SEARCH · PubMed Health

Results for “predictive modelling”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,531 records · Page 85Linked to original sources

PubMed1999

kPROT: a knowledge-based scale for the propensity of residue orientation in transmembrane segments. Application to membrane protein structure prediction.

Modeling of integral membrane proteins and the prediction of their functional sites requires the identification of transmembrane (TM) segments and the determination of their angular orientations. Hydrophobicity scales predict accurately the location of TM helices, but are less accurate in computing angular disposition. Estimating lipid-exposure propensities of the residues from statistics of solved membrane protein structures has the disadvantage of relying on relatively few proteins. As an alternative, we propose here a scale of knowledge-based Propensities for Residue Orientation in Transmembrane segments (kPROT), derived from the analysis of more than 5000 non-redundant protein sequences. We assume that residues that tend to be exposed to the membrane are more frequent in TM segments of single-span proteins, while residues that prefer to be buried in the transmembrane bundle interior are present mainly in multi-span TMs. The kPROT value for each residue is thus defined as the logarithm of the ratio of its proportions in single and multiple TM spans. The scale is refined further by defining it for three discrete sections of the TM segment; namely, extracellular, central, and intracellular. The capacity of the kPROT scale to predict angular helical orientation was compared to that of alternative methods in a benchmark test, using a diversity of multi-span alpha-helical transmembrane proteins with a solved 3D structure. kPROT yielded an average angular error of 41 degrees, significantly lower than that of alternative scales (62 degrees -68 degrees ). The new scale thus provides a useful general tool for modeling and prediction of functional residues in membrane proteins. A WWW server (http://bioinfo.weizmann.ac.il/kPROT) is available for automatic helix orientation prediction with kPROT.

Amino Acid Sequence↗

PubMed2003

Determinism in a transient assemblage: the roles of dispersal and local competition.

Both dispersal and local competitive ability may determine the outcome of competition among species that cannot coexist locally. I develop a spatially implicit model of two-species competition at a small spatial scale. The model predicts the relative fitness of two competitors based on local reproductive rates and regional dispersal rates in the context of the number, size, and extinction probability of habitat patches in the landscape. I test the predictions of this model experimentally using two genotypes of the bacteriophagous soil nematode Caenorhabditis elegans in patchy microcosms. One genotype has higher fecundity while the other is a better disperser. With such a fecundity-dispersal trade-off between competitors, the model predicts that relative fitness will be affected most by local population size when patches do not go extinct and by the number of patches when there is a high probability of patch extinction. The microcosm experiments support the model predictions. Both approaches suggest that competitive dominance in a patchily distributed transient assemblage will depend upon the architecture and predictability of the environment. These mechanisms, operating at a small scale with high spatial admixture, may be embedded in a larger metacommunity process.

Animals↗

PubMed1999

Re-evaluation of the oxygen diffusion model for predicting minimum contact lens Dk/t values needed to avoid corneal anoxia.

PURPOSE: (1) To update Fatt's mathematical model of the distribution of oxygen tension (pO2) across the cornea and contact lens (CL) to include the recent finding that corneal oxygen consumption increases with the acidification that occurs with CL wear. (2) To estimate the minimum transmissibility (CL Dk/t) to avoid epithelial anoxia or to avoid stromal anoxia. METHODS: A five-layer static and one-dimensional mathematical model of oxygen diffusion through the cornea based on Fatt's models was used. The relationships between acidosis and increased QO2, and acidosis and CL Dk/t were used to estimate corneal QO2 for a given CL Dk/t. RESULTS: (1) Revised model predictions are in agreement with direct tear pO2 measurements beneath CLs in the rabbit. (2) For the human eye, the minimum CL Dk/t for oxygen delivery to the basal epithelial cells was determined to be 23 for the open eye and 89 for the closed eye. To prevent anoxia throughout the entire corneal thickness the Dk/t requirements are 35 for the open eye and 125 for the closed eye. CONCLUSIONS: (1) Model predictions of the oxygen distribution beneath contact lenses are significantly lower than previous models that did not include the effect of acidosis on corneal QO2. (2) Minimum Dk/t values that allow oxygen delivery to the basal epithelium are in agreement with the Dk/t needed to avoid corneal edema.

Acidosis↗

PubMed1993

CO2 homeostasis during periodic breathing: predictions from a computer model.

