Score Function

In recent years, there is a growing interest about assisted living environments especially for the elderly who live alone, due to the increasing number of aged people. In order for them to live safe and healthy, we need to detect abnormal behavior that may cause severe and emergent situations for the elderly. In this work, we suggest a method that detects abnormal behavior using wireless sensor networks. We model an episode that is a series of events, which includes spatial and temporal information about the subject being monitored. We define a similarity scoring function that compares two episodes taking into consideration temporal aspects. We propose a way to determine a threshold to divide episodes into two groups that reduces wrong classification. Weights on individual functions that consist the similarity function are determined experimentally so that they can produce the good results in terms of area under curve in receiver operating characteristic (ROC) curve.

Efficient processing of top- k queries is a crucial requirement in many interactive environments that involve massive amounts of data. In particular, efficient top- k processing in domains such as the Web, multimedia search, and distributed systems has shown a great impact on performance. In this survey, we describe and classify top- k processing techniques in relational databases. We discuss different design dimensions in the current techniques including query models, data access methods, implementation levels, data and query certainty, and supported scoring functions. We show the implications of each dimension on the design of the underlying techniques. We also discuss top- k queries in XML domain, and show their connections to relational approaches.

A strategy for grafting protein-protein binding sites is described. Firstly, key interaction residues at the interface of ligand protein to be grafted are identified and suitable positions In scaffold protein for grafting these key residues are sought. Secondly, the scaffold proteins are superposed onto the ligand protein based on the corresponding C, and CB atoms. The complementarity between the scaffold protein and the receptor protein is evaluated and only matches with high score are accepted. The relative position between scaffold and receptor proteins is adjusted so that the interface has a reasonable packing density. Then the scaffold protein is mutated to corresponding residues in ligand protein at each candidate position. And the residues having bad steric contacts with the receptor proteins, or buried charged residues not involved in the formation of any salt bridge are mutated. Finally, the mutated scaffold protein in complex with receptor protein is co-minimized by Charmm. In addition, we deduce a scoring function to evaluate the affinity between mutated scaffold protein and receptor protein by statistical analysis of rigid binding data sets.

Mémoire proposes a general framework for reasoning from cases in biology and medicine. Part of this project is to propose a memory organization capable of handling large cases and case bases as occur in biomedical domains. This article presents the essential principles for an efficient memory organization based on pertinent work in information retrieval. Information retrieval systems have been able to scale up to terabyte of data taking advantage of large databases research to build Internet search engines. They search for pertinent documents to answer a query using term-based ranking and/or global ranking schemes. Similarly, CBR systems search for pertinent cases using a scoring function for ranking the cases. Mémoire proposes a memory organization based on inverted indexes, which may be powered by databases to search and rank efficiently through large case bases. It can be seen as a first step toward largescale CBR systems, and in addition provides a framework for tight cooperation between CBR and IR.

Two main languages have been defined to allow users to express their preference criteria: Top-k and Skyline. Top-k ranks the top k tuples in terms of a user-defined score function while Skyline identifies non-dominated tuples, ie such tuples that does not exists a better one in all user criteria. A third language, Top-k Skyline, integrates them. One of the drawbacks of relational engines is that they do not understand the notion of preferences. However, some solutions for Skyline and Top-k queries have been integrated into ...

This paper investigates the influence of different page features on the ranking of search engine results. We use Google (via its API) as our testbed and analyze the result rankings for several queries of different categories using statistical methods. We reformulate the problem of learning the underlying, hidden scores as a binary classification problem. To this problem we then apply both linear and non-linear methods. In all cases, we split the data into a training set and a test set to obtain a meaningful, unbiased estimator for the quality of our predictor. Although our results clearly show that the scoring function cannot be approximated well using only the observed features, we do obtain many interesting insights along the way and discuss ways of obtaining a better estimate and main limitations in trying to do so.

