing approaches for cross-resistance to multiple synthetic auxin herbicides.

The natural auxin indole-3-acetic acid (IAA) is a key regulator of many aspects of plant growth and development. Synthetic auxin herbicides such as 2,4-D mimic the effects of IAA by inducing strong auxinic-signaling responses in plants. To determine the mechanism of 2,4-D resistance in a Sisymbrium orientale (Indian hedge mustard) weed population, we performed a transcriptome analysis of 2,4-D-resistant (R) and -susceptible (S) genotypes that revealed an in-frame 27-nucleotide deletion removing nine amino acids in the degron tail (DT) of the auxin coreceptor Aux/IAA2 (SoIAA2). The deletion allele cosegregated with 2,4-D resistance in recombinant inbred lines. Further, this deletion was also detected in several 2,4-D-resistant field populations of this species. Arabidopsis transgenic lines expressing the SoIAA2 mutant allele were resistant to 2,4-D and dicamba. The IAA2-DT deletion reduced binding to TIR1 in vitro with both natural and synthetic auxins, causing reduced association and increased dissociation rates. This mechanism of synthetic auxin herbicide resistance assigns an in planta function to the DT region of this Aux/IAA coreceptor for its role in synthetic auxin binding kinetics and reveals a potential biotechnological approach to produce synthetic auxin-resistant crops using gene-editing.

Pseudonocardia humida sp. nov., an Actinomycete Isolated from Mangrove Soil Showing Distinct Distribution Pattern of Biosynthetic Gene Clusters

A novel actinomycete strain, designated S2-4^T, was isolated from a mangrove soil sample, and a polyphasic approach was employed to determine its taxonomic position. Phylogenetic analysis based on 16S rRNA gene indicated that strain S2-4^T formed a unique clade next to that harboring Pseudonocardia dioxanivorans CB1190^T, which shared the highest sequence similarity (98.20%) with the new isolate. Phylogenetic analysis based on core genes of genomic sequences displayed a different scenario, exhibiting closer phylogenetic relationship of strain S2-4^T with several species with 16S rRNA gene sequence similarities ranging from 96.95 to 98.06%, which was confirmed by the phylogenetic tree reconstructed based on genomic sequences. Further, substantial differences between the genotypic properties of strain S2-4^T and its closest neighbors, including digital DNA-DNA hybridization, average nucleotide identity, and distribution patterns of biosynthetic gene clusters (BGC), indicated the taxonomic position of strain S2-4^T as a novel species of the genus Pseudonocardia. Accordingly, strain S2-4^T was observed to show a different distribution pattern of a predicted BGC encoding ectoine by comparative genomic analysis, which could be strongly linked to its unique habitat distinct from where its close neighbors were isolated. The major cellular fatty acids were iso-C_15:0, C_21:0, and iso-C_16:0. The predominant menaquinone was MK-8(H₄). The polar lipids were composed of diphosphatidylglycerol, phosphatidylethanolamine, phosphatidylglycerol, phosphatidyl-N-monomethylethanolamine, phosphatidylcholine, phosphatidylinositol, phosphatidylinositol mannosides, and two unidentified glycolipids. Here, we propose a novel species of the genus Pseudonocardia: Pseudonocardia humida sp. nov. with the type strain S2-4^T (= JCM 34291^T = CGMCC 4.7706^T).

Uses and Abuses of the Atomic Displacement Parameters in Structural Biology

B-factors determined with X-ray crystallographic analyses are commonly used to estimate the flexibility degree of atoms, residues, and molecular moieties in biological macromolecules. In this chapter, the most recent studies and applications of B-factors in protein engineering and structural biology are briefly summarized. Particular emphasis is given to the limitations in using B-factors, in order to prevent inappropriate applications. It is eventually predicted that future applications will involve anisotropically refined B-factors, deep learning, and data produced by cryo-EM.

DeepREx-WS: A web server for characterising protein-solvent interaction starting from sequence

Protein–solvent interaction provides important features for protein surface engineering when the structure is absent or partially solved. Presently, we can integrate the notion of solvent exposed/buried residues with that of their flexibility and intrinsic disorder to highlight regions where mutations may increase or decrease protein stability in order to modify proteins for biotechnological reasons, while preserving their functional integrity. Here we describe a web server, which provides the unique possibility of integrating knowledge of solvent and non-solvent exposure with that of residue conservation, flexibility and disorder of a protein sequence, for a better understanding of which regions are relevant for protein integrity. The core of the webserver is DeepREx, a novel deep learning-based tool that classifies each residue in the sequence as buried or exposed. DeepREx is trained on a high-quality, non-redundant dataset derived from the Protein Data Bank comprising 2332 monomeric protein chains and benchmarked on a blind test set including 200 protein sequences unrelated with the training set. Results show that DeepREx performs at the state-of-the-art in the field. In turn, the Web Server, DeepREx-WS, supplements the predictions of DeepREx with features that allow a better characterisation of exposed and buried regions: i) residue conservation derived from multiple sequence alignment; ii) local sequence hydrophobicity; iii) residue flexibility computed with MEDUSA; iv) a predictor of secondary structure; v) the presence of disordered regions as derived from MobiDB-Lite3.0. The web server allows browsing, selecting and intersecting the different features. We demonstrate a possible application of the DeepREx-WS for assisting the identification of residues to be variated in protein surface engineering processes.

Discovering the Ultimate Limits of Protein Secondary Structure Prediction

Secondary structure prediction (SSP) of proteins is an important structural biology technique with many applications. There have been ~300 algorithms published in the past seven decades with fierce competition in accuracy. In the first 60 years, the accuracy of three-state SSP rose from ~56% to 81%; after that, it has long stayed at 81-86%. In the 1990s, the theoretical limit of three-state SSP accuracy had been estimated to be 88%. Thus, SSP is now generally considered not challenging or too challenging to improve. However, we found that the limit of three-state SSP might be underestimated. Besides, there is still much room for improving segment-based and eight-state SSPs, but the limits of these emerging topics have not been determined. This work performs large-scale sequence and structural analyses to estimate SSP accuracy limits and assess state-of-the-art SSP methods. The limit of three-state SSP is re-estimated to be ~92%, 4-5% higher than previously expected, indicating that SSP is still challenging. The estimated limit of eight-state SSP is 84-87%. Several proposals for improving future SSP algorithms are made based on our results. We hope that these findings will help move forward the development of SSP and all its applications.

PYTHIA: Deep Learning Approach for Local Protein Conformation Prediction

Protein Blocks (PBs) are a widely used structural alphabet describing local protein backbone conformation in terms of 16 possible conformational states, adopted by five consecutive amino acids. The representation of complex protein 3D structures as 1D PB sequences was previously successfully applied to protein structure alignment and protein structure prediction. In the current study, we present a new model, PYTHIA (predicting any conformation at high accuracy), for the prediction of the protein local conformations in terms of PBs directly from the amino acid sequence. PYTHIA is based on a deep residual inception-inside-inception neural network with convolutional block attention modules, predicting 1 of 16 PB classes from evolutionary information combined to physicochemical properties of individual amino acids. PYTHIA clearly outperforms the LOCUSTRA reference method for all PB classes and demonstrates great performance for PB prediction on particularly challenging proteins from the CASP14 free modelling category.