Single Protein Model Research Articles

Constructing a strongly interacting Pea-Cod binary protein system by introducing metal cations toward enhanced gelling properties

Developing prospective plant-animal binary protein systems with desirable nutritional and rheological properties stands as a significant and challenging pursuit within the food industry. Our understanding of the effect of adding salt on the aggregation behavior of food proteins is currently based on single model protein systems, however, this knowledge is rather limited following binary protein systems. Herein, various ionic strength settings are used to mitigate the repulsive forces between pea-cod mixed proteins during the thermal process, which further benefits the construction of a strengthened gel network. Transmission electron microscopy (TEM) and dynamic light scattering (DLS) collectively demonstrated that larger heat-induced protein aggregates were formed, which increased in size with higher ionic strength. In the presence of 2.5 mM CaCl2 and 50 mM NaCl, the disulfide bonds significantly increased from 19.3 to 27.53 and 30.5 μM/g, respectively. Notably, similar aggregation behavior could be found when introducing 2.5 mM CaCl2 or 25 mM NaCl, due to the enhanced aggregation tendency by specific binding of Ca2+ to proteins. With relevance to the strengthened cross-links between protein molecules, salt endowed composite gels with preferable gelling properties, evidenced by increased storage modulus. Additionally, the gelling temperature of mixed proteins decreased below 50 °C at elevated ionic strength. Simultaneously, the proportion of network proteins in composite gels increased remarkably from 82.05 % to 93.61 % and 92.31 % upon adding 5.0 mM CaCl2 and 100 mM NaCl, respectively. The findings provide a valuable foundation for designing economically viable and health-oriented plant-animal binary protein systems.

Clinical validation of a multi-protein, serum-based assay for disease activity assessments in multiple sclerosis

An 18-protein multiple sclerosis (MS) disease activity (DA) test was validated based on associations between algorithm scores and clinical/radiographic assessments (N = 614 serum samples; Train [n = 426; algorithm development] and Test [n = 188; evaluation] subsets). The multi-protein model was trained based on presence/absence of gadolinium-positive (Gd+) lesions and was also strongly associated with new/enlarging T2 lesions, and active versus stable disease (composite of radiographic and clinical evidence of DA) with improved performance (p < 0.05) compared to the neurofilament light single protein model. The odds of having ≥1 Gd+ lesions with a moderate/high DA score were 4.49 times that of a low DA score, and the odds of having ≥2 Gd+ lesions with a high DA score were 20.99 times that of a low/moderate DA score. The MSDA Test was clinically validated with improved performance compared to the top-performing single-protein model and can serve as a quantitative tool to enhance the care of MS patients.

Plasma Proteomic Signature of Cellular Senescence and Markers of Biological Aging Among Postmenopausal Women.

We aimed to investigate the association of circulatory senescence-associated secretory phenotypes (SASPs) produced by senescent cells with chronological and menopausal age in women aged 45 years or more. The proteomic profiles for 32 SASP factors of plasma samples were measured in 76 healthy postmenopausal women aged 46-82 years from the Korean Genome and Epidemiology Study Cardiovascular Disease Association Study (KoGES-CAVAS). We assessed the association between the SASP factors and aging indicators (chronological age, menopausal age, and years since menopause) using single- and multiprotein models. First, we composed a profile of proteins associated with chronological age, menopausal age, and years since menopause. In a single-protein model, three proteins (growth differentiation factor 15 [GDF15], insulin-like growth factor binding protein-2 [IGFBP-2], and tumor necrosis factor-α [TNF-α]) are positively associated with chronological age. Menopausal age and years since menopause are interrelated with interleukin-8 (IL-8). The direction of association between menopausal age and monocyte chemoattractant protein-1 (MCP-1) was only negative, and IGFBP-2 and TNF-α were significant in all three aging factors. We also constructed parsimonious multiprotein models to confirm the association of the proteomic signature for aging after adjusting for covariates and the combination of proteomic signature of 13 proteins (GDF15, interferon-γ [IFN-γ], IGFBP-2, IGFBP-7, IL-15, IL-1β, IL-17A, IL-8, MCP-1, tissue inhibitors of metalloproteinase-2 [TIMP-2], TNF-α, vascular endothelial growth factor-A [VEGF-A], and interferon-inducible protein 10 [IP-10]) appear to be associated with chronological age and menopausal state of individuals. Thus, by observing association between the selected SASPs and age-related markers among healthy postmenopausal women, we examine how menopause in women relates to proteomic indicators of aging and highlight the potential use of SASP factors as a marker to reflect the state of biological aging attributed by ovarian senescence.