We have developed a new scoring function, the template modeling score (TM-score), to assess the quality of protein structure templates and predicted full-length models by extending the approaches used in Global Distance Test (GDT) 1 and MaxSub. 2 First, a protein size-dependent scale is exploited to eliminate the inherent protein size dependence of the previous scores and appropriately account for random protein structure pairs. Second, rather than setting specific distance cutoffs and calculating only the fractions with errors below the cutoff, all residue pairs in alignment/modeling are evaluated in the proposed score. For comparison of various scoring functions, we have constructed a large-scale benchmark set of structure templates for 1489 small to medium size proteins using the threading program PROSPECTOR_3 and built the full-length models using MODELLER and TASSER. The TM-score of the initial threading alignments, compared to the GDT and MaxSub scoring functions, shows a much stronger correlation to the quality of the final full-length models. The TM-score is further exploited as an assessment of all 'new fold' targets in the recent CASP5 experiment and shows a close coincidence with the results of human-expert visual assessment. These data suggest that the TMscore is a useful complement to the fully automated assessment of protein structure predictions. The executable program of TM-score is freely downloadable at http://bioinformatics.buffalo.edu/TM-score. Proteins 2004;57:702-710.

This paper proposes a regression model where the response is beta distributed using a parameterization of the beta law that is indexed by mean and dispersion parameters. The proposed model is useful for situations where the variable of interest is continuous and restricted to the interval (0, 1) and is related to other variables through a regression structure. The regression parameters of the beta regression model are interpretable in terms of the mean of the response and, when the logit link is used, of an odds ratio, unlike the parameters of a linear regression that employs a transformed response. Estimation is performed by maximum likelihood. We provide closed-form expressions for the score function, for Fisher's information matrix and its inverse. Hypothesis testing is performed using approximations obtained from the asymptotic normality of the maximum likelihood estimator. Some diagnostic measures are introduced. Finally, practical applications that employ real data are presented and discussed.

We construct a zero net-worth uninformed ''naive investor'' who uses a random portfolio allocation strategy. We then compare the returns of the momentum strategist to the return distribution of naive investors. For this purpose we reward momentum profits relative to the return percentiles of the naive investors with scores that are symmetric around the median. The score function thus constructed is invariant and robust to risk factor models. We find that the average scores of the momentum strategies are close to zero (the score of the median) and statistically insignificant over the sample period between 1926 and 2005, various sub-sample periods including the periods examined in Jegadeesh and Titman (1993, 2001). The findings are robust with respect to sampling or period-specific effects, tightened score intervals, and the imposition of maximum-weight restrictions on the naive strategies to mitigate market friction considerations.

A novel and robust automated docking method that predicts the bound conformations of flexible ligands to macromolecular targets has been developed and tested, in combination with a new scoring function that estimates the free energy change upon binding. Interestingly, this method applies a Lamarckian model of genetics, in which environmental adaptations of an individual's phenotype are reverse transcribed into its genotype and become Ž . heritable traits sic . We consider three search methods, Monte Carlo simulated annealing, a traditional genetic algorithm, and the Lamarckian genetic algorithm, and compare their performance in dockings of seven protein᎐ligand test systems having known three-dimensional structure. We show that both the traditional and Lamarckian genetic algorithms can handle ligands with more degrees of freedom than the simulated annealing method used in earlier versions of AUTODOCK, and that the Lamarckian genetic algorithm is the most efficient, reliable, and successful of the three. The empirical free energy function was calibrated using a set of 30 structurally known protein᎐ligand complexes with experimentally determined binding constants. Linear regression analysis of the observed binding constants in terms of a wide variety of structure-derived molecular properties was performed. The final model had a residual standard y1 Ž y1 . error of 9.11 kJ mol 2.177 kcal mol and was chosen as the new energy Abstract: The authors describe the development and testing of a semiempirical free energy force field for use in AutoDock4 and similar grid-based docking methods. The force field is based on a comprehensive thermodynamic model that allows incorporation of intramolecular energies into the predicted free energy of binding. It also incorporates a charge-based method for evaluation of desolvation designed to use a typical set of atom types. The method has been calibrated on a set of 188 diverse protein-ligand complexes of known structure and binding energy, and tested on a set of 100 complexes of ligands with retroviral proteases. The force field shows improvement in redocking simulations over the previous AutoDock3 force field.