The Bohr/Riley model of CO2 homeostasis describes the relationship between CO2 production, ventilation, and arterial PCO2 and assumes that ventilation and CO2 delivery to the lung are both anatomically and temporally well matched. In contrast to normal breathing, periodic patterns of ventilation show temporal mismatch of ventilation to CO2 delivery. We developed a computer model of lung CO2 clearance that uses CO2 transfer equations to generate iterative solutions for PCO2 in multiple body compartments as a function of time. During continuous ventilatory patterns our model predicts steady-state arterial PCO2 identical to that of the Bohr model. During periodic ventilation, we predict mean PCO2 will be elevated unless mean ventilation is increased above that required by the Bohr model. Waxing and waning tidal volumes, low functional residual capacity, and low capillary blood volume potentiate the hypercapnia. However, if cardiac output oscillates in phase with breathing, hypercapnia is minimized. This analysis suggests a new mechanism for the development of sustained hypercapnia, separate from absolute hypoventilation or the presence of lung disease.

Carbon Dioxide↗

PubMed2006

Development and evaluation of models to predict the feed intake of dairy cows in early lactation.

Inaccurate prediction of dry matter intake (DMI) limits the ability of current models to anticipate the technical and economic consequences of adopting different strategies for production management on individual dairy farms. The objective of the present study was to develop an accurate, robust, and broadly applicable prediction model and to compare it with the current NRC model for dairy cows in early lactation. Among various functions, an exponential model was selected for its best fit to DMI data of dairy cows in early lactation. Daily DMI data (n = 8,547) for 3 groups of Holstein cows (at Illinois, New Hampshire, and Pennsylvania) were used in this study. Cows at Illinois and New Hampshire were fed totally mixed diets for the first 70 d of lactation. At Pennsylvania, data were for the first 63 d postpartum. Data from Illinois cows were used as the developmental dataset, and the other 2 datasets were used for model evaluation and validation. Data for BW, milk yield, and milk composition were only available for Illinois and New Hampshire cows; therefore, only these 2 datasets were used for model comparisons. The exponential model, fitted to the individual cow daily DMI data, explained an average of 74% of the total variation in daily DMI for Illinois data, 49% of the variation for New Hampshire data, 67% of the variation for Pennsylvania data, and 64% of the variation overall. Based on all model selection criteria used in this study, the exponential model for prediction of weekly DMI of individual cows was superior to the current NRC equation. The exponential model explained 85% of the variation in weekly mean DMI compared with 42% for the NRC equation. Compared with the relative prediction error of 6% for the exponential model, that associated with prediction using the NRC equation was 14%. The overall mean square prediction error value for individual cows was 5-fold higher for the NRC equation than for the exponential model (10.4 vs. 2.0 kg2/d2). The consistently accurate and robust prediction of DMI by the exponential model for all data-sets suggested that it could safely be used for predicting DMI in many circumstances.

Animal Nutritional Physiological Phenomena↗

PubMed2002

Development of models to predict dose of pesticides in professional turf applicators.

Epidemiologic studies designed to assess the chronic effects of pesticides are limited by inadequate measurements of exposures. Although cohort studies have been initiated to evaluate the effects of 2,4-dichlorophenoxyacetic acid (2,4-D) and other pesticides in professional turf applicators, they may have limited power to detect significant health risks and may be subject to bias from exposure measurement error. In this study, the doses of 2,4-D, mecoprop [2-(4-chloro-2 methylphenoxy) propionic acid, MCPP] and dicamba (3,6-dichloro-o-anisic acid) were evaluated in a group of 98 professional turf applicators from 20 companies across southwestern Ontario. During a 1-week period (Saturday to Thursday), the volume of pesticide (active ingredient) applied was only weakly related to the total dose of 2,4-D absorbed (R(2)=0.21). Two additional factors explained a large proportion of variation in dose: the type of spray nozzle used and the use of gloves while spraying. Individuals who used a fan-type nozzle had significantly higher doses than those who used a gun-type nozzle. Glove use was associated with significantly lower doses. Job satisfaction and current smoking influenced the dose but were not highly predictive. In the final multiple regression models predicting total absorbed dose of 2,4-D and mecoprop, approximately 63-68% of the variation was explained. The future application of these models for epidemiologic research will depend on the availability of information and records from employers, the feasibility of contacting study subjects and cost.