DISTEMA: distance map-based estimation of single protein model accuracy with attentive 2D convolutional neural network

BackgroundEstimation of the accuracy (quality) of protein structural models is important for both prediction and use of protein structural models. Deep learning methods have been used to integrate protein structure features to predict the quality of protein models. Inter-residue distances are key information for predicting protein’s tertiary structures and therefore have good potentials to predict the quality of protein structural models. However, few methods have been developed to fully take advantage of predicted inter-residue distance maps to estimate the accuracy of a single protein structural model.ResultWe developed an attentive 2D convolutional neural network (CNN) with channel-wise attention to take only a raw difference map between the inter-residue distance map calculated from a single protein model and the distance map predicted from the protein sequence as input to predict the quality of the model. The network comprises multiple convolutional layers, batch normalization layers, dense layers, and Squeeze-and-Excitation blocks with attention to automatically extract features relevant to protein model quality from the raw input without using any expert-curated features. We evaluated DISTEMA’s capability of selecting the best models for CASP13 targets in terms of ranking loss of GDT-TS score. The ranking loss of DISTEMA is 0.079, lower than several state-of-the-art single-model quality assessment methods.ConclusionThis work demonstrates that using raw inter-residue distance information with deep learning can predict the quality of protein structural models reasonably well. DISTEMA is freely at https://github.com/jianlin-cheng/DISTEMA

Protein simulation in supercritical CO2: The challenge of force field

Supercritical CO2 is one of the most important green solvents for biotechnological applications. Due to the high-pressure conditions and gas–liquid properties of the fluid, the structural investigations of the biological macromolecules such as proteins and enzymes in supercritical conditions is problematic. Molecular dynamics simulation has emerged as a good tool to study the molecular-levels phenomena in macromolecule/supercritical CO2 environments. Many unsolved aspects of such systems have been explored by this method in recent years. One of the most important challenges in biomolecular simulations in supercritical CO2 is the applied force field for the bio-macromolecules. Most of simulations in supercritical CO2 have been performed by united-atoms and GROMOS96-based force fields for proteins and enzymes. The reliability of the obtained results should be examined by applying the other protein force fields. There is no comprehensive work about the effect of the force field on a single protein model in supercritical CO2. In this work, I have applied four famous biomolecular force fields including: GROMOS96 (43a1), CHARMM (27), AMBER (03) and OPLS-AA for simulation of chymotrypsin inhibitor 2 as a model protein in supercritical CO2 and also in the aqueous conditions. Different structural properties of protein have been examined using the different force fields. Results show that all of the force fields confirm the structural deviations of the protein structure from its native state in supercritical CO2. GROMOS43a1, similar to the previous simulations, showed the higher flexibility of protein backbone in scCO2 in comparison to the aqueous condition. In contrast, in simulations with all-atom force fields, the protein flexibility was almost identical in two solvents. Simulations by all of the force fields confirm the previously proposed mechanism for protein denaturation in scCO2 which states that the changing of the interaction pattern of protein destabilizes the conformation in supercritical CO2. However, the extent of changes was different among the force fields. While united-atom force field show higher changes and deviations from the native state, all-atom force fields, in agreement with each other, show moderate changes and deviations.