This paper considers adaptive estimation in nonstationary autoregressive moving average models with the noise sequence satisfying a generalised autoregressive conditional heteroscedastic process. The locally asymptotic quadratic form of the log-likelihood ratio for the model is obtained. It is shown that the limit experiment is neither LAN nor LAMN, but is instead LABF. Adaptivity is discussed and it is found that the parameters in the model are generally not adaptively estimable if the density of the rescaled error is asymmetric. For the model with symmetric density of the rescaled error, a new efficiency criterion is established for a class of defined M ν -estimators. It is shown that such efficient estimators can be constructed when the density is known. Using the kernel estimator for the score function, adaptive estimators are constructed when the density of the rescaled error is symmetric, and it is shown that the adaptive procedure for the parameters in the conditional mean part uses the full sample without splitting. These estimators are demonstrated to be asymptotically efficient in the class of M ν -estimators. The paper includes the results that the stationary ARMA-GARCH model is LAN, and that the parameters in the model with symmetric density of the rescaled error are adaptively estimable after a reparameterisation of the GARCH process.

ABSTRACT Textual information in the form of digital documents quickly accumulates to create huge amounts of data. The majority of these documents are unstructured: it is unrestricted text and has not been organized into traditional databases. Processing documents is therefore a perfunctory task, mostly due to a lack of standards. It has thus become extremely difficult to implement automatic text analysis tasks. Automatic Text Summarization (ATS), by condensing the text while maintaining relevant information, can help to process this ever-increasing, difficult-to-handle, mass of information. This book examines the motivations and different algorithms for ATS. The author presents the recent state of the art before describing the main problems of ATS, as well as the difficulties and solutions provided by the community. The book provides recent advances in ATS, as well as current applications and trends. The approaches are statistical, linguistic and symbolic. Several examples are also included in order to clarify the theoretical concepts.

This paper considers adaptive estimation in nonstationary autoregressive moving average models with the noise sequence satisfying a generalized autoregressive conditional heteroscedastic process. The locally asymptotic quadratic form of the log-likelihood ratio for the model is obtained. It is shown that the limit experiment is neither LAN nor LAMN, but is instead LABF. For the model with symmetric density of the rescaled error, a new efficiency criterion is established for a class of defined M ν -estimators. It is shown that such efficient estimators can be constructed when the density is known. Using the kernel estimator for the score function, adaptive estimators are constructed when the density of the rescaled error is symmetric, and it is shown that the adaptive procedure for the parameters in the conditional mean part uses the full sample without splitting. These estimators are demonstrated to be asymptotically efficient in the class of M ν -estimators. The paper includes the results that the stationary ARMA-GARCH model is LAN, and that the parameters in the model with symmetric density of the rescaled error are adaptively estimable after a reparameterization of the GARCH process. This paper also establishes the locally asymptotic quadratic form of the loglikelihood ratio for nonlinear time series models with ARCH-type errors.

Efficient processing of top-k queries is a crucial requirement in many interactive environments that involve massive amounts of data. In particular, efficient top-k processing in domains such as the Web, multimedia search and distributed systems has shown a great impact on performance. In this survey, we describe and classify top-k processing techniques in relational databases. We discuss different design dimensions in the current techniques including query models, data access methods, implementation levels, data and query certainty, and supported scoring functions. We show the implications of each dimension on the design of the underlying techniques. We also discuss top-k queries in XML domain, and show their connections to relational approaches.

We have developed an evolutionary approach for flexible ligand docking. This approval, GEMDOCK, uses a Generic Evolutionary Method for molecular DOCKing and an empirical scoring function. The former combines both discrete and continuous global search strategies with local search strategies to speed up convergence, whereas the latter results in rapid recognition of potential ligands. GEMDOCK was tested on a diverse data set of 100 protein-ligand complexes from the Protein Data Bank. In 79% of these complexes, the docked lowest energy ligand structures had root-mean-square derivations (RMSDs) below 2.0 Å with respect to the corresponding crystal structures. The success rate increased to 85% if the structure water molecules were retained. We evaluated GEMDOCK on two cross-docking experiments in which each ligand of a protein ensemble was docked into each protein of the ensemble. Seventysix percent of the docked structures had RMSDs below 2.0 Å when the ligands were docked into foreign structures. We analyzed and validated GEM-DOCK with respect to various search spaces and scoring functions, and found that if the scoring function was perfect, then the predicted accuracy was also essentially perfect. This study suggests that GEMDOCK is a useful tool for molecular recognition and may be used to systematically evaluate and thus improve scoring functions. Proteins 2004; 55:288 -304.