2-Methyl-4-chlorophenoxyacetic Acid↗

PubMed2000

The use of clinical prediction formulas in the evaluation of obstructive sleep apnea.

STUDY OBJECTIVES: To prospectively study the utility of four clinical prediction models for either predicting the presence of obstructive sleep apnea (OSA, apnea-hypopnea index [AHI] > or = 10/hour), or prioritizing patients for a split-night protocol (AHI(3)20/hour). DESIGN: All patients presenting for OSA evaluation completed a research questionnaire that included questions from previously developed clinical prediction models. The probability of sleep apnea for each patient for each model was calculated based upon the equation used in the model. Based upon two cutoffs of apnea-hypopnea index, 10 and 20, the sensitivity, specificity, and positive predictive value were calculated. For the cutoffs AHI > or =10 and > or =20, receiver operating characteristic curves were generated and the areas under the curves calculated. Comparisons of demographic information and symptom response were compared between patients with and without OSA, and men vs. women. SETTING: Urban, accredited sleep disorders center. PATIENTS OR PARTICIPANTS: All patients referred for evaluation of OSA who underwent polysomnography. INTERVENTIONS: N/A. RESULTS: 370 patients (191 men, 179 women) completed the study. 248 of the 370 (67%) patients had an AHI(3)10; 180 of the 370 (49%) had an AHI> or =20. For AHI > or =10, the sensitivities ranged from 76 to 96%, specificities from 13%-54%, positive predictive values from 69%-77% using the probability cutoff of the original investigators; the areas under the curve from 0.669 to 0.736. For AHI(3)20, the areas under the ROC curves ranged from 0.700 to 0.757; using cutoffs to maximized specificity, the sensitivities ranged from 33%-39%, specificities from 87%-93%, and positive predictive values from 72%-85%. All the models performed better for men. CONCLUSIONS: The clinical prediction models tested are not be sufficiently accurate to discriminate between patients with or without OSA but could be useful in prioritizing patients for split-night polysomnography.

Adult↗

PubMed2006

A simplified clinical model to predict pulmonary embolism in patients with acute dyspnea.

UNLABELLED: The objective of the present study was to develop a simple clinical model for predicting pulmonary embolism (PE) in patients with acute dyspnea in the emergency room. PATIENTS AND MEASUREMENTS: We enrolled 56 patients diagnosed with PE, and 92 consecutive patients without PE, all of whom presented with acute dyspnea in the emergency room. Primary emergency-room physicians assessed the initial evaluation and interpretation of various laboratory findings. Some significantly independent predictors of PE were identified and integrated into a clinical model of pretest probability: low (< 30%), intermediate (>or= 30%, 70%). After setting up the model, another 40 patients (16 with PE, 24 without PE) were tested using the pretest model. Clinical variables associated with an increased likelihood of PE were being female and having unilateral low-leg edema, a high alveolar-arterial oxygen gradient, a clear chest x-ray, and electrocardiographic findings of right ventricular strain. Variables associated with a decreased likelihood of PE were cough, chest tightness, and unclear breath sounds. Our clinical model predicted that 95% of patients with PE had a high or low probability of PE. The positive predictive value for high probability was 94.1% and the negative predictive value for low probability was 94.4%. In the tested group, the positive predictive value for high probability was 92.9%. The negative predictive value for low probability was 91.3%. This simple and easily available prediction model was useful for estimating the pretest probability of PE in patients with acute dyspnea.

Acute Disease↗

PubMed1994

Using computer-based models for predicting human thermal responses to hot and cold environments.

Four influential models, capable of predicting human responses to hot and cold environments and potentially suitable for use in practical applications, were evaluated by comparing their predictions with human data published previously. The models were versions of the Pierce Lab 2-node and Stolwijk and Hardy 25-node models of human thermoregulation, the Givoni and Goldman model of rectal temperature response, and ISO/DIS 7933. Experimental data were available for a wide range of environmental conditions, with air temperatures ranging from -10 to 50 degrees C, and with different levels of air movement, humidity, clothing and work. The experimental data were grouped into environment categories to allow examination of the effects of variables, such as wind or clothing, on the accuracy of the models' predictions. This categorization also enables advice to be given regarding which model is likely to provide the most accurate predictions for a particular combination of environmental conditions. Usually at least one of the models was able to give predictions with an accuracy comparable with the degree of variation that occurred within the data from the human subjects. The evaluation suggests that it is possible to make useful predictions of deep-body and mean skin temperature responses to cool, neutral, warm and hot environmental conditions. The models' predictions of deep-body temperature in the cold were poor. Overall, the 25-node model provided the most consistently accurate predictions. The 2-node model was often accurate but could be poor for exercise conditions. The rectal-temperature model usually overestimated deep-body temperature, although its predictions for very hot or heavy exercise conditions could be useful. The ISO model's allowable exposure times would not have protected subjects for some exercise conditions.

