A comparative study of rank aggregation methods for partial and top ranked lists in genomic applications

Learning the true rank ordering among objects by aggregating a set of expert opinion rank order lists is an important and ubiquitous problem in many applications ranging from social choice theory to recommendation systems and search aggregation. We study the problem of unsupervised rank aggregation where no ground truth ordering information in available, neither about the true preference ordering among any set of objects nor about the quality of individual rank lists. Aggregating the often inconsistent and poor quality rank lists in such an unsupervised manner is a highly challenging problem, and standard consensus-based methods fall short in terms of both quality and scalability. In this manuscript we propose a novel framework to bypass these issues by using object attributes to augment the standard rank aggregation framework. We design algorithms that learn joint models on both rank lists and object features to obtain an aggregated rank ordering that is more accurate and robust, and also helps weed out rank lists of dubious validity. We validate our techniques on synthetic datasets where our algorithm is able to estimate the true rank ordering even when the rank lists are corrupted. Experiments on three real datasets, MQ2008, MQ2007 and OHSUMED, show that using object features can result in significant improvement in performance over existing rank aggregation methods that do not use object information. Furthermore, when at least some of the rank lists are of high quality, our methods are able to effectively exploit such information to output an aggregated rank ordering of high accuracy.

Rank aggregation is a popular problem that combines different ranked lists from various sources (frequently called voters or judges), and generates a single aggregated list with improved ranking of its items. In this context, a portion of the existing methods attempt to address the problem by treating all voters equally. Nevertheless, several related works proved that the careful and effective assignment of different weights to each voter leads to enhanced performance. In this article, we introduce an unsupervised algorithm for learning the weights of the voters for a specific topic or query. The proposed method is based on the fact that if a voter has submitted numerous elements which have been placed in high positions in the aggregated list, then this voter should be treated as an expert, compared to the voters whose suggestions appear in lower places or do not appear at all. The algorithm iteratively computes the distance of each input list with the aggregated list and modifies the weights of the voters until all weights converge. The effectiveness of the proposed method is experimentally demonstrated by aggregating input lists from six TREC conferences.CCS CONCEPTS• Information systems → Rank aggregation; •Theory of computation → Unsupervised learning and clustering.

In credit risk evaluation the accuracy of a classifier is very significant for classifying the high-risk loan applicants correctly. Feature selection is one way of improving the accuracy of a classifier. It provides the classifier with important and relevant features for model development. This study uses the ensemble of multiple feature ranking techniques for feature selection of credit data. It uses five individual rank based feature selection methods. It proposes a novel rank aggregation algorithm for combining the ranks of the individual feature selection methods of the ensemble. This algorithm uses the rank order along with the rank score of the features in the ranked list of each feature selection method for rank aggregation. The ensemble of multiple feature selection techniques uses the novel rank aggregation algorithm and selects the relevant features using the 80%, 60%, 40% and 20% thresholds from the top of the aggregated ranked list for building the C4.5, MLP, C4.5 based Bagging and MLP based Bagging models. It was observed that the performance of models using the ensemble of multiple feature selection techniques is better than the performance of 5 individual rank based feature selection methods. The average performance of all the models was observed as best for the ensemble of feature selection techniques at 60% threshold. Also, the bagging based models outperformed the individual models most significantly for the 60% threshold. This increase in performance is more significant from the fact that the number of features were reduced by 40% for building the highest performing models. This reduces the data dimensions and hence the overall data size phenomenally for model building. The use of the ensemble of feature selection techniques using the novel aggregation algorithm provided more accurate models which are simpler, faster and easy to interpret.

Imputation of missing values is a strategy for handling non-responses in surveys or data loss in measurement processes, which may be more effective than ignoring the losses and omitting them. The characteristics of variables presenting missing values must be considered when choosing the imputation method to be used; in particular when the variable is a count the literature dealing with this issue is scarce. If the variable has an excess of zeros it is necessary to consider models including parameters for handling zero-inflation. Likewise, if problems of over- or under-dispersion are observed, generalizations of the Poisson, such as the Hermite or Conway–Maxwell Poisson distributions are recommended for carrying out imputation. The aim of this study was to assess the performance of various regression models in the imputation of a discrete variable based on Poisson generalizations, in comparison with classical counting models, through a comprehensive simulation study considering a variety of scenarios and a real data example. To do so we compared the results of estimations using only complete data, and using imputations based on the most common count models. The COMPoisson distribution provides in general better results in any dispersion scenario, especially when the amount of missing information is large.