Fast and effective protein model refinement using deep graph neural networks

Protein model refinement is the last step applied to improve the quality of a predicted protein model. Currently the most successful refinement methods rely on extensive conformational sampling and thus, take hours or days to refine even a single protein model. Here we propose a fast and effective model refinement method that applies GNN (graph neural networks) to predict refined inter-atom distance probability distribution from an initial model and then rebuilds 3D models from the predicted distance distribution. Tested on the CASP (Critical Assessment of Structure Prediction) refinement targets, our method has comparable accuracy as two leading human groups Feig and Baker, but runs substantially faster. Our method may refine one protein model within ~11 minutes on 1 CPU while Baker needs ~30 hours on 60 CPUs and Feig needs ~16 hours on 1 GPU. Finally, our study shows that GNN outperforms ResNet (convolutional residual neural networks) for model refinement when very limited conformational sampling is allowed.

A human stem cell phenotypic drug screening assay for identification of opioid and non‐opioid analgesics with lowered risk of addiction

Mu opioid receptors, a subclass of G protein-coupled receptors (GPCRs), are involved in the control of pain perception and reward properties. They are also the primary molecular target of opioid drugs and mediate their efficacious analgesic effects, euphoric properties, as well as undesired negative side effects such as tolerance and physical dependence. The biology of pain is complex and has considerable redundancy, thus screens that observe the effect of a single protein, cell lineage, or animal model(s) may be fated for failure. In an effort to find analgesics with lessened risk of addiction we designed Biological Assay for Next Generation Analgesics (BANGA). BANGA is a phenotypic approach using human stem cells (hESC, hiPSC, and/or directly converted neurons) to generate human neurons involved in pain signaling such as sensory neurons (pain), and dopaminergic neurons (reward and addiction) that together with other neurons and glia, used for 3D printed tissue organoids, authentically represent Dorsal Root Ganglion (DRG) and Ventral Tegmental Area (VTA), in different “banks” of drug screening assays. High throughput calcium imaging of human nociceptors, in the DRG organoids, quantitates sensory pain input and its dose response inhibition by candidate analgesic drugs. To reveal addiction risk, dopaminergic neurons’ quantitative secretion of dopamine, in the VTA organoids, measures analgesic drugs’ level of risk for dependence and addiction. All reprogrammed and directly converted neurons and glia used in this assay are patient specific, making studies of pain and addiction patient cohorts possible. The BANGA assay is a dramatic shift from the commonly used models for anti-pain drug discovery and could revolutionize the discovery of the next generation analgesics. Our presentation will include results of BANGA screening of commercially available FDA approved compound libraries that has so far identified novel classes of opioid and non-opioid efficacious analgesic drugs with low to no risk of dependence and addiction.

Methodological considerations for identifying multiple plasma proteins associated with all-cause mortality in a population-based prospective cohort

Novel methods to characterize the plasma proteome has made it possible to examine a wide range of proteins in large longitudinal cohort studies, but the complexity of the human proteome makes it difficult to identify robust protein-disease associations. Nevertheless, identification of individuals at high risk of early mortality is a central issue in clinical decision making and novel biomarkers may be useful to improve risk stratification. With adjustment for established risk factors, we examined the associations between 138 plasma proteins measured using two proximity extension assays and long-term risk of all-cause mortality in 3,918 participants of the population-based Malmö Diet and Cancer Study. To examine the reproducibility of protein-mortality associations we used a two-step random-split approach to simulate a discovery and replication cohort and conducted analyses using four different methods: Cox regression, stepwise Cox regression, Lasso-Cox regression, and random survival forest (RSF). In the total study population, we identified eight proteins that associated with all-cause mortality after adjustment for established risk factors and with Bonferroni correction for multiple testing. In the two-step analyses, the number of proteins selected for model inclusion in both random samples ranged from 6 to 21 depending on the method used. However, only three proteins were consistently included in both samples across all four methods (growth/differentiation factor-15 (GDF-15), N-terminal pro-B-type natriuretic peptide, and epididymal secretory protein E4). Using the total study population, the C-statistic for a model including established risk factors was 0.7222 and increased to 0.7284 with inclusion of the most predictive protein (GDF-15; P < 0.0001). All multiple protein models showed additional improvement in the C-statistic compared to the single protein model (all P < 0.0001). We identified several plasma proteins associated with increased risk of all-cause mortality independently of established risk factors. Further investigation into the putatively causal role of these proteins for longevity is needed. In addition, the examined methods for identifying multiple proteins showed tendencies for overfitting by including several putatively false positive findings. Thus, the reproducibility of findings using such approaches may be limited.