Body Temperature Regulation↗

PubMed2006

In silico ADME modelling 2: computational models to predict human serum albumin binding affinity using ant colony systems.

Modelling of in vitro human serum albumin (HSA) binding data of 94 diverse drugs and drug-like compounds is performed to develop global predictive models that are applicable to the whole medicinal chemistry space. For this aim, ant colony systems, a stochastic method along with multiple linear regression (MLR), is employed to exhaustively search and select multivariate linear equations, from a pool of 327 molecular descriptors. This methodology helped us to derive optimal quantitative structure-property relationship (QSPR) models based on five and six descriptors with excellent predictive power. The best five-descriptor model is based on Kier and Hall valence connectivity index--Order 5 (path), Auto-correlation descriptor (Broto-Moreau) weighted by atomic masses--Order 4, Auto-correlation descriptor (Broto-Moreau) weighted by atomic polarizabilities--Order 5, AlogP98, SklogS (calculated buffer water solubility) [R=0.8942, Q=0.86790, F=62.24 and SE=0.2626]; the best six-variable model is based on Kier and Hall valence connectivity index of Order 3 (cluster), Auto-correlation descriptor (Broto-Moreau) weighted by atomic masses--Order 4, Auto-correlation descriptor (Broto-Moreau) weighted by atomic polarizabilities--Order 5, Atomic-Level-Based AI topological descriptors--AIdsCH, AlogP98, SklogS (calculated buffer water solubility) [R=0.9128, Q=0.89220, F=64.09 and SE=0.2411]. From the analysis of the physical meaning of the selected descriptors, it is inferred that the binding affinity of small organic compounds to human serum albumin is principally dependent on the following fundamental properties: (1) hydrophobic interactions, (2) solubility, (3) size and (4) shape. Finally, as the models reported herein are based on computed properties, they appear to be a valuable tool in virtual screening, where selection and prioritisation of candidates is required.

Algorithms↗

PubMed2006

Modeling, identification, and validation of models for predictive ammonia control in a wastewater treatment plant--a case study.

The aim of this work is to develop the ammonia models that could be used for model predictive control (MPC) of nitrification process in a wastewater treatment plant. First, a reduced nonlinear model is presented, which is based on expression for nitrification reaction rate in activated sludge model No. 1 and modified for attached biomass processes, while second, a linear black-box model is shown. The data used for model identification were collected during several weeks of experiments on a real plant so that good identification data were obtained. The designed models were validated based on open loop simulations and predictions. Validation results show that the reduced nonlinear model performs better compared to the linear model, however, both models show relatively large errors compared to the real plant data. Hence, a closed loop simulation study was performed to see the differences between the performance of model predictive controller using previously estimated linear and nonlinear models and a standard proportional integral (PI) controller. From the simulation study results it was seen that in spite of relatively large model errors the MPC algorithms give better results in terms of ammonia removal compared to the PI controller, while MPC with the nonlinear model shows additional improvements over the MPC with the linear model.

Ammonia↗

PubMed2006

Performance of the partition-limited model on predicting ryegrass uptake of polycyclic aromatic hydrocarbons.

Accurate modeling of the uptake and accumulation behavior of organic contaminants like polycyclic aromatic hydrocarbons (PAHs) in plants is essential to assess crop contamination and subsequent human exposure. In this study, the performance of a partition-limited model on predicting ryegrass uptake of PAHs (acenaphthene, fluorene, phenanthrene and pyrene) from water was evaluated and the major factors were examined. It was found that model predictions of PAH concentrations in roots and shoots of ryegrass were all within an order of magnitude of the observed values with the differences between estimated and measured concentrations less than 42.1% for roots and 78.4% for shoots. Since the model considered soil/water-plant pathway only, it was inevitable that simulated concentrations in shoots suffered a bigger error than those in roots due to the influence of foliar uptake, the other important pathway for PAHs. If the impact of foliar uptake was excluded, the accuracy of simulated shoot concentrations would be greatly enhanced, with the maximum prediction error reduced from 78.4% to 47.1% for pyrene. Other factors aside from foliar uptake were also examined, including aqueous PAH concentrations, uptake time, plant composition and chemical properties. These factors were found to influence the model performance generally through acting on the quasi-equilibrium factor (alpha(pt)). Results from this study substantiated the utility of this partition-limited model for vegetation-uptake assessment, and then provided some testimony valuable for the modification of a model with a better performance.