<title>Abstract</title> In this paper, a novel distribution called record-based transmuted exponential power is proposed by using the record-based transmutation map. The suggested distribution serves as a flexible alternative to the exponential power and its modified ones, offering improved adaptability in modeling skewed or heavytailed real-life data structures. Some distributional properties such as moments, median, hazard function, R´enyi entropy, and stochastic ordering are examined in detail. In point estimation, nine different estimators, namely maximum likelihood, least squares, weighted least squares, Anderson-Darling and right tail Anderson Darling, Cram´er-von Mises, maximum product spacings, minimum spacing absolute distance and minimum spacing absolute log-distance estimators are used to estimate the parameters of proposed distribution. Then, a comprehensive Monte-Carlo simulation study is designed to compare these estimators in terms of estimation procedure. Among the nine estimation methods considered, the minimum spacing absolute distance estimator consistently yielded the lowest mean squared error across various sample sizes and parameter settings. Three practical data examples are provided to assess the fits of suggested model as well as its competitor ones. Three real-world data examples are considered assess both the potential of the introduced model for modeling real-life data and its superiority over competing distributions, the main one being exponential power. Based on goodness-of-fit measures such as −2 log ℓ, Kolmogorov-Smirnov statistics, Anderson-Darling statistics, Cramer-von-Mises statistics and their p-values, the RBTEP distribution outperformed classical distributions including exponential power, Weibull, transmuted Weibull, transmuted record type Weibull, and transmuted Lindley in modeling the failure time, aircraft and recidivism time datasets.

Compendia of large-scale datasets made available in public repositories provide a precious opportunity to discover new biomedical phenomena and to fill gaps in our current knowledge. In order to foster novel insights it is necessary to ensure that these data are made readily accessible to research investigators in an interpretable format. Here we make a curated, public, collection of transcriptome datasets relevant to human placenta biology available for further analysis and interpretation via an interactive data browsing interface. We identified and retrieved a total of 24 datasets encompassing 759 transcriptome profiles associated with the development of the human placenta and associated pathologies from the NCBI Gene Expression Omnibus (GEO) and present them in a custom web-based application designed for interactive query and visualization of integrated large-scale datasets (http://placentalendocrinology.gxbsidra.org/dm3/landing.gsp). We also performed quality control checks using relevant biological markers. Multiple sample groupings and rank lists were subsequently created to facilitate data query and interpretation. Via this interface, users can create web-links to customized graphical views which may be inserted into manuscripts for further dissemination, or e-mailed to collaborators for discussion. The tool also enables users to browse a single gene across different projects, providing a mechanism for developing new perspectives on the role of a molecule of interest across multiple biological states. The dataset collection we created here is available at: http://placentalendocrinology.gxbsidra.org/dm3.

For a given query and a set of images ranked lists retrieved from multiple search engines, the metasearch technique aims at combining these lists to build an unified ranking with improved relevance. Rank aggregation is an approach that has been widely used to support this task. This paper investigates the use of rank aggregation methods in the metasearch scenario for diverse image retrieval. Although metasearch systems are usually driven by the relevance of the final result, the impact on diversification has also been analyzed. The experimental findings suggest metasearch based on rank aggregation allows significant improvements, both in terms of relevance and diversity.

Existing methods dealing with the problem of rank aggregation in the context of meta-search in information retrieval are considered as a passive learner machine and suffer in the presence of unreliable ranking lists. This paper proposes a novel approach which selects the most informative features and instances to be labeled from which the model will learn. To train an efficient ranking aggregation model from few labeled instances, we use Multiple Hyperplane Ranker algorithm in an active learning environment. Experimental results on the OHSUMED dataset show that our method outperforms the existing methods.