Human purine nucleoside phosphorylase (HsPNP) is a target for inhibitor development aiming at T-cell immune response modulation. In this work, we report the development of a new set of empirical scoring functions and its application to evaluate binding affinities and docking results. To test these new functions, we solved the structure of HsPNP and 2-mercapto-4(3H)-quinazolinone (HsPNP:MQU) binary complex at 2.7 Å resolution using synchrotron radiation, and used these functions to predict ligand position obtained in docking simulations. We also employed molecular dynamics simulations to analyze HsPNP in two conditions, as apoenzyme and in the binary complex form, in order to assess the structural features responsible for stability. Analysis of the structural differences between systems provides explanation for inhibitor binding. The use of these scoring functions to evaluate binding affinities and molecular docking results may be used to guide future efforts on virtual screening focused on HsPNP.

Erythropoietin plus granulocyte colony-stimulating factor is better than erythropoietin alone to treat anemia in low-risk myelodysplastic syndromes: results from a randomized single-centre study Abstract Haemopoietic growth factors (HGF), i.e. erythropoietin [recombinant human erythropoietin (rHEPO)] or granulocyte colony stimulating factor (G-CSF), alone or in combination, have largely been used to treat anemia in myelodysplastic syndromes (MDS), but whether combined rHEPO and G-CSF is really superior to rHEPO alone is still under debate. In particular, randomized studies comparing front-line rHEPO vs rHEPO+G-CSF are still lacking. The aim of this study was to compare the effects of "standard" doses of rHEPO with the combination of rHEPO and G-CSF in the treatment of anemic patients with low-risk MDS in a prospective randomized trial.

We investigate the finite sample properties of the maximum likelihood estimator for the spatial autoregressive model. A stochastic expansion of the score function is used to develop the secondorder bias and mean squared error of the maximum likelihood estimator. We show that the results can be expressed in terms of the expectations of cross products of quadratic forms, or ratios of quadratic forms in a normal vector which can be evaluated using the top order invariant polynomial. Our numerical calculations demonstrate that the second-order behaviors of the maximum likelihood estimator depend on the degree of sparseness of the weights matrix. r

Efficient processing of top-k queries is a crucial requirement in many interactive environments that involve massive amounts of data. In particular, efficient top-k processing in domains such as the Web, multimedia search and distributed systems has shown a great impact on performance. In this survey, we describe and classify top-k processing techniques in relational databases. We discuss different design dimensions in the current techniques including query models, data access methods, implementation levels, data and query certainty, and supported scoring functions. We show the implications of each dimension on the design of the underlying techniques. We also discuss top-k queries in XML domain, and show their connections to relational approaches.

To analyse cross-cultural validity of the Functional Independence Measure (FIM 2) in patients with stroke using the Rasch model. Settings: Thirty-one rehabilitation facilities within 6 different countries in Europe. Participants: A total of 2546 in-patients at admission, median age 63 years. Methods: Data from the FIM 2 were evaluated with the Rasch model, using the Rasch analysis package RUMM2020. A detailed analysis of scoring functions of the 7 categories of the FIM items was undertaken prior to testing fit to the model. Categories were re-scored where necessary. Analysis of Differential Item Functioning was undertaken in pooled data for each of the FIM motor and social-cognitive scales, respectively. Results: Disordered thresholds were found on most items when using 7 categories. Fit to the Rasch model varied between countries. Differential Item Functioning was found by country for most items. Adequate fit to the Rasch model was achieved when items were treated as unique for each country and after a few country-specific items were removed. Conclusion: Clinical collected data from FIM for patients with stroke cannot be pooled in its raw form, or compared across countries. Comparisons can be made after adjusting for country-specific Differential Item Functioning, though the adjustments for Differential Item Functioning and rating scales may not generalize to other samples.

This paper presents an algorithm which is able to extract discriminant rules from oligopeptides for protease proteolytic cleavage activity prediction. The algorithm is developed using genetic programming. Three important components in the algorithm are a min-max scoring function, the reverse Polish notation (RPN) and the use of minimum description length. The min-max scoring function is developed using amino acid similarity matrices for measuring the similarity between an oligopeptide and a rule, which is a complex algebraic equation of amino acids rather than a simple pattern sequence. The Fisher ratio is then calculated on the scoring values using the class label associated with the oligopeptides. The discriminant ability of each rule can therefore be evaluated. The use of RPN makes the evolutionary operations simpler and therefore reduces the computational cost. To prevent overfitting, the concept of minimum description length is used to penalize over-complicated rules. A fitness function is therefore composed of the Fisher ratio and the use of minimum description length for an efficient evolutionary process. In the application to four protease datasets (Trypsin, Factor Xa, Hepatitis C Virus and HIV protease cleavage site prediction), our algorithm is superior to C5, a conventional method for deriving decision trees.