The effects of mutation and modification on the structure and stability of human lysozyme: A molecular link between carbamylation and atherosclerosis

Amino acid mutations in some proteins such as lysozyme lead to genetically disorder variants and adverse pathogenic consequences. Recently, amino acid modifications were known as a risk factor in many related diseases such as uremia and atherosclerosis, showing the importance of these surface-structure changes. Although the structural consequences of the hereditary proteins have been examined extensively, such effects for the protein modifications are known to a lesser extent. One drawback in the examination of protein modifications is hardness in experimental detection of modifications by techniques such as NMR and crystallography. Molecular modeling and simulation can help to understand such phenomena at the molecular levels. It is more rational that the effects of both mutation and modification can be compared in a single protein model. Here, molecular dynamics simulation is used to compare the effects of a disease-related carbamylation modification and an amyloidogenic mutation (D67H) in human lysozyme as a model protein. The results show that the carbamylation adversely effects on the tertiary structure, leading to the similar unfolding pathway to the hereditary amyloidogenic form. The carbamylation leads to the instability of the overall protein conformation, especially on the β-domain, which is a characteristic of hereditary amyloidosis in human lysozymes. The aggregation behaviors of both modified and mutant lysozyme were examined by molecular docking calculations. The results showed that the partially unfolded lysozyme might form tight protein aggregates upon carbamylation similar to the amyloidogenic variant. Both single and all-residues carbamylations impose serious conformational changes to the tertiary structure of lysozyme. It was obtained that carbamylation of lysozyme strongly effects on the stability of N-terminal β-sheet, which can produce a highly unstable conformation. The results of this study not only show the adverse structural consequences of a disease-associated post-translational modification, but it also may be very helpful to understand the molecular basis for many carbamylation-related diseases such as atherosclerosis in ESRD patients. The results show that non-native post-translational modifications may be as structurally important as hereditary mutations.

Predicting Drug Binding to Human Serum Albumin and Alpha One Acid Glycoprotein in Diseased and Age Patient Populations

Plasma protein binding, namely the fraction unbound (fu), can be an important determinant of the disposition and response of drugs. The primary objective of this study was to predict fu values of 183 drugs utilizing either a single binding protein model, where the predominant binding protein had been established, or a multiple binding protein model (MBPM), where the relative binding contribution of human serum albumin (HSA) or alpha 1 acid glycoprotein (AAG) is known. Mean protein concentrations, dependent on disease or age, were used to account for changes in fu. A simple scaling approach for binding protein concentration was employed to account for quantitative changes in molar concentrations of either HSA or AAG in their respective conditions. The MBPM predictive model works best if the relative binding contribution of HSA and AAG is known, and a scaler for the change in protein concentration can be adjusted accordingly. The value of MBPM was most evident when considering reported changes in lidocaine binding because of increasing AAG concentration in response to trauma. The present approach enhances the ability to predict fu in diseased and age populations because of quantitative changes in major binding proteins.

CABS-flex 2.0: a web server for fast simulations of flexibility of protein structures.

Classical simulations of protein flexibility remain computationally expensive, especially for large proteins. A few years ago, we developed a fast method for predicting protein structure fluctuations that uses a single protein model as the input. The method has been made available as the CABS-flex web server and applied in numerous studies of protein structure-function relationships. Here, we present a major update of the CABS-flex web server to version 2.0. The new features include: extension of the method to significantly larger and multimeric proteins, customizable distance restraints and simulation parameters, contact maps and a new, enhanced web server interface. CABS-flex 2.0 is freely available at http://biocomp.chem.uw.edu.pl/CABSflex2.