The interplay between genetic and environmental factors plays a significant role in interindividual variation in immune and inflammatory responses. The availability of high-throughput low-cost genotyping and next-generation sequencing has revolutionized our ability to identify human genetic variation and understand how this varies within and between populations, and the relationship with disease. In this review, we explore the potential of genomics for patient benefit, specifically in the diagnosis, prognosis and treatment of inflammatory and immune-related diseases. We summarize the knowledge arising from genetic and functional genomic approaches, and the opportunity for personalized medicine. The review covers applications in infectious diseases, rare immunodeficiencies and autoimmune diseases, illustrating advances in diagnosis and understanding risk including use of polygenic risk scores. We further explore the application for patient stratification and drug target prioritization. The review highlights a key challenge to the field arising from the lack of sufficient representation of genetically diverse populations in genomic studies. This currently limits the clinical utility of genetic-based diagnostic and risk-based applications in non-Caucasian populations. We highlight current genome projects, initiatives and biobanks from diverse populations and how this is being used to improve healthcare globally by improving our understanding of genetic susceptibility to diseases and regional pathogens such as malaria and tuberculosis. Future directions and opportunities for personalized medicine and wider application of genomics in health care are described, for the benefit of individual patients and populations worldwide.

The Neyman–Pearson (NP) classification paradigm addresses an important binary classification problem where users want to minimize type II error while controlling type I error under some specified level α, usually a small number. This problem is often faced in many genomic applications involving binary classification tasks. The terminology Neyman–Pearson classification paradigm arises from its connection to the Neyman–Pearson paradigm in hypothesis testing. The NP paradigm is applicable when one type of error (e.g., type I error) is far more important than the other type (e.g., type II error), and users have a specific target bound for the former. In this chapter, we review the NP classification literature, with a focus on the genomic applications as well as our contribution to the NP classification theory and algorithms. We also provide simulation examples and a genomic case study to demonstrate how to use the NP classification algorithm in practice.

The kuruma prawn, Marsupenaeus japonicus, is one of the most cultivated and consumed species of shrimp. However, very few molecular genetic/genomic resources are publically available for it. Thus, the characterization and distribution of simple sequence repeats (SSRs) remains ambiguous and the use of SSR markers in genomic studies and marker-assisted selection is limited. The goal of this study is to characterize and develop genome-wide SSR markers in M. japonicus by genome survey sequencing for application in comparative genomics and breeding. A total of 326 945 perfect SSRs were identified, among which dinucleotide repeats were the most frequent class (44.08%), followed by mononucleotides (29.67%), trinucleotides (18.96%), tetranucleotides (5.66%), hexanucleotides (1.07%), and pentanucleotides (0.56%). In total, 151 541 SSR loci primers were successfully designed. A subset of 30 SSR primer pairs were synthesized and tested in 42 individuals from a wild population, of which 27 loci (90.0%) were successfully amplified with specific products and 24 (80.0%) were polymorphic. For the amplified polymorphic loci, the alleles ranged from 5 to 17 (with an average of 9.63), and the average PIC value was 0.796. A total of 58 256 SSR-containing sequences had significant Gene Ontology annotation; these are good functional molecular marker candidates for association studies and comparative genomic analysis. The newly identified SSRs significantly contribute to the M. japonicus genomic resources and will facilitate a number of genetic and genomic studies, including high density linkage mapping, genome-wide association analysis, marker-aided selection, comparative genomics analysis, population genetics, and evolution.

With the rapid accumulation of microarray data, it is a significant challenge to integrate available data sets addressing the same biological questions that can provide more samples and better experimental results. Sometimes, different microarray platforms make it difficult to effectively integrate data from several studies and there is no consensus on which method is the best to produce a single and unified data set. Methods using median rank score, quantile discretization and standardization (which directly combine rescaled gene expression values) and meta-analysis (which combine the results of individual studies at the interpretative level) are reviewed. Real data examples downloaded from GEO are used to compare the performance of these methods and to evaluate if the combined data set detects more reliable information from the separated data sets or not.

It is already more than 10 years after the completion of Human Genome Project. However, the mechanism about how genomic information guides the gene expression in a particular space and time remains to be elucidated. Structure determines function is a consensus that we understand the laws of nature. In biology, this consensus is not only applicable to RNA, protein and other small-scale organic molecules; the same should also apply to macromolecules such as the entire DNA composition of the whole genome. Recent studies have shown that three-dimensional (3D) structure of the genome has an important impact on the regulation of gene transcription and other nuclear functions. Therefore, the study of the 3D structure and function of the whole genome (referred to as 3D genomics) has become a new trend of development in genomics. With the consideration of the genome sequence, gene structure and regulatory elements at the same time, 3D genomics studies the 3D structure of the genome in the nucleus, and its impact on gene transcription, replication, repair, regulation and others biological processes. With the progress of large-scale genome sequencing technology, we already have the relevant technologies such as ChIA-PET and Hi-C for the study of 3D genomics. Preliminary results showed the influence of genome 3D structures to genome functions, which provides the basis for further study of genome structure and function in the spatial nuclear space. This paper summarizes the emergence of 3D genomics, lists the issues and the current progress in this new field, and discusses possible applications of 3D genomics to human health and agriculture researches. We anticipate that 3D genomics will provide novel insights into the genome regulatory functions and provide new opportunities of improving human life qualities.