Lolium↗

PubMed2004

Assessment of prediction confidence and domain extrapolation of two structure-activity relationship models for predicting estrogen receptor binding activity.

Quantitative structure-activity relationship (QSAR) methods have been widely applied in drug discovery, lead optimization, toxicity prediction, and regulatory decisions. Despite major advances in algorithms and software, QSAR models have inherent limitations associated with a size and chemical-structure diversity of the training set, experimental error, and many characteristics of structure representation and correlation algorithms. Whereas excellent fit to the training data may be readily attainable, often models fail to predict accurately chemicals that are outside their domain of applicability. A QSAR's utility and, in the case of regulatory decisions, justification for usage increasingly depend on the ability to quantify a model's potential for predicting unknown chemicals with some known degree of certainty. It is never possible to predict an unknown chemical with absolute certainty. Here we report on two QSAR models based on different data sets for classification of chemicals according to their ability to bind to the estrogen receptor. The models were developed by using a novel QSAR method, Decision Forest, which combines the results of multiple heterogeneous but comparable Decision Tree models to produce a consensus prediction. We used an extensive cross-validation process to define an applicability domain for model predictions based on two quantitative measures: prediction confidence and domain extrapolation. Together, these measures quantify the accuracy of each prediction within and outside of the training domain. Despite being based on large and diverse training sets, both QSAR models had poor accuracy for chemicals within the domain of low confidence, whereas good accuracy was obtained for those within the domain of high confidence. For prediction in the high confidence domain, accuracy was inversely proportional to the degree of domain extrapolation. The model with a larger training set of 1,092, compared with 232 for the other, was more accurate in predicting chemicals at larger domain extrapolation, and could be particularly useful for rapidly prioritizing potential endocrine disruptors from large chemical universe.

Animals↗

PubMed2004

Dose-response model for 13 strains of salmonella.

Data from a human feeding trial with healthy men were used to develop a dose-response model for 13 strains of Salmonella and to determine the effects of strain variation on the shape of the dose-response curve. Dose-response data for individual strains were fit to a three-phase linear model to determine minimum, median, and maximum illness doses, which were used to define Pert distributions in a computer simulation model. Pert distributions for illness dose of individual strains were combined in an Excel spreadsheet using a discrete distribution to model strain prevalence. In addition, a discrete distribution was used to model dose groups and thus create a model that simulated human feeding trials. During simulation of the model with @Risk, an illness dose and a dose consumed were randomly assigned to each consumption event in the simulated feeding trial and if the illness dose was greater than the dose consumed then the model predicted no illness, otherwise the model predicted that an illness would occur. To verify the dose-response model predictions, the original feeding trial was simulated. The dose-response model predicted a median of 69 (range of 43-101) illnesses compared to 74 in the original trial. Thus, its predictions were in agreement with the data used to develop it. However, predictions of the model are only valid for eggnog, healthy men, and the strains and doses of Salmonella used to develop it. When multiple strains of Salmonella were simulated together, the predicted dose-response curves were irregular in shape. Thus, the sigmoid shape of dose-response curves in feeding trials with one strain of Salmonella may not accurately reflect dose response in naturally contaminated food where multiple strains may be present.

Colony Count, Microbial↗

PubMed2006

A simple statistical model for prediction of acute coronary syndrome in chest pain patients in the emergency department.