Arginine prevents thermal aggregation of hen egg white proteins

The control of aggregation and solubilization of hen egg white protein (HEWP) is an important issue for industrial applications of one of the most familiar food protein sources. Here, we investigated the effects of edible amino acids on heat-induced aggregation of HEWP. The addition of 0.6M arginine (Arg) completely suppressed the formation of insoluble aggregates of 1mgmL−1 HEWP following heat treatment, even at 90°C for 20min. In contrast, lysine (Lys), glycine (Gly), and sodium chloride (NaCl) did little to suppress the aggregation of HEWP under the same conditions. SDS-PAGE indicated that Arg suppresses the thermal aggregation of almost all types of HEWP at 1mgmL−1. However, Arg did not suppress the thermal aggregation of HEWP at concentrations ≥10mgmL−1 and prompted the formation of aggregates. Transmission electron micrographs revealed a high-density structure of unfolded proteins in the presence of Arg. These results indicate that Arg exerts a greater suppressive effect on a protein mixture, such as HEWP, than on a single model protein. These observations may propose Arg as a safe and reasonable additive to HEWP for the elimination of microorganisms by allowing an increase in sterilization temperature.

DeepQA: improving the estimation of single protein model quality with deep belief networks.

BackgroundProtein quality assessment (QA) useful for ranking and selecting protein models has long been viewed as one of the major challenges for protein tertiary structure prediction. Especially, estimating the quality of a single protein model, which is important for selecting a few good models out of a large model pool consisting of mostly low-quality models, is still a largely unsolved problem.ResultsWe introduce a novel single-model quality assessment method DeepQA based on deep belief network that utilizes a number of selected features describing the quality of a model from different perspectives, such as energy, physio-chemical characteristics, and structural information. The deep belief network is trained on several large datasets consisting of models from the Critical Assessment of Protein Structure Prediction (CASP) experiments, several publicly available datasets, and models generated by our in-house ab initio method. Our experiments demonstrate that deep belief network has better performance compared to Support Vector Machines and Neural Networks on the protein model quality assessment problem, and our method DeepQA achieves the state-of-the-art performance on CASP11 dataset. It also outperformed two well-established methods in selecting good outlier models from a large set of models of mostly low quality generated by ab initio modeling methods.ConclusionDeepQA is a useful deep learning tool for protein single model quality assessment and protein structure prediction. The source code, executable, document and training/test datasets of DeepQA for Linux is freely available to non-commercial users at http://cactus.rnet.missouri.edu/DeepQA/.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-1405-y) contains supplementary material, which is available to authorized users.

QAcon: single model quality assessment using protein structural and contact information with machine learning techniques.

Protein model quality assessment (QA) plays a very important role in protein structure prediction. It can be divided into two groups of methods: single model and consensus QA method. The consensus QA methods may fail when there is a large portion of low quality models in the model pool. In this paper, we develop a novel single-model quality assessment method QAcon utilizing structural features, physicochemical properties, and residue contact predictions. We apply residue-residue contact information predicted by two protein contact prediction methods PSICOV and DNcon to generate a new score as feature for quality assessment. This novel feature and other 11 features are used as input to train a two-layer neural network on CASP9 datasets to predict the quality of a single protein model. We blindly benchmarked our method QAcon on CASP11 dataset as the MULTICOM-CLUSTER server. Based on the evaluation, our method is ranked as one of the top single model QA methods. The good performance of the features based on contact prediction illustrates the value of using contact information in protein quality assessment. The web server and the source code of QAcon are freely available at: http://cactus.rnet.missouri.edu/QAcon. chengji@missouri.edu. Supplementary data are available at Bioinformatics online.