Models of macromolecular assemblies are essential for a mechanistic description of cellular processes. Such models are increasingly obtained by fitting atomic-resolution structures of components into a density map of the whole assembly. Yet, current density-fitting techniques are frequently insufficient for an unambiguous determination of the positions and orientations of all components. Here, we describe MultiFit, a method for simultaneously fitting atomic structures of components into their assembly density map at resolutions as low as 25 Å. The component positions and orientations are optimized with respect to a scoring function that includes the quality-of-fit of components in the map, the protrusion of components from the map envelope, as well as the shape complementarity between pairs of components. The scoring function is optimized by our exact inference optimizer DOMINO that efficiently finds the global minimum in a discrete sampling space. MultiFit was benchmarked on 7 assemblies of known structure, consisting of up to 7 proteins each. The input atomic structures of the components were obtained from the Protein Data Bank as well as by comparative modeling based on 16-99% sequence identity to a template structure. A near-native configuration was usually found as the top-scoring model. Therefore, MultiFit can provide initial configurations for further refinement of many multi-component assembly structures described by electron microscopy.

In this paper, we propose a generic text summarization method that generates summaries of Turkish texts by ranking sentences according to their scores calculated using their surface level features and extracting the highest ranked ones from the original documents. In order to extract sentences which form a summary with an extensive coverage of main content of the text and less redundancy, we use the features such as term frequency, key phrase, centrality, title similarity and position of the sentence in the original text. Sentence rank is computed using a score function that uses its feature values and the weights of the features. The best feature weights are learned using machine learning techniques with the help of human constructed summaries. Performance evaluation is conducted by comparing summarization outputs with manual summaries generated by 25 independent human evaluators. This paper presents one of the first Turkish summarization systems, and its results are promising.

R. Koenker a G. Basset (1978) proposed the regression quantiles as any generalization of usual quantiles to linear regression model. They characterized the regression quantile as the solution of the linear program. called the componenents of the optimal solution of dual problem as the regression rank scores. They showed that many aspects of the duality of order statistics and ranks in the location model generalize naturally to the linear model. proposed some tests based on regression rank scores generated by truncated score functions. A general class of tests based on regression rank scores, parallel to classical rank tests as the Wilcoxon, normal scores and median, was constructed in . The tests of the Kolmogorov-Smirnov type were proposed by Jurečková (1992) and the tests of homoscedasticity in the linear model based regression rank scores were proposed by Gutenbrunner (1994). The concept of regression rank scores was extended in , as autoregression rank scores, to the autoregressive (AR) model. The tests of the linear hypothesis on the AR parameter based on the autoregression rank scores were constructed in and in Hallin and el. (1997). A nonparametric test of independence of two autoregressive time series was considered Hallin and el. (2001). Goodness-of-fit tests in the model with nuisance regression and scale parameters were constructed by Jurečková, .

Bayesian networks are a useful tool in the representation of uncertain knowledge. This paper proposes a new algorithm called ACO-E, to learn the structure of a Bayesian network. It does this by conducting a search through the space of equivalence classes of Bayesian networks using Ant Colony Optimization (ACO). To this end, two novel extensions of traditional ACO techniques are proposed and implemented. Firstly, multiple types of moves are allowed. Secondly, moves can be given in terms of indices that are not based on construction graph nodes. The results of testing show that ACO-E performs better than a greedy search and other state-of-the-art and metaheuristic algorithms whilst searching in the space of equivalence classes.

Virtual screening is becoming an important tool for drug discovery. However, the application of virtual screening has been limited by the lack of accurate scoring functions. Here, we present a novel scoring function, MedusaScore, for evaluating protein-ligand binding. MedusaScore is based on models of physical interactions that include van der Waals, solvation and hydrogen bonding energies. To ensure the best transferability of the scoring function, we do not use any protein-ligand experimental data for parameter training. We then test the MedusaScore for docking decoy recognition and binding affinity prediction and find superior performance compared to other widely used scoring functions. Statistical analysis indicates that one source of inaccuracy of MedusaScore may arise from the unaccounted entropic loss upon ligand binding, which suggests avenues of approach for further MedusaScore improvement.