BACKGROUND: Several models for prediction of acute coronary syndrome (ACS) among chest pain patients in the emergency department (ED) have been presented, but many models predict only the likelihood of acute myocardial infarction, or include a large number of variables, which make them less than optimal for implementation at a busy ED. We report here a simple statistical model for ACS prediction that could be used in routine care at a busy ED. METHODS: Multivariable analysis and logistic regression were used on data from 634 ED visits for chest pain. Only data immediately available at patient presentation were used. To make ACS prediction stable and the model useful for personnel inexperienced in electrocardiogram (ECG) reading, simple ECG data suitable for computerized reading were included. RESULTS: Besides ECG, eight variables were found to be important for ACS prediction, and included in the model: age, chest discomfort at presentation, symptom duration and previous hypertension, angina pectoris, AMI, congestive heart failure or PCI/CABG. At an ACS prevalence of 21% and a set sensitivity of 95%, the negative predictive value of the model was 96%. CONCLUSION: The present prediction model, combined with the clinical judgment of ED personnel, could be useful for the early discharge of chest pain patients in populations with a low prevalence of ACS.

Acute Disease↗

PubMed1994

Effect of health education and primary care on diarrhoeal disease morbidity in children: evaluation of a predictive intervention model.

Two groups of infants (intervention group n = 66, and control group n = 45) at risk of persistent diarrhoea (PD) as identified by a predictive model were followed for 12 months. Families were visited at home weekly; mothers in the intervention group received information about prevention and treatment of diarrhoea and were encouraged to seek help in the field station when their children became ill. These children suffered fewer days with any illness, diarrhoea, or respiratory episodes (p < 0.00001 each). Also, in this group, the incidence of PD decreased to the levels of nonselected population only in children whose mothers consulted for the episode of diarrhoea (2.8%). Results show that the intervention decreased the time children suffered diarrhoea and also respiratory and other illnesses. This suggests that the predictive model identified children with high risk but is not disease-specific. Risk appears to be related to maternal behaviours. The model may be useful in the community for detecting groups vulnerable to common paediatric illnesses, including diarrhoeal disease.

Adolescent↗

PubMed2004

Judging trauma center quality: does it depend on the choice of outcomes?

BACKGROUND: Trauma centers routinely benchmark their survival outcomes against a national norm using the TRISS methodology. However, the use of survival as a measure of the effectiveness of trauma care may be too limited in scope because it fails to capture information regarding functional outcomes. METHODS: The objective of this study was to develop a prediction model that allows hospitals to benchmark their functional outcomes in blunt trauma patients, and to determine whether the assessment of hospital "quality" depends on the choice of outcome measure: survival or survival combined with functional outcome. This retrospective cohort study was based on patients, aged 18 years or older, in the National Trauma Database who sustained blunt trauma in 1999 without associated head or spinal cord injury. We developed a sequential logistic model to predict the probability of a good functional outcome. The TRISS methodology was customized to this data set to obtain a survival model. Using each of these prediction models, we then obtained two standardized measures of hospital performance: one based on the number of survivors and the other based on the number of survivors with good functional outcomes. These standardized outcome measures were then used to identify low-performance and high-performance hospitals. The ranking based on these two different measures were compared. RESULTS: Fifteen of the 27 hospitals in the study cohort were categorized differently when their performance was benchmarked using survival versus functional outcome. Kappa analysis revealed minimal agreement between these two quality measures on the identity of hospital quality outliers (kappa = 0.04; p = 0.35). CONCLUSION: The evaluation of hospital quality depends on whether hospital performance is judged by looking at survival or at survival combined with functional outcome. Because functional status is an important outcome of major concern to survivors, it is important to include it in hospital performance assessment. Consideration should be given to including functional outcome in the evaluation of trauma center performance.

Benchmarking↗

PubMed2002

Use of structure comparison methods for the refinement of protein structure predictions. I. Identifying the structural family of a protein from low-resolution models.

Predicting the three-dimensional structure of proteins is still one of the most challenging problems in molecular biology. Despite its difficulty, several investigators have started to produce consistently low-resolution predictions for small proteins. However, in most of these cases, the prediction accuracy is still too low to make them useful. In the present article, we address the problem of obtaining better-quality predictions, starting from low-resolution models. To this end, we have devised a new procedure that uses these models, together with structure comparison methods, to identify the structural family of the target protein. This would allow, in a second step not described in the present work, to refine the predictions using conserved features of the identified family. In our approach, the structure database is investigated using predictions, at different accuracy levels, for a given protein. As query structures, we used both low-resolution versions of the native structures, as well as different sets of low accuracy predictions. In general, we found that for predictions with a resolution of > or =5-7 A, structure comparison methods were able to identify the fold of a protein in the top positions.

Databases, Protein↗