The pivotal role of reactivity in the design of novel biotinylation reagents for the chemical-proteomics-based identification of vascular accessible biomarkers

A promising approach for the development of novel therapeutics with fewer side effects in healthy tissues is the targeted delivery of bioactive molecules directly to the site of disease. Thus, one prerequisite is the identification of a robust, disease-specific, vascular accessible biomarker localized on the surface of diseased cells, in the surrounding extracellular matrix or on newly formed blood vessels. One avenue towards the identification of such biomarkers consists in the enrichment of the vascular accessible surface proteome fraction prior to analysis. This can be achieved by covalent modification of the target proteins with membrane-impermeable ester derivatives of biotin, followed by streptavidin-based affinity capturing. The properties of the respective reagents are determined by the linker between the biotin moiety and the reactive group for protein coupling. In the frame of this study, novel, reactivity-improved peptide-based biotinylation reagents as well as reagents based on highly hydrophilic heparin linkers were synthesized and validated. The comprehensive evaluation of different biotinylation reagent classes with aliphatic, PEGylated, peptide-based and heparin-based linkers on single model protein BSA, HeLa cells as well as perfused kidney tissue revealed that the linker-dependent chemical reactivity is the crucial factor for the design of novel biotinylation reagents for vascular targeting approaches. SignificanceTo obtain a reliable identification and stable quantification of vascular accessible protein targets by means of mass spectrometry, covalent modification with a membrane-impermeable ester derivative of biotin, followed by streptavidin-based affinity capturing, is frequently applied for in vivo or ex vivo biomarker identification studies. Nevertheless, no comprehensive evaluation of different biotinylation reagent classes has been performed so far. Within this study, we systematically evaluated novel peptide- and heparin-based biotinylation reagents as well as established compounds based on aliphatic and PEGylated linkers. We identified the linker-dependant chemical reactivity of biotinylation reagents to be the critical factor for the design of novel reagents with high efficiency. The novel, site-specifically activated peptide-based reagents were found to be efficient compounds for application in mass spectrometry-based discovery of novel vascular-accessible biomarkers.

Chemo- and Regioselective Ethynylation of Tryptophan-Containing Peptides and Proteins.

Ethynylation of various tryptophan-containing peptides and a single model protein was achieved using Waser's reagent, 1-[(triisopropylsilyl)ethynyl]-1,2-benziodoxol-3(1 H)-one (TIPS-EBX), under gold(I) catalysis. It was demonstrated by NMR that the ethynylation occurred selectively at the C2-position of the indole ring of tryptophan. Further, MS/MS showed that the tryptophan residues could be modified selectively with ethynyl functionalities even when the tryptophan was present as a part of the protein. Finally, the terminal alkyne was used to label a model peptide with a fluorophore by means of copper-catalyzed click chemistry.

EXPLORING THE CORRELATION BETWEEN INTRINSIC FLUORESCENCE AND PROTEIN AGGREGATION

Progressive nerve and brain cell damage associated with Alzheimer's disease, has been attributed to the build‐up of aggregated proteins. Detection of amyloid aggregation, through the measuring their intrinsic fluorescence, could be a valuable approach for early detection of the disease.The purpose of our study was to teach the correlation between intrinsic fluorescence and protein aggregation within the visible range in an interdisciplinary undergraduate summer research course. Students used purified green fluorescent protein (GFP, Biorad), unpurified enhanced green fluorescent protein (EGFP, Edvotek) and liquid silk fibroin protein (SFP, Silktap) as model proteins. We focused our experimentation on sonication conditions that led to gelation of SFP due to β‐sheet formation. Using single model proteins and combined protein solutions, samples were sonicated and fluorescence was measured at 405 and 500nm wavelengths. UV light was also used to compare the green fluorescence emitted by the GFP and EGFP samples.Our results showed an increase in fluorescence of single model protein solutions post‐sonication when compared with the pre‐sonication protein solutions. When EGFP and SFP solutions were combined and sonicated together, we observed a decrease in fluorescence. Similar results were obtained when purified GFP and SFP solutions were combined and sonicated.In future experiments, students are planning to design a minimally invasive device able to detect early stages of amyloid aggregation in the brain as a senior project for an undergraduate biomedical engineering major.