We propose and evaluate several new scoring rules based on likelihood ratios, for comparing forecast densities in the context of VaR modelling and expected loss estimation. Our approach is motivated by the observation that existing scoring rules tend to favour fat-tailed models when compared with thin-tailed models. Rather than restricting the weight functions, we impose some restrictions on the score functions. Our benchmark case has fixed weights, equal to one in the left tail and zero elsewhere. Two different scoring rules based on partial likelihood are proposed for this extreme case. After generalizing these scoring rules to smooth weight functions, their properties are investigated numerically and illustrated by an empirical application.

The purpose of the paper is to explore the connection between multivariate homogeneity tests and AUC optimization. The latter problem has recently received much attention in the statistical learning literature. From the elementary observation that, in the two-sample problem setup, the null assumption corresponds to the situation where the area under the optimal ROC curve is equal to 1/2, we propose a two-stage testing method based on data splitting. A nearly optimal scoring function in the AUC sense is first learnt from one of the two half-samples. Data from the remaining half-sample are then projected onto the real line and eventually ranked according to the scoring function computed at the first stage. The last step amounts to performing a standard Mann-Whitney Wilcoxon test in the onedimensional framework. We show that the learning step of the procedure does not affect the consistency of the test as well as its properties in terms of power, provided the ranking produced is accurate enough in the AUC sense. The results of a numerical experiment are eventually displayed in order to show the efficiency of the method.

Background: Molecular-docking-based virtual screening is an important tool in drug discovery that is used to significantly reduce the number of possible chemical compounds to be investigated. In addition to the selection of a sound docking strategy with appropriate scoring functions, another technical challenge is to in silico screen millions of compounds in a reasonable time. To meet this challenge, it is necessary to use high performance computing (HPC) platforms and techniques. However, the development of an integrated HPC system that makes efficient use of its elements is not trivial.

We formulate the local ranking problem in the framework of bipartite ranking where the goal is to focus on the best instances. We propose a methodology based on the construction of real-valued scoring functions. We study empirical risk minimization of dedicated statistics which involve empirical quantiles of the scores. We first state the problem of finding the best instances which

The state of the art of various computational aspects of docking-based virtual screening of database of small molecules is presented. The review encompasses the different search algorithms and the scoring functions used in docking methods and their applications to protein and nucleic acid drug targets. Recent progress made in the development and application of methods to include target flexibility are summarized. The fundamental issues and challenges involved in comparing various docking methods are discussed. Limitations of current technologies as well as future prospects are presented.

A wide selection of classical and recent tests for exponentiality are discussed and compared. The classical procedures include the statistics of Kolmogorov-Smirnov and Crame´r-von Mises, a statistic based on spacings, and a method involving the score function. Among the most recent approaches emphasized are methods based on the empirical Laplace transform and the empirical characteristic function, a method based on entropy as well as tests of the Kolmogorov-Smirnov and Crame´r-von Mises type that utilize a characterization of exponentiality via the mean residual life function. We also propose a new goodness-of-fit test utilizing a novel characterization of the exponential distribution through its characteristic function. The finitesample performance of the tests is investigated in an extensive simulation study.

Email Retrieval task has recently taken much attention to help the user retrieve the email(s) related to the submitted query. Up to our knowledge, existing email retrieval ranking approaches sort the retrieved emails based on some heuristic rules, which are either search clues or some predefined user criteria rooted in email fields. Unfortunately, the user usually does not know the effective rule that acquires best ranking related to his query. This paper presents a new email retrieval ranking approach to tackle this problem. It ranks the retrieved emails based on a scoring function that depends on crucial email fields, namely subject, content, and sender. The paper also proposes an architecture to allow every user in a network/group of users to be able, if permissible, to know the most important network senders who are interested in his submitted query words. The experimental evaluation on Enron corpus prove that our approach outperforms known email retrieval ranking approaches.

