Doina Caragea

Predicted effector molecules in the salivary secretome of the pea aphid (Acyrthosiphon pisum): a dual transcriptomic/proteomic approach.

The relationship between aphids and their host plants is thought to be functionally analogous to plant-pathogen interactions. Although virulence effector proteins that mediate plant defenses are well-characterized for pathogens such as bacteria, oomycetes, and nematodes, equivalent molecules in aphids and other phloem-feeders are poorly understood. A dual transcriptomic-proteomic approach was adopted to generate a catalog of candidate effector proteins from the salivary glands of the pea aphid, Acyrthosiphon pisum. Of the 1557 transcript supported and 925 mass spectrometry identified proteins, over 300 proteins were identified with secretion signals, including proteins that had previously been identified directly from the secreted saliva. Almost half of the identified proteins have no homologue outside aphids and are of unknown function. Many of the genes encoding the putative effector proteins appear to be evolving at a faster rate than homologues in other insects, and there is strong evidence that genes with multiple copies in the genome are under positive selection. Many of the candidate aphid effector proteins were previously characterized in typical phytopathogenic organisms (e.g., nematodes and fungi) and our results highlight remarkable similarities in the saliva from plant-feeding nematodes and aphids that may indicate the evolution of common solutions to the plant-parasitic lifestyle.

A framework for semantic web services discovery

This paper describes a framework for ontology-based flexible discovery of Semantic Web services. The proposed approach relies on user-supplied, context-specific mappings from an user ontology to relevant domain ontologies used to specify Web services. We show how a users query for a Web service that meets certain selection criteria can be transformed into queries that can be processed by a matchmaking engine that is aware of the relevant domain ontologies and Web services. We also describe how user-specified preferences for Web services in terms of non-functional requirements (e.g., QoS) can be incorporated into the Web service discovery mechanism to generate a partially ordered list of services that meet user-specified functional requirements.

A Framework for Learning from Distributed Data Using Sufficient Statistics and Its Application to Learning Decision Trees

This paper motivates and precisely formulates the problem of learning from distributed data; describes a general strategy for transforming traditional machine learning algorithms into algorithms for learning from distributed data; demonstrates the application of this strategy to devise algorithms for decision tree induction from distributed data; and identifies the conditions under which the algorithms in the distributed setting are superior to their centralized counterparts in terms of time and communication complexity; The resulting algorithms are provably exact in that the decision tree constructed from distributed data is identical to that obtained in the centralized setting. Some natural extensions leading to algorithms for learning from heterogeneous distributed data and learning under privacy constraints are outlined.

A massive expansion of effector genes underlies gall-formation in the wheat pest Mayetiola destructor.

Gall-forming arthropods are highly specialized herbivores that, in combination with their hosts, produce extended phenotypes with unique morphologies [1]. Many are economically important, and others have improved our understanding of ecology and adaptive radiation [2]. However, the mechanisms that these arthropods use to induce plant galls are poorly understood. We sequenced the genome of the Hessian fly (Mayetiola destructor; Diptera: Cecidomyiidae), a plant parasitic gall midge and a pest of wheat (Triticum spp.), with the aim of identifying genic modifications that contribute to its plant-parasitic lifestyle. Among several adaptive modifications, we discovered an expansive reservoir of potential effector proteins. Nearly 5% of the 20,163 predicted gene models matched putative effector gene transcripts present in the M. destructor larval salivary gland. Another 466 putative effectors were discovered among the genes that have no sequence similarities in other organisms. The largest known arthropod gene family (family SSGP-71) was also discovered within the effector reservoir. SSGP-71 proteins lack sequence homologies to other proteins, but their structures resemble both ubiquitin E3 ligases in plants and E3-ligase-mimicking effectors in plant pathogenic bacteria. SSGP-71 proteins and wheat Skp proteins interact in vivo. Mutations in different SSGP-71 genes avoid the effector-triggered immunity that is directed by the wheat resistance genes H6 and H9. Results point to effectors as the agents responsible for arthropod-induced plant gall formation.

An empirical study on using the national vulnerability database to predict software vulnerabilities

Software vulnerabilities represent a major cause of cybersecurity problems. The National Vulnerability Database (NVD) is a public data source that maintains standardized information about reported software vulnerabilities. Since its inception in 1997, NVD has published information about more than 43,000 software vulnerabilities affecting more than 17,000 software applications. This information is potentially valuable in understanding trends and patterns in software vulnerabilities, so that one can better manage the security of computer systems that are pestered by the ubiquitous software security flaws. In particular, one would like to be able to predict the likelihood that a piece of software contains a yet-to-be-discovered vulnerability, which must be taken into account in security management due to the increasing trend in zero-day attacks. We conducted an empirical study on applying data-mining techniques on NVD data with the objective of predicting the time to next vulnerability for a given software application. We experimented with various features constructed using the information available in NVD, and applied various machine learning algorithms to examine the predictive power of the data. Our results show that the data in NVD generally have poor prediction capability, with the exception of a few vendors and software applications. By doing a large number of experiments and observing the data, we suggest several reasons for why the NVD data have not produced a reasonable prediction model for time to next vulnerability with our current approach.