CABS-flex predictions of protein flexibility compared with NMR ensembles.

Motivation: Identification of flexible regions of protein structures is important for understanding of their biological functions. Recently, we have developed a fast approach for predicting protein structure fluctuations from a single protein model: the CABS-flex. CABS-flex was shown to be an efficient alternative to conventional all-atom molecular dynamics (MD). In this work, we evaluate CABS-flex and MD predictions by comparison with protein structural variations within NMR ensembles.Results: Based on a benchmark set of 140 proteins, we show that the relative fluctuations of protein residues obtained from CABS-flex are well correlated to those of NMR ensembles. On average, this correlation is stronger than that between MD and NMR ensembles. In conclusion, CABS-flex is useful and complementary to MD in predicting protein regions that undergo conformational changes as well as the extent of such changes.Availability and implementation: The CABS-flex is freely available to all users at http://biocomp.chem.uw.edu.pl/CABSflex.Contact: sekmi@chem.uw.edu.plSupplementary information: Supplementary data are available at Bioinformatics online.

SMOQ: a tool for predicting the absolute residue-specific quality of a single protein model with support vector machines.

BackgroundIt is important to predict the quality of a protein structural model before its native structure is known. The method that can predict the absolute local quality of individual residues in a single protein model is rare, yet particularly needed for using, ranking and refining protein models.ResultsWe developed a machine learning tool (SMOQ) that can predict the distance deviation of each residue in a single protein model. SMOQ uses support vector machines (SVM) with protein sequence and structural features (i.e. basic feature set), including amino acid sequence, secondary structures, solvent accessibilities, and residue-residue contacts to make predictions. We also trained a SVM model with two new additional features (profiles and SOV scores) on 20 CASP8 targets and found that including them can only improve the performance when real deviations between native and model are higher than 5Å. The SMOQ tool finally released uses the basic feature set trained on 85 CASP8 targets. Moreover, SMOQ implemented a way to convert predicted local quality scores into a global quality score. SMOQ was tested on the 84 CASP9 single-domain targets. The average difference between the residue-specific distance deviation predicted by our method and the actual distance deviation on the test data is 2.637Å. The global quality prediction accuracy of the tool is comparable to other good tools on the same benchmark.ConclusionSMOQ is a useful tool for protein single model quality assessment. Its source code and executable are available at: http://sysbio.rnet.missouri.edu/multicom_toolbox/.

Designing and evaluating the MULTICOM protein local and global model quality prediction methods in the CASP10 experiment.

BackgroundProtein model quality assessment is an essential component of generating and using protein structural models. During the Tenth Critical Assessment of Techniques for Protein Structure Prediction (CASP10), we developed and tested four automated methods (MULTICOM-REFINE, MULTICOM-CLUSTER, MULTICOM-NOVEL, and MULTICOM-CONSTRUCT) that predicted both local and global quality of protein structural models.ResultsMULTICOM-REFINE was a clustering approach that used the average pairwise structural similarity between models to measure the global quality and the average Euclidean distance between a model and several top ranked models to measure the local quality. MULTICOM-CLUSTER and MULTICOM-NOVEL were two new support vector machine-based methods of predicting both the local and global quality of a single protein model. MULTICOM-CONSTRUCT was a new weighted pairwise model comparison (clustering) method that used the weighted average similarity between models in a pool to measure the global model quality. Our experiments showed that the pairwise model assessment methods worked better when a large portion of models in the pool were of good quality, whereas single-model quality assessment methods performed better on some hard targets when only a small portion of models in the pool were of reasonable quality.ConclusionsSince digging out a few good models from a large pool of low-quality models is a major challenge in protein structure prediction, single model quality assessment methods appear to be poised to make important contributions to protein structure modeling. The other interesting finding was that single-model quality assessment scores could be used to weight the models by the consensus pairwise model comparison method to improve its accuracy.