Relevance ranking consists in sorting a set of objects with respect to a given criterion. However, in personalized retrieval systems, the relevance criteria may usually vary among different users and may not be predefined. In this case, ranking algorithms that adapt their behaviour from users' feedbacks must be devised. Two main approaches are proposed in the literature for learning to rank: the use of a scoring function, learned by examples, that evaluates a feature-based representation of each object yielding an absolute relevance score; a pairwise approach, where a preference function is learned to determine the object that has to be ranked first in a given pair. In this paper, we present a preference learning method for learning to rank. A neural network, the Comparative Neural Network (CmpNN), is trained from examples to approximate the comparison function for a pair of objects. The CmpNN adopts a particular architecture designed to implement the symmetries naturally present in a preference function. The learned preference function can be embedded as the comparator into a classical sorting algorithm to provide a global ranking of a set of objects. To improve the ranking performances, an active-learning procedure is devised, that aims at selecting the most informative patterns in the training set. The proposed algorithm is evaluated on the LETOR dataset showing promising performances in comparison with other state of the art algorithms.

Management of large giant cell tumors of the proximal humerus is controversial because wide resection with reconstruction results in a poor functional outcome for most patients. We retrospectively reviewed the cases of six patients with Campanacci Grade III giant cell tumors of the proximal humerus to determine the feasibility of avoiding en bloc resections for large giant cell tumors in this location. We evacuated the tumor through curettage and then used burring (unless the remaining cavity was thinned and at risk for fracture) and phenolization, followed by packing of the defect with allograft cancellous bone. The mean age of the patients at surgery was 30 years, and the minimum followup was 2.5 years (mean, 5.6 years; range, 2.5-9.7 years). One of the six patients had local recurrence 1.2 years postoperatively and was treated with repeat intralesional surgery with no additional recurrence 5 years later. No other patient required additional treatment, had pulmonary metastases develop, or had progression to osteoarthritis. The mean Musculoskeletal Tumor Society and Toronto Extremity Salvage Score functional scores at last followup were 26 of 30 (range, 21-30) and 95% (range, 90%-100%), respectively. These functional scores are higher than reported scores for patients with segmental resection and reconstruction of the proximal humerus. Level of Evidence: Level IV, therapeutic study. See the Guidelines for Authors for a complete description of levels of evidence.

Protein-protein docking is a challenging computational problem in functional genomics, particularly when one or both proteins undergo conformational change(s) upon binding. The major challenge is to define scoring function soft enough to tolerate these changes and specific enough to distinguish between near-native and "misdocked" conformations. Using a linear programming technique, we derived protein docking potentials (PDPs) that comply with this requirement. We considered a set of 63 nonredundant complexes to this aim, and generated 400,000 putative docked complexes (decoys) based on shape complementarity criterion for each complex. The PDPs were required to yield for the native (correctly docked) structure a potential energy lower than those of all the nonnative (misdocked) structures. The energy constraints applied to all complexes led to ca. 25 million inequalities, the simultaneous solution of which yielded an optimal set of PDPs that discriminated the correctly docked (up to 4.0 Å root-mean-square deviation from known complex structure) structure among the 85 topranking (0.02%) decoys in 59/63 examined boundbound cases. The high performance of the potentials was further verified in jackknife tests and by ranking putative docked conformation submitted to CAPRI. In addition to their utility in identifying correctly folded complexes, the PDPs reveal biologically meaningful features that distinguish docking potentials from folding potentials. Proteins 2006;62:970 -981.

Phenomenological analysis of existing hydrogen bond (HB) donor and acceptor scales and apparent physical considerations have enabled the establishment of new quantitative scales of hydrogen bond basicity and acidity. Chemical structures represented by molecular graphs and the orbital electronegativities of Hinze and Jaffe are utilized as an input data. The scales obtained correlate well with several experimental solvent polarity scales such as ∑ 2 H and ∑R 2 H , pK(HB), and E T (30). To demonstrate the applicability of the new quantities, we have applied them to seven equilibrium partitioning data sets: octanol-water, hexadecanewater, chloroform-water, gas-water, gas-octanol, gas-hexadecane, and gas-chloroform partition coefficients. The hydrogen bond descriptors when supplemented by a cavity-forming term and a dipolarity term show high performance in correlations of the partition coefficients of aliphatic compounds. These new HB descriptors can be used in studying hydrogen bonding and fluid phase equilibria as well as scoring functions in ligand docking and descriptors in ADME evaluations. * Corresponding address