ABSTRACTCarbapenem-resistant Klebsiella pneumoniae (CRKP) has disseminated globally and is difficult to treat, causing increased morbidity and mortality rates in critically ill patients. We conducted a multicenter cross-sectional study of intensive care unit (ICU) inpatients in 78 hospitals to investigate the prevalence and molecular characteristics of CRKP in Henan Province, China, a hyperepidemic region. A total of 327 isolates were collected and downsampled to 189 for whole-genome sequencing. Molecular typing revealed that sequence type 11 (ST11) of clonal group 258 (CG258) was predominant (88.9%, n = 168), followed by ST2237 (5.8%, n = 11) and ST15 (2.6%, n = 5). We used core genome multilocus sequence typing (cgMLST) to further classified the population into 13 subtypes. Capsule polysaccharide (K-antigen) and lipopolysaccharide (LPS; O-antigen) typing revealed that K64 (48.1%, n = 91) and O2a (49.2%, n = 93) were the most common. We studied isolates collected from both the airway and the gut of the same patients and showed that intestinal carriage was associated with respiratory colonization (odds ratio = 10.80, P < 0.0001). Most isolates (95.2%, n = 180) showed multiple drug resistance (MDR), while 59.8% (n = 113) exhibited extensive drug resistance (XDR), and all isolates harbored either blaKPC-2 (98.9%, n = 187) or blaCTX-M and blaSHV extended-spectrum beta-lactamases (ESBLs) (75.7%, n = 143). However, most were susceptible to ceftazidime-avibactam (CZA) (94.7%, n = 179) and colistin (97.9%, n = 185). We found mgrB truncations in isolates conferring resistance to colistin and mutations in blaSHV and OmpK35 and OmpK36 osmoporins in CZA-resistant isolates. Using a regularized regression model, we found that the aerobactin sequence type and the salmochelin sequence type, among others, were predictors of the hypermucoviscosity phenotype.IMPORTANCE In this study, we address the ongoing epidemic of carbapenem-resistant Klebsiella pneumoniae, a critical threat to public health. The alarming genotypic and phenotypic convergence of multidrug resistance and virulence highlights the increasingly aggravated threat posed by K. pneumoniae. This calls for a combined effort of physicians and scientists to study the potential mechanisms and establish guidelines for antimicrobial therapies and interventions. To this end, we have conducted a genomic epidemiology and characterization study using isolates collected in a coordinated effort of multiple hospitals. Innovative biological discoveries of clinical importance are made and brought to the attention of clinical researchers and practitioners. This study presents an important advance in the application of genomics and statistics to recognize, understand, and control an infectious disease of concern.

Quadratic forms of multivariate normal variables play a critical role in statistical applications, particularly in genomics and bioinformatics. However, accurately computing small right-tail probabilities ( p -values) for large-scale quadratic forms is computationally challenging due to the intractability of their probability distributions, as well as significant numerical constraints and computational burdens. To address these problems, we propose MCMC-CE, an innovative algorithm that integrates Markov Chain Monte Carlo (MCMC) sampling with the cross-entropy (CE) method, coupled with leading-eigenvalue extraction and Satterthwaite-type approximation techniques. Our approach efficiently estimates small p -values for quadratic forms with their ranks exceeding 10,000. Through extensive simulation studies and real-world applications in genomics, including genome-wide association studies and pathway enrichment analyses, our method demonstrates advantageous numerical accuracy and computational reliability compared to existing approaches such as Davies', Imhof's, Farebrother's, Liu-Tang-Zhang's and saddlepoint approximation methods. MCMC-CE provides a robust and scalable solution for accurately computing small p -values for quadratic forms, facilitating more precise statistical inference in large-scale genomic studies.