Information extraction and integration from heterogeneous, distributed, autonomous information sources - a federated ontology-driven query-centric approach

This paper motivates and describes the data integration component of INDUS (intelligent data understanding system) environment for data-driven information extraction and integration from heterogeneous, distributed, autonomous information sources. The design of INDUS is motivated by the requirements of applications such as scientific discovery, in which it is desirable for users to be able to access, flexibly interpret, and analyze data from diverse sources from different perspectives in different contexts. INDUS implements a federated, query-centric approach to data integration using user-specified ontologies.

Decision Tree Induction from Distributed Heterogeneous Autonomous Data Sources

With the growing use of distributed information networks, there is an increasing need for algorithmic and system solutions for data-driven knowledge acquisition using distributed, heterogeneous and autonomous data repositories. In many applications, practical constraints require such systems to provide support for data analysis where the data and the computational resources are available. This presents us with distributed learning problems. We precisely formulate a class of distributed learning problems; present a general strategy for transforming traditional machine learning algorithms into distributed learning algorithms; and demonstrate the application of this strategy to devise algorithms for decision tree induction (using a variety of splitting criteria) from distributed data. The resulting algorithms are provably exact in that the decision tree constructed from distributed data is identical to that obtained by the corresponding algorithm when in the batch setting. The distributed decision tree induction algorithms have been implemented as part of INDUS, an agent-based system for data-driven knowledge acquisition from heterogeneous, distributed, autonomous data sources.

Towards Collaborative Environments for Ontology Construction and Sharing

Ontologies that explicitly identify objects, properties, and relationships in specific domains are essential for collaborations that involve sharing of data, knowledge, or resources among autonomous individuals. Against this background, this paper motivates the need for collaborative environments for ontology construction, sharing, and usage; identifies the desiderata of such environments; and proposes package based description logics (P-DL) that extend classic description logic (DL) based ontology languages to support modularity and (selective) knowledge hiding. In P-DL, each ontology consists of packages (or modules) with well-defined interfaces. Each package encapsulates a closely related set of terms and relations between terms. Together, these terms and relations represent the ontological commitments about a small, coherent part of the universe of discourse. Packages can be hierarchically nested, thereby imposing an organizational structure on the ontology. Package-based ontologies also allow creators of packages to exert control over the visibility of each term or relation within the package thereby allowing the selective sharing (or conversely, hiding) of ontological commitments captured by a package.

Gaining insights into support vector machine pattern classifiers using projection-based tour methods

This paper discusses visual methods that can be used to understand and interpret the results of classification using support vector machines (SVM) on data with continuous real-valued variables. SVM induction algorithms build pattern classifiers by identifying a maximal margin separating hyperplane from training examples in high dimensional pattern spaces or spaces induced by suitable nonlinear kernel transformations over pattern spaces. SVM have been demonstrated to be quite effective in a number of practical pattern classification tasks. Since the separating hyperplane is defined in terms of more than two variables it is necessary to use visual techniques that can navigate the viewer through high-dimensional spaces. We demonstrate the use of projection-based tour methods to gain useful insights into SVM classifiers with linear kernels on 8-dimensional data.

Learning Classifiers from Semantically Heterogeneous Data

Semantically heterogeneous and distributed data sources are quite common in several application domains such as bioinformatics and security informatics. In such a setting, each data source has an associated ontology. Different users or applications need to be able to query such data sources for statistics of interest (e.g., statistics needed to learn a predictive model from data). Because no single ontology meets the needs of all applications or users in every context, or for that matter, even a single user in different contexts, there is a need for principled approaches to acquiring statistics from semantically heterogeneous data. In this paper, we introduce ontology-extended data sources and define a user perspective consisting of an ontology and a set of interoperation constraints between data source ontologies and the user ontology. We show how these constraints can be used to derive mappings from source ontologies to the user ontology. We observe that most of the learning algorithms use only certain statistics computed from data in the process of generating the hypothesis that they output. We show how the ontology mappings can be used to answer statistical queries needed by algorithms for learning classifiers from data viewed from a certain user perspective.